The phrase “deleted scene” can refer to different situations. A passage that was written but never filmed does not provide the same evidence as a sequence that was shot or a fragment removed after a screening.
A written alternative is not an invisible film
A screenplay reveals a proposed set of actions and dialogue when its version is identified. It does not prove that everything written was filmed. Between the page and the shoot, the scene’s duration, cast, location, and organization may change. Discussing discarded material requires first specifying what kind of material exists, rather than imagining a complete film behind every page that differs from the released version.
The script versions help distinguish these possibilities. A scene present in a draft but absent from a revision may document a change in the writing. Another source of evidence is needed to claim that it was filmed. Nor is finding a scene number in a shooting schedule enough: a schedule records an intended course of work, not necessarily what was carried out.
The search for locations raises a similar limitation. A photograph taken to evaluate a space may relate to a scene that was never made. The object exists and has documentary value, but it should not be turned into a frame from an alternate version. Preserving its original purpose makes it possible to explain what was considered without inventing a later stage.
This distinction does not diminish the interest of preparatory materials. A written alternative may show a different distribution of information or a different relationship between characters. These differences can be analyzed as narrative possibilities. What they cannot do is describe the acting, editing rhythm, or sound mix of a sequence known only through words.
Filmed material, prepared music, and work-in-progress versions
The Dr. Strangelove pie fight is an example of filmed material that was removed. The archive catalog describes a late version of the screenplay that no longer includes the sequence. That record helps establish a difference between materials, but it does not by itself reproduce the scene’s edit or determine all the reasons for its removal.
A still photograph from a shoot does not preserve duration, either. It may show performers, objects, and a situation, but not the order of shots, the pauses, or the relationship to the sound. Reconstructing a sequence requires materials that can establish those connections. Arranging known images in a plausible order produces a researcher’s proposal; it does not automatically recover a historical version.
2001 offers a documented difference between continuity documents. The British copies from March 1968 include transcripts of scientists for a planned prologue; the American copies from July do not. The comparison identifies a documentary variation. It is not enough to claim that each corresponds exactly to a copy shown to the public or to reconstruct how the prologue would have fit into the film.
The unused score by Alex North for 2001 is another kind of discarded material: music prepared for a film whose final soundtrack took a different direction. Comparing a recording with images can be useful, but it requires knowing the version, excerpt, and synchronization. Adding music to a scene on one’s own initiative does not make the result an authorized reconstruction of the edit that would have been released.
The sound stage matters, too. A filmed scene may retain guide vocals or lack a finished mix. Its provisional appearance should not be judged as though it necessarily represents the definitive result the team would have chosen. At the same time, recognizing that it is incomplete does not justify attributing hypothetical improvements to it to argue that it would have been better than the released version.
Removal from an edit does not mean disappearance from history
The versions of The Shining remind us that comparisons can be made between edits with different public release histories. In that case, the question changes: it is no longer only about what was planned or filmed, but about identifying which version was shown, where, and when. A runtime on its own does not resolve the comparison when presentation or playback-speed differences are involved.
The vocabulary should preserve these stages. Written, prepared, filmed, edited, and released are not synonyms. A scene may pass through some stages but not others. To support the claim that something was removed after a screening, it is necessary to document both the screening and the change, rather than infer them solely from the material’s apparently advanced state of completion.
The archive may bring together traces from several stages without providing a complete sequence. A letter may explain a decision; a screenplay may reflect its result; a photograph may preserve an earlier moment. None should be treated as universal proof of the entire process. Its value lies in what it actually records and in the relationships that can be established and verified.
Artistic evaluation is a separate question. A deleted scene may change the pacing, information, or tone, but its existence does not prove that the final edit is mutilated or that removing it was necessarily the right decision. A comparison needs to describe what changes and acknowledge which parts of the alternative material have not survived or are not known with sufficient precision.
Working with discarded material requires resisting the promise of a secret film that can always be recovered. Materials can shed light on decisions and possibilities without forming a complete version. Identifying their status, attributing reasons when testimony exists, and distinguishing reconstruction from documentation make it possible to study genuine alternatives. This also preserves the value of what is incomplete: partial evidence can answer a specific question without resolving the entire history of a film.
Sources: Stanley Kubrick Archive, UAL, SK/11/1/26, Late script for Dr. Strangelove; SK/12/4/2/1, Dialogue Continuity Scripts – Printed in England, March 1968; SK/12/4/2/2, Continuity Scripts – Printed in the USA, July 10, 1968; SK/12/8/1/42, correspondence about music for 2001; theatrical versions of The Shining.
View all articles in Kubrick’s Working Method, Documents, and Archive