This dialecticized or oscillating structure is absent from Dark Passage, and not only from the scene where Parry escapes from Madge's apartment. As Parry jumps to the ground and rushes out into the street to board a trolley car, the camera continues to keep him fully in view. The absence of an answering look is now marked not only by the continued refusal of a reverse shot, but additionally by the disturbing depopulation of the field we are given to see. The street and the trolley car are inexplicably empty. Our disquiet continues throughout the bus sequence that follows. Narratively this sequence is set up for suspense: the bus will not leave until two more tickets are sold. We are therefore led to expect a tense countdown as the selling of the tickets is measured against the arrival of the police. But this is not what we get. The sequence is cut to disperse rather than to build suspense. A policeman does arrive, nonclimactically, at the station and, as we watch the four other passengers board the bus one by one, we wonder if Parry will make it on before he is spotted. As the last of the other passengers walks toward her seat, the camera pans slightly forward to reveal Parry, already seated. Not only does his sudden appearance here not answer to the diegetic look of the police, it also fails to respond to the extradiegetic look of the spectator; that is, Parry's appearance neither meets nor surprises our expectations, it simply disregards them. To whom, then, is Parry visible?

It can be argued that he is visible to no one; not articulated within a dialecticized structure, his visibility is for no other subject. If he is seen, nevertheless, it is by something more 'rigid' than a subject, something incapable of being deceived, the way a subject is deceived. But if the dialectic of desire - hence suspense - and the possibility of deceit are both eliminated from the world of Dark Passage, then we have not only left the world of detective fiction far behind, we have also left behind many of the commonplaces of film noir criticism. As an example of film noir, the film would be described by this criticism as a dark sort of 'thriller', filled with deceptions, false leads and sudden reversals. Would we deny these basic descriptions of the genre? Yes. What is taken for deception is, in fact, something much more sinister: the emergence of a split between power and those whom power subjects such that the very world of these subjects appears incomprehensible to them. The world no longer unfolds in nonsimultaneous parts, as in detective fiction; in film noir it breaks up into inconsistent and always alien fragments. As a consequence of this, nothing can remain hidden in the noir universe only to become visible in a future moment. The noir hero is embarrassed by a visibility that he carries around like an excess body for which he can find no proper place. Already encumbered by his own overexposed being, the noir hero has no desire to seek his being through another. The links of intersubjectivity, invisible yet essential to detective fiction, are here dissolved. A different kind of struggle with a different, unyielding adversary remains.

Film theory still operates largely according to the structuralist assumption Lefort indicts. We conceive a film as a system of relations among 'minimal significant units', at various levels of articulation. Anyone familiar with the study of film noir, for example, can recite a list of such units or elements thought to be necessary to the definition of the genre: a femme fatale, a morally compromised detective, an urban setting, voice-over narration, convoluted plot structure, chiaroscuro lighting, skewed framing, and so on. Lefort's implied argument that any such list, along with a specification of the relations among its individual terms, is bound to remain a mere heap of traits with no real coherency is borne out by the continued suspicion regarding the very status of film noir. We pay too much attention to the established terms and their relations without ever inquiring into the principle by which they are established. In virtue of what, we neglect to ask, is this particular organization instituted?

Sometimes, however, we are fooled into believing that we have asked this question, and are able to answer it, by making reference to a film's historic context. Film noir criticism correlates filmic elements with historical 'sources' - World War Two, an increase in crime, mounting paranoia regarding the working woman's place in society, and so on - thinking that it has thereby located the 'generative principle' of the films. But this reference to external sources in no way resolves the question of the internal logic of the films; the question is merely deferred. Raymond Borde and Etienne Chaumeton, authors of the first (1955) book-length study of film noir, acknoweldge [sic] this point even as they examine what they take to be the sources of the films. After discussing the rising incidence of crime during the period of the production of films noirs, they comment:

The question remains: why did crime suddenly assume such importance in Hollywood productions during these last ten years? For the existence of a phenomenon does not itself explain its artistic exploitation: artists and the public could misrecognize it, intentionally reject it, or consider it unaesthetic. Why were American spectators so sensible to films of violence and murder?