Originally published in Filmnews vol. 21, no. 3, April 1991, pp. 8-9. It appears here with minor corrections. Republished with the permission of the author and the then editor of Filmnews, Tina Kaufman.

The Coen brothers’ third feature is dense, paradoxical, unusually strong in both sensuousness and intellection, with striking shifts between expressionism and impressionism. It recalls various other films in many of its details but is organically distinct from all of them, and its creative amplitude is far superior to its contemporaries. Its meaning is occluded, its values problematical, and there is no “safe place” for the audience to resolve them.

The plotting is labyrinthine, while the narrative is linear. The characters talk a lot about what’s going on (and what’s been going on), but they talk in restricted codes and we can’t always believe them even when we understand them. For every card on the table they have three on their chest and maybe another up their sleeve. Consequently the plot – or “histoire”, or story we deduce – moves faster than the narrative – the story we see. This is a rare, even paradoxical method of storytelling.

Miller's Crossing

In place of a moral centre there is a central enigma, the cryptic character of Tommy (Gabriel Byrne in one cinema’s great performances) who is of dubious heroic, or anti-heroic, stature.

Sometimes it helps to get our mind around a “difficult” movie by considering what it is not. Miller’s Crossing is not part of the gangster genre – unless you want to stretch the criteria of the genre out of shape. Of course there are superficial resemblances, all its characters are gangsters or their associates. And the concurrent release of GoodFellas and The Godfather: Part III, which are genre films, creates a climate of receptivity we can hardly ignore.

I regard The Godfather: Part III as a film without a flaw, but second-rate, whereas I see Miller’s Crossing as a first-rate film with flaws. By “first-rate” I’m thinking of those infrequent films which are neither structured nor constrained by the fashions of their time, though they may use those fashions. They do not confirm our thinking about cinema so much as they revise it; they attempt more, though they may achieve less.

A “classic” by definition exhausts some of the creative possibilities of its form; beyond the classic is the kind of film which carries out the promise of new possibilities, another form, in Jean Duvignaud’s term, “the wager”. Art that endures does so because it takes a bet on the future. There is no contradiction in reworking material from earlier periods or other art forms to do this.

GoodFellas and The Godfather: Part III beguile us with their apparent authenticity; we believe that the world they depict is “out there”, co-existent with our own history. Thus the theatre seat becomes a “safe place” from which to view whatever shocks and abominations may be their subject matter. They rely on conventions, principally from the gangster genre, for their harmony and power – and for our reassurance.

GoodFellas forces us to confront some of the social and historical reality which is suppressed or euphemised by the genre’s mythic machinery, the seductive production values of The Godfather: Part III gives spurious life to an elegantly galvanised corpse. Fine achievements both, but neither of them could exist without the genre they seek to qualify.

Miller’s Crossing is somewhere else, doing something else to us. Though it refers to the genre, it can exist without it. It even looks different from the other two. Part of the aesthetic pleasure of The Godfather: Part III come from a pictorial method that confirms us in our safe place next to “reality”; the rich colours of its pro-filmic world are made dark by the cinematographer’s art – the style itself is our protection. GoodFellas and The Godfather: Part III are lit dark. Miller’s Crossing is designed dark.

The “look” of the film has a different cumulative effect upon us, even in its “public” scenes which are designed in lighter hues. (One of the minor paradoxes of this stylistic discipline is that the ladies’ powder room at the Shenandoah Club is experienced as “public” while a gun battle on the street is “private”).

Miller’s Crossing has some serious flaws. Performances lack the consistency of stylisation which is one of the strengths of GoodFellas. Narrative pace is sporadically sluggish compared to the majestic slowness of The Godfather: Part III. Some of its moments of satire are shamelessly crude alongside comparable elements in the other two films. But for the Coens satire and/or parody are mere ornaments, their irony is embedded in the “what if” of their fictional world rather than in simulation of the “real world”.

The storyline may be the least immediately explicable element of the film but it commands our attention. While the pace within scenes tends to drag, the pace between scenes is rapid, ahead of our expectations. The situations change faster than the events, so the viewer’s intellection is more retrospective that anticipatory. Most screen narratives operate on tension between the what-happens-next and the what-happens-now. In Miller’s Crossing this is twisted so strongly and unusually we might call it “torque”. It could be a useful metaphor towards understanding the Coens’ story and how they tell it. Their combination of expressionist images and performances with impressionist storyline and dialogue builds up a curious, twisted tension – torque.

Miller’s Crossing is ostensibly about gangsters, with a few fringe politicians. Mythically it is about politics. We see an imaginary balance between “chisel” and “muscle”, the instruments of power in any system, and between brain and nerve, the human resources which administer them, and between circumstance and opportunity, the factors which divide loyalties and thereby destabilise the system. Everyone knows his/her place, but some aspire to a higher place.

Chisel is the Name of the Game

Chisel is the name of the game, and Leo is in the highest place. He has nerve but he needs brain, which is supplied by Tommy who “can see all the angles”. Tommy is an oddity, uninterested in gaining either power or its pay-offs, but addicted to gambling with money or with lives, including his own. Leo and Tommy are friends, not boss and henchman. Between friends is Verna; she uses nerve and brain to save her brother Bernie, a minor player who has “broken the rules” and angered Johnny Caspar, one of Leo’s satraps. Caspar’s state of mind is strictly muscle, but actual violence is left to his henchman the Dane, the only one with brain enough to match Tommy. However, the Dane takes orders, Tommy does not. Against Tommy’s advice, Leo “breaks the rules” for Verna’s sake to protect Bernie, and suffers a briefly successful revolt by Caspar.

Miller's Crossing

When the shooting starts, Leo is able to revert to what he does best – muscle (the film enters its most mannered and unrealistic passage when he wields his Thompson gun with an unholy joy which the responsibilities of office have long suppressed). In a characteristically subtle move Tommy manipulates Leo to break their friendship, gambling that he can then manipulate Caspar. When Bernie perishes, Verna dumps both her lovers. Although Tommy loses control of the game temporarily he is able to restore Leo’s power but their friendship is ended.

The plotting of Miller’s Crossing is as intricate and inevitable as the curse on the house of Atreus, its pattern of escalation from minor transgression to outright war as inexorable as the burning of Njal. If Tommy were a man of honour (you may believe he is), he would be in Lancelot’s dilemma towards Verna and Leo. However, the process we need to deduce this is radically different from that of the literary references. We have to “read” the game and the players similar means of inference to our “reading” of political events. Think of movies about politics: most have to be “read” as literature, though there are exceptions such as The Scarlet Empress (1934) and Salvatore Giuliano (1962).

Our consolation for such unfamiliar effort is to see the political aptness of its psychological revelations. Tommy refuses financial help from Leo because it would end their friendship; he refuses a similar offer from Caspar because it would begin their friendship. Bernie is ready to betray Tommy because hatred for being made to plea for his life is stronger than gratitude for being spared. This is emotional territory where movies seldom venture. The Coens expect us to infer these insights without benefit of exposition. And this is cognitive territory where moves seldom venture.

The film’s thematic territory, drawing upon the parallel logic of crime, business and government, is more familiar to us, as we look back from The Godfather (1972) to such forerunners from the fifties as The Big Heat (1953), New York Confidential (1955) and The Brothers Rico. But the first movie to examine this trinity of power is one that historians tend to forget: Rowland Brown’s cynical Quick Millions (1931). Its final image of the black hat moving as though of its own volition, accompanied by the hardboiled envoi “Us gangsters get the best funerals”, finds ironic resonance in Miller’s Crossing.

Apart from occasional references to literature and cinema, the dominant artistic strategy of the Coens’ film can be attributed to two models: one is literary and has similar theme, the other is filmic but may not be easy to recognise because it is thematically remote from the Coens’ work.

A Problematical Place

By strategy, I don’t mean style or technique – the Coens have their own. I’m thinking about a theoretical approach to any representational art which raises the workload of its audience, putting them in a problematical place between the representation and the represented. The opening dialogue scene of Miller’s Crossing registers its literary model: Dashiell Hammett’s The Glass Key (1931). The relationship of Leo and Tommy is immediately identifiable which that of the novel’s Paul and Ned, a peculiar theme with a distinctive treatment. As the film proceeds, there are some major divergences in storyline. The Coens do not concern themselves with the sociological dimensions which give the novel its title. However, this is the very point of Rowland Brown’s image of the wayward hat, it symbolises the breaking of the glass key. (Since Hammett’s novel and Brown’s movie came out in the same year it is unlikely they compared notes.)

It was Hammett who politicised storytelling by presenting externally observed personality, action and speech as matters of inference. What characters do and say, or are reported to do and say, are cyphers which have to be interpreted, decrypted. Some can be read at face value, others are signifiers of their own contradictions. Gaps, the absence of elements which orthodoxy leads us to expect, also have to be interpreted. Some key players are rarely seen but often referred to (as in Mink in Miller’s Crossing). The detection process is not what the work is about, but how it is about.

We have to be able to understand people and mores in the real world in order to make our deductions on the run about the author’s imaginary world. The comfortable, amusing clue-spotting of genre whodunits is far removed from the radical strategy of The Glass Key and Miller’s Crossing which engages us in a high level of intellection.

In addition the Coens’ film has a high level of sensuousness, especially in its mannered pictorialism. This is where we can connect with its filmic mode: the Paramount films of Josef von Sternberg (1930-35). The irrational pleasure of imagery, the ravishment of the eye, serves a rational purpose which is the opposite of indulgence. It distances us from naïve naturalism, filtering and refracting the meanings of gesture and intonation, of garb and décor, of things said which may be true or untrue, and of things unsaid.

It’s a paradoxical strategy: sensuousness operates as a distancing device to stimulate intellection. Like Hammett, the Coens locate their story in one (any one) of those cold dirty cities of New England, those ethnic melting pots of an earlier part of this century when notions of domain and rank, rule and obligation, were feudal in their simplicity. Hammett, working with words on the page, didn’t have the problem of the filmmaker who wants to escape the mimetic tyranny of the photographic image.

The Coens, making a movie not an illustrated novel, use some of the rhetoric of the gangster genre (hats, coats, doorways, argot and wisecracks) to avoid being pinned down to the specifics of history and geography. Von Sternberg insisted that his Morocco was a “country of the mind”. So too is the unnamed city which has Miller’s Crossing on its outskirts.

The Coens’ values are vastly different from von Sternberg’s. Apart from a similar, but less sophisticated, use of absurdity as a measure of “bad faith”, they have little to say about the existential struggle between Self and Role. This was the dynamic matrix of von Sternberg’s masterpieces a generation before it became the philosophical underpinning of the nouvelle vague. Its absence from the centre of the Coens’ work may be part of the explanation of why they have yet to make their female characters as problematical as their males.

Miller's Crossing

Miller’s Crossing sometimes topples into the traps of expressionism, especially in the use of sound, and some performances are miscalculated. Albert Finney shows strength but not much range as Leo. A more flexible performance would have given us a more complex character without altering the screenplay. John Polito’s Caspar is played with impressive energy, but the full pedal treatment allows no modulation. As a result the ironic potential of his naivety and his ruthlessness, his grotesque paternalism and his loss of nerve when chisel replaces muscle all come over as witless parody. I suppose the actors could have done better, but the responsibility for their shortcomings is the director’s.

Perhaps it was in a mistaken attempt to render filmically one of the least appreciated aspects of Hammett’s way of describing behaviour accompanying dialogue. Characters frown, wince, pale, flush, stiffen, tremble and otherwise over-react in the manner of 19th century fiction. This is in disconcerting contrast to Hammet’s description, or non-description, of almost everything else. I like to think this is deliberate – a form of mannerism rather than cliché, it privileges the reader with the discerning powers of a special kind of observer – such as a detective (which was Hammett’s career before writing). Whereas he remains a nice distance from caricature, the Coens’ filmic emulation lacks rigour and consistency, and often slips over the edge into thespian bombast (“hamming” to you).

The Enigmatic Centre

Yet they succeed triumphantly with the enigmatic centre of the film, the character of Tommy which Gabriel Byrne plays, and is seen to play, via the mise en scène. It is one of the most exquisitely disciplined performances in American cinema – historically a milestone, stylistically a touchstone. Like Hammett’s Ned he is a stoic with “a face cruelly placid”. If the Coens are to be blamed for Polito and Finney they must be praised for Byrne’s metallic calm and refined detail (the droop of the eye, the sardonic grimace, the Gaelic aspirates in his speech, etc.).

What Hammett created in Ned was a rare kind of fantasy for alienated intellectuals: playing the power game without taking responsibility, being on familiar terms with the mighty, being needed by them but not needing them in return, at liberty to come and go in the corridors and the smoke-filled rooms, attending the seat of power whilst disdaining its rewards. Tommy also embodies this fantasy. But, along with the lone wolf pride and the chess player’s ability to think three moves ahead, goes a weakness. He has a gambler’s addiction to risk; he needs the game not the prize. Even when he gets his hands bloody or his stomach betrays him, the lust to gamble steadies his nerve.

The Coens have found a filmic style for this difficult and fascinating concept, particularly in their “soliloquy shots” which combine the still figure with the infrequent dolly-in and the slightly overripe music. Their Tommy is not a copy of Ned but a provocative variation, as is the entire film of its literary source.

Miller's Crossing

The Coens’ ending is their own: in a final bitter gesture, Tommy pulls down his hat brim in the manner of Dane. He realises, as Leo cannot, that in taking up the gun to save Leo’s power, he has performed the office of henchman, and his quirky pride, or a grievously twisted sense of honour, sees it as the end of their friendship. If you don’t buy the alienated fantasy interpretation, you might say that once Verna is no longer between them, the friends can no longer be together.

Although this is not Hammett’s ending, he deserves the last word: “Whatever you’ve done you’ve paid for and been paid for and that goes for all us”.

About The Author

John Flaus began writing film criticism in 1954, and was sacked the following year when he wrote that On the Waterfront was right-wing propaganda. He has been writing film reviews intermittently ever since. These days he makes a living as an actor, script editor and occasional lecturer.