Versions and restorations of Fear and Desire: what is being compared

Fear and Desire circulated in different cuts and copies of varying lengths, including an early version close to seventy minutes and a shortened version of about sixty-two minutes. Comparing them requires identifying the materials, playback speed and editing decisions—not just catalogue numbers.

What is known about the running times

Documentation relating to the 1952 Venice presentation puts an early version at around seventy minutes, when the film was still identified as Shape of Fear. Other records and accounts describe a later, shortened version of about sixty-two minutes. Modern restorations have made it possible to screen longer materials, sometimes catalogued at running times close to seventy or seventy-two minutes. These figures come from different contexts and should not be set out in a table as if they were identical measurements of the same print.

The running time can vary because of editing, but also because of projection speed, titles or cataloguing criteria. A difference of a few minutes does not, by itself, identify which shots are missing. To claim that a scene was removed, it is necessary to know both copies, their sequence and their visible elements. Similarly, finding a longer version does not automatically mean it is the final form Kubrick would have preferred at any later point.

The shortened version is associated with a circulation history in which the director altered his own debut. Without precise documentation, it is not appropriate to explain every cut as the result of censorship, commercial pressure or a single philosophical theory. The history of the versions is a material and chronological problem before it is a psychological explanation. Kubrick’s self-criticism is another question, linked to statements and decisions about the film’s circulation.

Preserved materials and restoration work

A restoration may use photochemical elements from different sources: release prints, positives, negatives or other materials identified by archives. To describe a particular process, consult the accounts provided by those who carried it out. The fact that a film can be screened again in good quality does not automatically reveal which element was scanned or how many frames were recovered. Precision starts with naming the material that is actually documented.

Eye Filmmuseum has scheduled a revival of the longer version and reports the existence of historical materials related to the initial presentation. This information is useful for distinguishing the long period when certain prints were difficult to see from the complete disappearance of all of them. A return to cinemas or festivals can change availability for researchers, but each edition must be identified by its own source and configuration.

The question of deleted scenes requires an even more precise classification. A scene present in a screenplay but absent from all known copies may never have been filmed, may have been discarded during editing, or may have been lost from a later material; the text alone does not allow us to choose between these explanations. By contrast, if a sequence appears in an identified historical print and is missing from another, there is a verifiable material difference. This distinction prevents any passage described only in a written document from being presented as “recovered”. It also clarifies why a restoration should not promise to reconstruct artistic decisions that the surviving elements do not allow us to know.

The aspect ratio requires the same care. An image with different edges in two materials does not, by itself, prove that there are two different narrative cuts. It may reflect transfer, presentation or format decisions. Before discussing a lost framing, it is necessary to know the standards and conditions of the print being compared. Nor should supposedly deleted scenes be reconstructed from a screenplay draft: written, filmed and preserved material are distinct categories.

How to compare two copies without inventing differences

A useful comparison starts by recording the running time, source, titles, sequence order, and the presence of dialogue or images. It can then identify specific changes in their narrative context. If a passage appears in one copy but not another, it is possible to discuss how it affects the pacing or understanding of a character. If only the total running time changes, there is not yet enough evidence to attribute the effect to a specific deletion.

Comments on the philosophical scope of Fear and Desire should be made in relation to the version viewed. An assessment of the prologue, the construction of the doubles or the ending may change when editing alters the order or duration of a scene. It is therefore imprecise to speak of “the film” as if all historical copies were identical and known in detail. Analysis is stronger when it identifies the object of study.

A valid comparison also requires recording whether the copies have identical openings and endings, and whether they contain the same intertitles or credits, where relevant. Finding a scene that runs a few seconds shorter in one viewing is not enough: a change in speed or transfer may be involved. Only the identification and comparison of specific materials can support the claim that an editing intervention took place, after which its effect on the story can be assessed.

The versions of Fear and Desire tell a story of production, circulation and recovery that cannot be resolved by isolated figures. Running times of approximately seventy and sixty-two minutes are starting points for investigating specific cuts; restorations provide access to materials that must be described by their provenance. Rigorous comparison distinguishes verifiable facts from inferences and avoids equating a longer print with the director’s final artistic wishes.

View all articles in Fear and Desire (Fear and Desire): a film guide