Improvisation and prepared text in Kubrick

A line can emerge during a rehearsal and arrive on set as learned text. Discussing improvisation requires identifying when the contribution was made and what we know about its incorporation into the film.

A spontaneous contribution can end up in writing

In his 2013 interview for DGA Quarterly, James B. Harris explained that, during his collaboration with Kubrick, they recorded suggestions that arose in rehearsals and later integrated them into the script. The process he recalls separates discovery from execution: material could originate in a spontaneous contribution, but the actors would then receive prepared pages to work with.

That distinction avoids an overly simple opposition between freedom and control. A revised text can preserve an idea from a performer; an apparently free take can be based on dialogue that has already been set. The contribution does not disappear because it has been written down. It changes form within the collective work and becomes available to other actors, so they know when to respond, step in or remain silent.

The difference affects how we read a performance. A verbal hesitation may seem accidental while being carefully calculated; a perfectly articulated retort may come from a suggestion improvised days earlier. The result can be analysed through its rhythm, the relationship between characters and the information conveyed. Reconstructing its origins requires testimony or preparatory materials that identify that specific stage.

It also matters what counts as a contribution. It might be new words, a physical action, a change in register or a way of connecting two existing lines. collaboration on the script is not just a matter of adding memorable lines. A suggestion can change the length of a scene or its power dynamics without introducing a line that the audience remembers on its own.

Sellers and Ermey: contributions in different contexts

The appearances of Clare Quilty in Lolita offer a way to observe shifts in social persona within a single performance. Sellers changes the way he speaks and approaches Humbert depending on the situation. The repetitions, circumlocutions and apparent difficulty getting to the point prolong the discomfort. That effect can be studied without deciding, on the basis of Sellers’ ease, which word he made up in front of the camera.

In the hotel encounter, the conversation seems constantly to veer away from any fixed purpose. Humbert tries to maintain a version of his situation while his interlocutor knows more than he lets on. The comedy depends on the imbalance between them, not simply on one of them talking a lot. A verbal variation only works within that relationship: it can come close to what Humbert fears without stating it entirely directly.

The study of Sellers as Kubrick’s collaborator explores his characters and specific testimonies. Here, the important thing is to preserve the distinction between interpretive flexibility and where the text came from. The same actor can produce very different results with prepared material. Attributing the whole scene to a flash of inspiration would erase both the scene partner’s response and the construction of the dramatic situation.

In Tim Cahill’s 1987 interview published in Rolling Stone, Kubrick attributed a significant part of Hartman’s language to R. Lee Ermey, which emerged during tests with prospective cast members. He also described how he prepared the dialogue. The testimony allows us to distinguish the generation of material from its later performance, without turning every contribution from the instructor into a surprise that happened during the take used in the film.

Hartman turns his attacks into an exercise in classification. He changes names, demands answers and makes an individual characteristic a public matter. The recruits have less freedom to interrupt or rephrase. The verbal content and control of the turn reinforce each other. Ermey’s contribution to the material does not change that inequality within the fiction: the character exercises an authority that the scene makes audible.

The film preserves a selection, not everything that happened

A quip that works during a rehearsal still has to find a place in a sequence. It may need an earlier set-up, a legible reaction or a length that does not unnecessarily slow the story. That integration distinguishes a useful suggestion from an entertaining anecdote about the shoot. The film presents the material that was selected and organised, not the full range of possibilities that may have been tried.

The other performers’ reactions are part of the construction. When a character seems bewildered, we cannot conclude that the performer did not know what they were about to hear. They may be portraying that bewilderment precisely. In the same way, a quick response does not prove that everything was set from the first draft of the script. The character’s surprise and the actor’s knowledge belong to different levels.

The rhythm of the performance helps us study what remains accessible. We can identify an interruption, an extended pause or a change of pace. We can also observe whether a retort wraps up an exchange or opens it up again. These are perceptible choices whose effectiveness does not depend on assigning them a spontaneous origin. Improvisation is one possible way of producing material, not an automatic guarantee of quality.

When versions of the script exist, comparisons must take their dates into account. The fact that a line does not appear in an early version does not prove that it originated during filming: it may have been added in an intermediate revision. And its appearance in a transcript published after the release does not prove that it was written before filming either. The document must be placed within the process before it can support a conclusion about authorship or when the line was incorporated.

Testimony about the Full Metal Jacket cast’s work helps to clarify some of these stages. In other cases, the available information only takes us as far as the finished result. Respecting that limit allows us to recognise performers’ initiative without replacing collaboration with a story of instant genius. The useful question is what a contribution changed and how it became part of a shared scene.

Sources: F. X. Feeney, interview with James B. Harris, “In the Trenches with Stanley Kubrick”, DGA Quarterly, Spring 2013; Tim Cahill, “The Rolling Stone Interview: Stanley Kubrick”, 27 August 1987. Films analysed: Lolita (1962) and Full Metal Jacket (1987).

View all articles in Directing actors and staging in Kubrick