Introduction to ESSENTIAL CINEMA (December 2002)

A slightly different version of the Introduction to my 2004 collection, Essential Cinema: On the Necessity of Film Canons. — J.R.



As the son and grandson of small-town exhibitors — a legacy explored in detail in my first book, Moving Places: A Life at the Movies (1980/1995) — I find it difficult to pinpoint with any exactitude when my film education started. But I can recall two pivotal early steps during my freshman year at New York University in 1961, when I was an English major still aspiring to become a professional novelist: taking the first and only film course I’ve ever had in my life and purchasing my first film magazine.

The course was an introductory survey taught by the late Haig Manoogian, who was serving as Martin Scorsese’s mentor in production courses around the same time. For me, it mainly afforded me my first opportunity to see The Birth of a Nation, The Last Laugh, and a few other film history staples; since I had no interest in making movies — or at this point in writing about them — I couldn’t work up much enthusiasm for such matters as “story values” that Manoogian tended to emphasize. I was also skeptical about the inclusion in the curriculum of the 40s movie Champion — something I’d already seen in the 50s as a revival at one of my family’s theaters — and the exclusion of Citizen Kane, which I’d seen a year or so earlier at boarding school in Vermont. When I asked Manoogian about this omission after class one day, he reiterated the standard line about Kane that I’d already encountered in Arthur Knight’s The Liveliest Art and Paul Rotha and Richard Griffith’s The Film Till Now, two of the few standard reference books available at the time — that Orson Welles’ film was basically uncinematic and therefore impressive only to amateurs who understood little about the medium. (It took me years to grasp that Knight, Griffith, and Manoogian were all American film industry apologists to varying degrees, and that their hostility towards Kane probably bore some relation to this position during the 50s and 60s — when the film was still perceived chiefly as being in opposition to the industry, before Pauline Kael maintained it was just another newspaper comedy.) In any event, this position was so irritating to me at the time that I got Manoogian’s grudging permission to call my final paper “In Defense of Citizen Kane,” and won a grudging B+ from him for my extended polemical efforts. (A few portions of this paper wound up getting recycled in my first extended article for Film Comment — a review of Kael’s “Raising Kane”. [1])

In the long as well as the short run, the film magazine I purchased proved to be much more consequential in developing my taste in film. It was the Winter 1961-62 issue of Sight and Sound, and I partially bought it because it had reviews of films I’d recently been seeing or hearing about in New York such as Last Year at Marienbad and La notte. (By this time, I’d already given myself a crash course in the French New Wave and the Italian film renaissance during my trips to and through New York over two years of boarding school.) But the bonus in that particular issue, which proved to be even more important to me in some ways, was the results of an international poll of film critics concerning the greatest films ever made. Citizen Kane, I was happy to discover, placed first, and I was astonished to discover in second place L’avventura – a film by Michelangelo Antonioni preceding La notte that I had only just discovered and was still trying to process. Determined to learn more about film history than Manoogian’s course could offer, I vowed to see as many films on the list as I could, and for the next several years proceeded like a butterfly collector, dutifully underlining each title in that issue of Sight and Sound as soon as I saw the film. It was a better way of surveying the lay of the land, I quickly discovered, than the indexes of Knight and Rotha/Griffith, because it led me towards objects of critical veneration more than historical markers —- objects that would eventually be joined by those found in Andrew Sarris’s The American Cinema (see below) and Noël Burch’s Theory of Film Practice, among other essential guidebooks —-and it brought me more directly into the process of forming my own values and tastes as I used it to make my own discoveries. Some critical favorites on the list proved to be disappointments, others were greater than I even hoped for, but in both cases these responses represented not so much endpoints as the beginnings of evaluations and re-evaluations that would continue over decades, and that are still taking place.

Academic film study had barely started by then, and several years would pass before I was informed that ranking movies qualitatively in such a canonical manner was foolish and naïve–that I was subscribing to myths about solitary geniuses and timeless art that only showed how provincial I was. Further down the road, these objections became amplified by demurrals about dead white males and bourgeois complacency, though by the time these came along, I’m afraid the damage to me had already been done: in 1968 I had quit graduate school to accept a job of editing a book of film criticism and, by the next year, to move from New York to Paris; and during all this period I stuck to the Sight and Sound list and others like it in order to carry out my self-education in film. Little did I realize that by seeing all these films willy-nilly and helter-skelter at the Paris Cinémathèque and elsewhere, I was fleeing from various theoretical and sociological considerations and simply having too much irresponsible fun by finding pleasure in most of them.

Roughly a quarter of a century later, while putting together a proposal for a first collection of my pieces at the invitation of the University of California Press’s Ed Dimendberg, I outlined a section entitled “Masterpieces,” which Ed urged me to change to “Touchstones”. The sort of arguments that made the very notion of a masterpiece academically unacceptable by the mid-90s can’t be easily synopsized, but I don’t think it would be misleading to say that a fundamental distrust of art, often unacknowledged as such, played a significant role in this unraveling.

Indeed, the unraveling of literary canons over the same period in what used to be my own academic field has ultimately proved to be even more decisive in certain ways, making me all the more happy that I opted for journalism over teaching in the late 60s. The havoc wreaked on literary studies through the virtual outlawing of canons has been discussed at some length by Harold Bloom in The Western Canon (2), and although Bloom has inspired as well as informed some of my own strategies in this collection, I should stress at the outset some major differences in our assumptions and approaches. Most of them can be found in the following check list:

1. The cinematic canon I’m proposing is not Western — and given that the unacknowledged canons currently ruling film culture, which I’ll speak about shortly, are all unabashedly Western and North American, I can’t see much merit in perpetuating such a habit, especially in an art and medium such as film that is ruled by linguistic differences considerably less than literature.

2. My canon, unlike Bloom’s, is not founded exclusively on aesthetic considerations. For far too long, in my opinion, an overall consumerist orientation in our culture has dictated the assumption that all films can be subsumed under the categories of either art or entertainment — a needlessly restrictive notion that has lasted as long as it has only because it has rarely been examined. The fact is, when images of Iranians in the mass media are almost invariably reductive stereotypes and images of Iranians in many Iranian films are more varied and accurate, this alone gives Iranian films a certain value apart from their credentials as art and/or entertainment — a value that we can’t afford to ignore, even though we haven’t yet learned how to validate this material for such a reason.

3. Unlike Bloom, I regard canon formation as an active process of selection rather than a passive one of reportage. Next to the history of literature, the history of film is minuscule, and one could argue as a consequence that it’s much too early to speak about film canons in the same way we might speak about literary canons. I would certainly agree with this premise. However, Bloom, asserting that “cultural prophecy is always a mug’s game” and that “critics do not make canons, any more than resentful networks [deriving from what Bloom calls a “School of Resentment”] can create them” (3), generally treats literary canons as descriptive rather than prescriptive, and my treatment of film canons is the reverse of this. I should add that Bloom’s avoidance of such matters as market considerations dictating what books get reprinted and why limit for me the value of his descriptive approach.

4. Consequently, I differ with Bloom’s view of the political and social aspects of canonizing. “Cultural criticism is another dismal social science,” Bloom writes (4), “but literary criticism, as an art, always was and always will be an elitist phenomenon.” My assumption is that criticism as an art can include cultural as well as literary (and film) criticism, and that all these forms of criticism can be either elitist or populist. Furthermore, notwithstanding Bloom’s caveat that art doesn’t make people into better citizens, I would argue that information can on occasion do precisely that, and that art and information never can and never should be viewed as wholly independent entities. Bearing all this in mind, the film canon proposed at the end of this book can’t be limited simply or exclusively to art, entertainment, information, communication, or models of thought or perception, though all these categories are applicable at one time or another. Insofar as books and printed matter in general can’t be limited to art and entertainment, it would be both presumptuous and self-defeating to claim something comparable for everything that exists on film. Consequently, the value of film canons can’t be limited to a need for priorities involving art and entertainment, even if these priorities understandably tend to dominate our discussions of film, mine included.


When I described a particular issue of Sight and Sound as being “more consequential” in the development of my film taste than an introductory film course at NYU, I don’t mean to imply that the former proposed a particular film canon while the latter didn’t. Indeed, despite the frequent–or at least implicit–claim of film professors that they don’t buy into the canon-formation game, I would argue, on the contrary, that canons have never left us; all that’s really gone away is their acknowledgment and academics agreeing to play a conscious and active role in promoting them. (5)

I suspect that part of what turned many academics against canons was the threat posed by one particular canon formulated by an American film critic who subsequently became an academic himself, Andrew Sarris. It’s hard to overestimate the multifaceted impact of Sarris’s The American Cinema when it first appeared during the 60s — initially in the pages of a New York “underground” magazine called Film Culture and subsequently as a revised and expanded book. Here, in one fell swoop, was not so much an evaluative reordering and fresh ranking of Hollywood feature directors as a grand gesture declaring that it was meaningful to order, rank, and classify them in the first place, with style taking precedence over other considerations. The threat posed to critics outside academia was so palpable that most film lovers first heard about Sarris’s so-called auteur theory only when it was attacked–most famously by Pauline Kael in the pages of Film Quarterly (6), but also by Dwight Macdonald, the film critic for Esquire, who was so incensed that he and Sarris contributed to the same feature of Film Quarterly called “Films of the Quarter” that he promptly resigned as a contributor once Sarris was brought on.

Inside academia, it would be safe to say that Sarris’s implied program was at first much more widely embraced as a way of organizing and defining the syllabus for film courses, and it wasn’t long before more theoretically inclined film academics such as Peter Wollen were proposing theoretically upgraded versions of the auteur theory. (7) But eventually Sarris’s methodology began to be mistrusted as too facile, too romantic, too apolitical, too redolent of film buffery, and too much grounded in art as opposed to the social sciences — the latter especially significant insofar as Sarris was himself reacting against some of the social and political inflections of previous film writers such as Lewis Jacobs and Siegfried Kracauer, brandishing the more obvious cinephilia of André Bazin and his disciples at Cahiers du Cinéma as more useful, pleasure-seeking, and energizing critical models. Paradoxically, the writer whose critical authority ultimately superseded that of both Sarris and Bazin in academia, Roland Barthes, was pleasure-seeking and energizing in his own right, but not in relation to film; if anything, Barthes’ relative distrust of film pleasure is what seemed to recommend him to film academics over everything else.

Having taught film for well over a decade myself, I can certainly see where part of this distrust of pleasure comes from, even after one rules out (as one shouldn’t) the puritanical factors which make this distrust quintessentially American: the frivolous associations with movies fostered by the studios and related institutions, such as those involving fans. For it’s worth stressing that during the same time that academic film study was canonizing theorists rather than films and filmmakers, the latter two were becoming canonized with increasing frivolity by the various promotional arms of the studios–starting with the Academy Awards, which has become increasingly prominent over the years, and continuing with the practice since the 80s of listing the biggest box office grossers week after week in newspapers, in magazines, and on TV.

In both these cases, popularity contests as largely determined–sometimes misleadingly or erroneously–by box office receipts have comprised the only film canons taken seriously in American culture. And the disinclination of American film academics to offer any alternative canons has continued to give the industry a clear and unchallenged playing field, assisted by such recent promotional campaigns as the American Film Institute’s various polls relating to the 100 greatest American films, stars, comedies, and so on. The restriction of such lists to Hollywood features only begins to describe the promotional aims of promoting particular products coming exclusively from the studios, mainly within the narrow range of what’s already available and out on the market. To cite only one random example of what I mean, my own choice of the greatest American film comedy, Charlie Chaplin’s Monsieur Verdoux, was missing from the 500 candidates proposed by the AFI on their ballot, no doubt because it qualified too clearly as independent — or else because the people composing the ballot were unfamiliar with it, which perhaps comes down to the same thing. (8)

It’s frequently remarked nowadays that young people today know next to nothing about the classics of American cinema — for instance, westerns by John Ford and Howard Hawks; crime pictures of the 30s; horror films produced by Val Lewton; or comedies by Chaplin, Keaton, and Preston Sturges. What’s much less frequently remarked upon is why this should be the case, particularly in relation to previous generations. Merely citing the absence of canons apart from those put together by ill-informed studio publicists (who typically don’t even have a clear sense of what could be found in the studio vaults) doesn’t suffice to account for a problem made only worse by the decimation of state funding for the arts, the downgrading of film discourse in general (both within the journalistic sectors, which increasingly show preference for promotion over criticism, and within the academic sectors, which increasingly show preference for the social sciences over art), and the cheap nostalgia of older film fans who refuse to examine or interrogate the current situation any further than arrogantly declaring their own generation and its canons superior to any of those succeeding them.

One fortunate development in recent years that has the potentiality for breaking down some of the barriers that keep academic film study and mainstream film culture in mutual ignorance about and alienation from one another is the DVD “bonus”. Specifically, what I have in mind is some of the most illuminating scholarly materials made available to the public on DVDs that are now readily available in outlets such as Tower Records. I hasten to add that, despite the fact that this book carries a university press imprint, my own main stomping ground as a critic has been an alternative weekly, the Chicago Reader, where I’ve been based since 1987 —- a paper in which I’ve consciously sought to bridge academic and mainstream perspectives on film. Bearing this in mind, in the 60 essays found in this collection, most of them derived from Reader pieces, I’ve pared away certain journalistic details regarding specific venues that no longer seem functional, retained others that seem relevant for historical, polemical, and/or rhetorical reasons, and, in some cases, added information, afterthoughts, or updates where these appear to be useful. (I’ve also eliminated all my original warnings about giving away the plots of certain films.)

Readers of my two previous collections, Placing Movies and Movies as Politics, as well as Movie Wars, will find that some of the arguments or information found here are already familiar from those earlier books, and indeed, some redundancies can be found among the 60 pieces in this book. I apologize for the recycling, but sometimes it was difficult to remove these repetitions without disturbing the arguments of the essays in question. And I’m afraid this also sometimes applies to the few cases where I deal more than once in this volume with the same filmmakers—-Altman, Davies, Demy, Godard, Kubrick, Maddin, and Ruiz.                                                          ***

In proposing that we – that is, the readers of this book — find ways of recanonizing cinema in order to combat the reductive canons of studio publicists, I’ve organized this collection around a series of alternate strategies. Most of these strategies can be loosely categorized according to the titles of this book’s five sections. “Classics”–another word that is disliked and mistrusted in academia nowadays, almost as much as “masterpieces” (though both remain in constant use in mainstream criticism) — starts off by proposing a purely reactionary strategy: a reversion to a “classical” canon that is basically a new and improved version of the ones we used to have when canonizing could still be carried out respectably beyond the annual ten-best lists of critics and the promotional campaigns of studios and their lackeys (including some critics and professors).

As John Guillory has pointed out in a thoughtful essay (9), the word “classic” has in recent years been displaced (and essentially replaced) by “canon” in the academic world; outside that world, one might add, the term continues to be used without shame or apology. Yet according to most common usages, a film can only be “classic” or “canonical” if it’s readily available, and I have sought to challenge that assumption in a few cases in this section.

As this book goes to press, at least five of the films discussed here at length —- the color Jour de fête, Satántángo, Blush, A Tale of the Wind, and The Asthenic Syndrome –are not readily available in this country, at least on video. Practically speaking, this would exclude them from most lists that might be used to make up a syllabus or a journalistic roster of recommendations, but this seems to me an absurd criterion for establishing what does or doesn’t deserve “classic” status. Consequently I’ve kept them in, in the hopes that accounts of their importance and relevance might help to make them available again. Indeed, most traces of the polemical arguments initially used to defend the dozen films examined here have been retained, even when they cite relatively forgotten commercial films that have been mentioned for purposes of comparison or contrast, because the fact that so many of them are now barely remembered helps to support those arguments.

Another provocation: five of the films discussed in    this section qualify as “golden oldies” by having been made between the 20s and 60s, each one coming from a separate decade (Greed, M, Jour de fête, Rear Window, and The Umbrellas of Cherbourg), but the remaining seven, also discussed in chronological order, all qualify as relatively recent, having been made between 1988 and 1999.

What do we mean—-and what do we assume-—when we speak about a film classic? My capsule review of The Bicycle Thief for the Chicago Reader, written on the occasion of a recent rerelease, broaches just a few of the many issues involved:

An unemployed worker (Lamberto Maggiorani) in postwar Rome finds a job putting up movie posters after his wife pawns the family sheets to get his bicycle out of hock. But right after he starts work the bike is stolen, and with his little boy in tow he travels across the city trying to recover it. This masterpiece — whose Italian title translates as “bicycle thieves” — is generally and correctly known as one of the key works of Italian neorealism, but French critic André Bazin also recognized it as one of the great communist films. (The fact that it received the 1949 Oscar for best foreign film suggests that it wasn’t perceived widely as such over here at the time; ironically, the only thing American censors cared about was a scene in which the little boy takes a pee on the street.) The dominance of auteurist criticism over the past three decades has made this extraordinary movie unfashionable because its power doesn’t derive from a single creative intelligence, but the work of screenwriter Cesare Zavattini, director Vittorio De Sica, the nonprofessional actors, and many others is so charged with a common purpose that there’s no point in even trying to separate their achievements. This is possibly the greatest depiction of a relationship between a father and son in the history of cinema, and it’s an awesome heartbreaker. If you set it alongside something like Life Is Beautiful you get some notion of how much mainstream world cinema and its relation to reality have been infantilized over the past half century. 90 min.; in Italian with subtitles.


This partially illustrates one of the points made in Guillory’s essay: “The reduction of the text to the `voice’ of an author has been without question politically strategic in the short term, but at a certain long-term cost.” (10) Though he isn’t referring to cinema here, the downgrading of Italian neorealist films that couldn’t readily be viewed as auteurist, in both the U.S. and Europe, is a 60s phenomenon that lamentably eclipsed the reputation of films such as The Bicycle Thief and Umberto D (both memorably resurrected and appreciated in Martin Scorsese’s recent video documentary My Voyage to Italy). And two other points seem worth raising: the fact that The Bicycle Thief may indeed be a communist classic, as it was widely regarded in Europe at the time, and not simply a humanist landmark, as it was taken to be in 1949 America, is one good reason for redefining it as a classic; and similarly, a comparison of this film with a more recent Italian feature about a father and son, Life is Beautiful, may point up virtues that might have been less apparent in previous decades.



“Special Problems” foregrounds some of the complications and practical difficulties that ensue once one decides that canon formation is a desirable activity in film criticism. Much of this — though by no means all of it – involves an interrogation of some of the ideological limitations of alleged classics, an approach that more specifically informs the fourth section, “Disputable Contenders,” but figures here more generally as an inquiry into the critical methodologies that accompany canonizing. Nevertheless, “Special Problems” and “Disputable Contenders” are close enough in spirit to one another to qualify as siblings, much as “Classics” and “Filmmakers” are. On the other hand, “Other Canons, Other Canonizers,” by focusing on filmmakers who are themselves cinéphiles, critics, historians and canonizers, proposes a more self-contained set of mirror structures. And “Filmmakers” more simply focuses on the salient characteristics of artists whom I believe are worth canonizing–comprising another reactionary set of examples of old-fashioned canonizing to provide a bookend with my first section.

Assuming that this latter point needs clarifying, I’d like to point out that my use of the word “reactionary” here is mainly ironic, because I don’t think history can ever be repeated–either in film history or in the history of film reception and film appreciation. That’s why I can only laugh when one of my colleagues remarks that none of the recent films can “duplicate” or “match” or “equal” or “approximate” the masterpieces or classics of the past, because my own definition of the singularity of a major film, tautologically speaking, is that it’s singular. I should stress that the process of canonization, which the pieces of this book explore in diverse ways, should not be equated with the pantheon movie picks that I provisionally propose at the very end. The fact that in “Filmmakers” I explore the relationships between directing, writing, and acting in my pieces on Samuel Fuller and Elaine May and between the screenplays (filmed and unfilmed) and novels of Rudy Wurlitzer already begins to suggest some of the less obvious reasons why as well as how certain works deserve to be canonized. All three of these essays were originally written for a magazine published by the Screenwriters Guild, so the fact that I concentrate on writing in all three is unsurprising. What may strike some readers as more of a stretch —- that some unpublished as well as some realized Wurlitzer scripts are worthy of canonizing, in part because of their relationship to his novels, and that Fuller as well as May should be valued as a writer-performer—-only begins to suggest that the way we value works always depends on particular contexts. (The same applies to my piece on I Am Cuba, which seeks to complicate what we mean by “filmmaker” as well as canonical films.)

In this book’s coda, spurred by its own context, I’ve made a stab at proposing a particular film canon of my own — that is to say, one that’s prescriptive (and proscriptive) rather than descriptive, reflecting my own tastes and preferences, which can be regarded either as a possible (or ideal) viewing list or as a critical manifesto that can be debated. Recalling that “Pantheon Directors” was the highest plateau in Sarris’s own canon, my own “Pantheon Movie Picks” are none the less designed to be used in launching a new discussion, not in concluding an old one. And the terms “pantheon” and “picks” are intended to form a dialectic, pointing to two different kinds of critical modes, with the hopes that this book can help to build a bridge between them.

Jonathan Rosenbaum

December 2002


End Notes


1. Reprinted in Perspectives on Citizen Kane, edited by Ronald Gottesman, New York: G.K. Hall, 1996, 133-140.

2. The Western Canon: The Books and School of the Ages, New York/San Diego/London: Harcourt Brace & Co., 1994.

3. Ibid., p. 548.

4. Ibid., p. 17.

5. For a more detailed discussion of this phenomenon, see the discussion between James Naremore and Adrian Martin, “The Future of Academic Film Study,” in Movie Mutations: The Changing Face of World Cinephilia, edited by Jonathan Rosenbaum and Adrian Martin, London: British Film Institute, 2003 (forthcoming).

6. “Circles and Squares,” I Lost It At The Movies, New York: Bantam Books, 1966, 264-288. Significantly, Kael omits this essay in her final collection of selected works, For Keeps.

7. Wollen, Peter, Signs and Meaning in the Cinema, expanded edition, London: BFI Publishing, 1998, 50-78.

8. See Chapter 6 of my book Movie Wars: How Hollywood and the Media Limit What Movies We Can See (Chicago: a cappella books, 2002), 91-106, for a detailed critique of the AFI’s first poll of this kind, on the 100 greatest American films, as well as a proposed alternative list — my own first major step in canonizing prior to writing this book.

9. Guillory, John, “Canon, Syllabus, List: A Note on the Pedagogic Imaginary,” reprinted in The Best American Essays 1992, edited by Susan Sontag, New York: Ticknor & Fields, 1992, 158-180. Originally published in the journal Transition (1991).

10. Ibid., 161.

This entry was posted in Notes. Bookmark the permalink.