2001 was initially shown in a longer version and was cut shortly after its release. To compare runtimes, it is necessary to distinguish narrative footage from elements of the theatrical presentation.
Release and changes in the first few days
Press sources compiled by the AFI Catalog document that Kubrick made cuts after the first screenings in April 1968. Various contemporary records mention that approximately nineteen minutes were removed, although published runtimes do not always refer to the same set of elements. This information needs to be presented in context: a standalone figure may include or exclude the overture, intermission, or exit music. The organization of the screening explains why these components are not equivalent to fictional scenes.
The changes occurred when the film was already being seen by critics and audiences. This makes it possible to compare reactions to different presentations, but does not justify claiming that every attendee saw one single initial version. The distribution of prints and theater schedules introduced differences. A historical reconstruction should identify the date, location, and runtime when citing a specific screening. The structure of the movements belongs to narrative analysis and should not be confused with the material history of these changes.
What can be said about the deleted scenes
Period press accounts and audience testimony from the first screenings described differences in passages involving the Discovery and the conflict with HAL. These accounts are valuable, but they need attribution. A viewer may remember a scene only in part or mix their experience with later information. Editing documents and surviving prints provide a different kind of evidence. Finding material in a screenplay is not enough to claim that a sequence was filmed; claiming it was screened requires evidence tied to an exhibited version.
The history of the cuts must also distinguish material prepared for release from proposals discarded before the final version. The scientific interviews planned as a prologue belong to a different process. A scene removed after a screening and an idea that was never filmed cannot be added together as if they formed one and the same lost version. This caution prevents the difference in runtimes from turning into an imaginary catalog of scenes whose existence has not been verified.
Comparing versions requires distinguishing at least four possible stages for a scene: written, prepared for filming, filmed, and exhibited. A screenplay may preserve material that was never filmed. A filmed scene may be absent from every known cut. A passage shown at an initial screening may disappear from later prints. The evidence needed differs in each case. Press accounts from April 1968 document cuts made after release, but do not automatically turn every draft of the project into a lost scene. The film's material history depends on preserving these distinctions.
Contemporary news reports mentioned changes to scenes involving the Discovery and the conflict with HAL. These references can be compared with audience recollections and production documents. However, a reconstruction should not treat a single account as an exact description of all the footage that was removed. Memories of a screening may be partial, and surviving prints do not always correspond to the same date. Identifying the location and format helps place each claim in context. Differences between versions are a documentation problem, not an invitation to invent scenes to make up a runtime.
The overture, intermission, and exit music add another complication to calculating runtimes. Two listings may give different times because one includes elements of the screening that the other leaves out. A responsible comparison should separate these components before calculating how many minutes were removed. We can say that 2001 was cut, but determining the extent and content of those cuts requires caution. The goal is not to reconstruct an imaginary version from runtime figures, but to identify which changes are supported by concrete evidence.
Runtime, preservation, and responsible comparison
A home-video copy, a 70mm screening, and a catalog listing may give different runtimes for reasons that do not involve narrative changes. The intermission may or may not be counted; presentation music may be added to the total; even cataloging practices can vary. The photochemical print presented by Nolan raises additional questions about which version was intended to be recreated and how that goal was documented. Comparison requires identifying the format and the elements included, not simply subtracting two numbers and automatically attributing the result to lost scenes.
The history of 2001 includes documented changes without requiring a legend about a hidden definitive cut. The released film and its modifications are concrete historical objects. Researchers can acknowledge the limits of available information and leave open questions about materials they have not consulted. This precision makes it possible to study how a long film was adapted for commercial circulation without confusing recollections, screenplays, prints, and published runtimes.
Information about the first screenings should preserve the provenance of each figure. A runtime taken from a review and one taken from an exhibition listing may measure different elements. Without identifying the screening, format, and included components, subtracting one from the other cannot establish with confidence which images disappeared.
View all articles in 2001: A Space Odyssey: Characters, Images, and the Construction of a Journey