Subtitling - An Issue of Speed?

Executive summary

This research set out to evaluate deaf people's experiences of subtitling, looking specifically at how comprehension and enjoyment of different types of programmes are affected by the speed of the words on screen. The sample of respondents consisted of a range of moderately, severely and profoundly deaf subtitle users and included people from across a wide range of literacy levels.

There are a number of interacting variables which make the issue of speed extremely complex. Degree of deafness, age, reliance on subtitles, and literacy level all play a part. For people who utilise more than one sensory input, subtitles are often used as a useful reference point if they miss a word, so speed is not such a critical factor. For those who rely more heavily on subtitles, speed is much more of an issue because if they cannot read them fast enough they will literally lose the plot.

Almost all subtitle users are extremely appreciative and grateful for subtitles. They believe UK subtitles in general are of a high quality and expect them to remain at the same standard or improve in the future.

The majority expect subtitles to remain at the same standard or improve still further in the future.

There is considerable sensitivity and antagonism towards the idea of editing, especially amongst the profoundly deaf who want access to the same information as hearing viewers. But the benefits of editing, especially to the wider deaf audience, are appreciated when considered more fully.

When prompted to comment about a potential increase in subtitling speed, the majority of respondents reacted negatively believing that there was no advantage in increasing subtitle speed. It was considered that faster subtitles would reduce many deaf people's enjoyment of television and alienate those with lower literacy skills.

Spontaneous issues raised are more to do with presentational style. Specific concerns include the quality of the editing, spelling, positioning, speaker identification and the need for the subtitles to faithfully mirror the action.

Where subtitle users do have an issue with speed, it is more frequently to do with subtitles being too fast rather than too slow.

But, under observation, respondents did not easily distinguish between the different subtitling speeds of the programme clips shown to them, especially between 160 and 180 words per minute. Above 180wpm respondents were more likely to find the subtitles too fast, and more difficult to follow.

There is an important distinction between being able to follow fast subtitles and being comfortable with fast subtitles. When viewing the programme clips, nearly 90% of all groups, regardless of age, degree of hearing loss or literacy level, were able to read all or nearly all of the subtitles. However, almost 40% thought that current subtitling as depicted by these examples, were too fast.

If forced to choose between whether the speed of each clip was too fast or too slow, viewers across all groups were significantly more likely to believe that the speed of the clip was "a bit too fast".

The most important defining criterion of subtitle users is first language: signing versus oral. In general, British Sign Language users find subtitles faster and more difficult than deaf people with oral English. Profoundly deaf users tend to have the most problems with fast or complex subtitles.

Among the over 45s, especially those who had been using subtitles for 5 years or longer, almost all felt their ability to speed read quickly had improved. Those who rarely watched television or who had only recently become deaf were the most likely to find subtitles problematic.

Those aged 54 or younger, seem to cope better with faster speeds, as do those who use computers regularly and those of a higher literacy level.

The higher the quality of the subtitles, e.g. speaker identification, consistency, the easier it is to follow fast subtitles.

Any format or programme that is familiar to the viewer causes fewer problems with speed. People who watch soaps, for example, know the characters well and anticipate the dialogue and action easily. But familiarity with genre format will vary from viewer to viewer, so the possibility of having faster speeds for selected programme types was deemed unworkable.

Most users feel that 'words per minute' is not the most appropriate means of evaluating subtitling speed, although they acknowledge that there must be guidelines for organisations to work within. They point out that programmes are so individual in terms of bursts of dialogue followed by periods of silence, or complex multi-speaker scenes followed by simple monologues that such guidelines do not account for the differences in viewing experiences. Any increase in subtitle speed must take account of such factors so that bursts of dialogue are made manageable, and other issues such as spelling, matching the action and consistency are also addressed.

The majority of deaf viewers would like subtitle speed to stay the same, but there is little evidence that they would be able to accurately identify the difference between 160wpm and 180wpm. There remains the issue, however, that while many people may be able to read faster subtitles, they do not necessarily want very fast subtitles on a day-to-day basis when they are watching television for leisure reasons.

It is recommended that if an increase in speed is proposed that it is done so in conjunction with broadcasters’ reaffirming their commitment to the quality and style of presentation, which are the subtitle issues that are of most concern to deaf viewers.

Whilst any increase in speed potentially will alienate a proportion of deaf viewers, this research suggests that subtitling speed should not normally exceed a threshold of 180 words per minute, or three lines of text on screen. To do so could make following subtitles more difficult for a significantly increased section of the deaf and hard of hearing audience.

Please see the link below for the full version