Why speech has to be shortened
People often speak faster than they can comfortably read. Spoken dialogue also contains repetition, hesitation and filler that carry little information. If every word were put on screen, the text would either race past or crowd the picture. Writers therefore condense: they rewrite what was said into a shorter form that fits the time the subtitle will be visible.
This is not the same as summarizing the whole film. The aim is that a viewer who reads the subtitle understands what a viewer who heard the line understood, in about the same moment. Good condensation is usually invisible; the viewer does not notice that words were dropped.
Not every project condenses to the same degree. Some contexts, such as certain educational or accessibility services, aim to stay closer to the spoken words, and others accept more trimming. The client’s guidelines decide how far to go.
What is usually kept
Writers generally protect the parts of a line that move the story or the information forward.
- The key information. Who did what, to whom, and when, if it matters.
- Names, numbers and technical terms the viewer needs, since these cannot be guessed.
- Tone. A threat, a joke, an apology or a hesitation should still feel like one.
- Words that other lines depend on. If a later line refers back to something, the earlier line must keep it.
- Distinctive voice, where it is important. A character’s habitual phrase may be worth a few characters; a generic filler is not.
What is usually cut
Some material can be removed with very little loss.
- Fillers and hesitations, such as “well”, “you know” or “I mean”, when they add nothing.
- Repetition, where a speaker says the same thing twice for emphasis and the picture or tone already conveys it.
- Greetings and tag questions that are obvious from context.
- Information already visible on screen, such as a name on a sign that the viewer can read for themselves.
- Redundant modifiers, like a stack of adjectives where one carries the sense.
Be careful with cutting anything that sets up a joke, a clue or a later callback. The line may look disposable but the story may need it.
Common techniques
Condensing is mostly rewriting, not just deleting. A few typical moves:
- Choose a shorter word with the same sense.
- Switch to a simpler structure, such as active voice, or a direct statement in place of a long polite request.
- Merge two short sentences into one, or split a long one into two cues.
- Drop a subject or verb that is clear from context, as spoken language often allows.
- Replace a phrase with a pronoun when the reference is obvious.
A hypothetical example: the spoken line “Well, I mean, I think that maybe we should probably go now” could become “Maybe we should go.” The hesitation is gone, but so is a little of the speaker’s nervousness. If that nervousness matters in the scene, a slightly longer version such as “I think we should… go now” might be worth the extra space.
A hypothetical timing check
Condensing is driven by arithmetic as well as taste. Suppose a guide allows roughly 17 characters per second and a line will be on screen for 3 seconds. That allows about 51 characters (17 times 3). If the full spoken line would need 85 characters, the writer must find about 34 characters of savings, or find a way to show it for longer. The numbers here are only an illustration, since limits vary by client and language, but the sums are the same everywhere. How long a cue can stay up is covered under display duration, and what the limits mean for viewers is covered in the article on reading speed.
Condensing is a judgment about the audience
Who is watching changes what is safe to cut. Children subtitles usually call for simple wording, and older viewers may benefit from text that stays up longer rather than text that is cut harder. Viewers who are deaf or hard of hearing cannot compare the text with the sound, so they may be relying on the subtitles for information that hearing viewers get from tone. A heavier trim that works for a hearing audience watching a foreign-language film may be less suitable there.
Subtitles that translate from another language add a layer: translation and condensation happen together, and the translator must also manage text expansion when the target language runs longer than the source. In those cases the shortening is often as much a translation decision as a style rule.
Avoiding common mistakes
- Cutting the point. Removing the one word that carries the answer, such as a negation, reverses the meaning.
- Over-flattening the voice. If every character ends up sounding alike, too much personality was trimmed.
- Creating mismatches with the picture. If a speaker clearly says a long sentence and the text is two words, viewers who know some of the language notice.
- Condensing by habit. If there is time and room, the full line is better than a shortened one.
- Introducing errors. Rewriting for length can accidentally change a fact. These are the kinds of subtitle errors that proofreading is meant to catch.
Checking the result
Read the condensed subtitle aloud, then ask three questions. Does it sound like something a person would say? Does it keep what the viewer needs? Can it be read comfortably in the time it is on screen? If the answer to any is no, rewrite it. A second person reading the subtitles without the script is a good test: if they can follow the scene, the condensing worked. More on reviewing text this way appears in the guide to subtitle proofreading.