Ing. Mgr. Jiří Rambousek

Katedra anglistiky a amerikanistiky

Filozofická fakulta Masarykovy univerzity

 

TV news subtitles: yet another instance of “little texts”

Titulky v TV zprávách: ještě jeden případ „krátkých textů

 

Key words: informative text, little texts, subtitles, TV news, language of news.

 

Annotation: The article analyzes a specific type of TV subtitle used on CNN International. It is the subtitle given in a permanent area on the screen and summarizing the reported news. The analysis is based on a corpus of 263 subtitles recorded during five hours of broadcasting on March 28–30, during the Iraq war.

Before presenting the actual analysis, the article briefly outlines the role of printed text in various types of media. It points out that the three most recent types of media (teletext, dedicated TV news channels, the Internet) brought a higher proportion of printed text in reporting news.

This tendency is supported by the analyzed subtitles. They differ substantially from traditional TV news subtitles: an overwhelming majority (74.5 %) could be used as standalone texts conveying the essence of the news. Traditional news subtitles, on the other hand, were used to add name to a person or a place (and could therefore be seen as predicates to the subjects shown on the screen).

The analysis reveals that the subtitles share some features with other types of “little texts” (Hallliday’s term); this applies especially to the high proportion of non-finite clauses – 43.4 % (40.5 % in newspaper headlines); within this specific category, TV subtitles show a slightly higher occurrence of elipted clauses (in which the finite verb is missing; 82.4 %) and a lower occurrence of nominal clauses (17.6 %) compared to newspaper headlines (ratio 61.2 % : 38.8 %).

Some other features are specific to TV subtitles. The most prominent of them – at the grammatical level – is the innovative use of the comma to replace the conjunction and (even in cases when it is not motivated by the need to save space). At the semantic level, the subtitles are characterized by the frequent presence of the source of the news (brought about probably by the specifics of real-time reporting in which it is not always possible to verify the news from an independent source). In style, the examined type of TV subtitles differs from newspaper headlines by the absence of language play and stylistic embellishments. This difference, too, is ascribed to the real-time production, as well as to the fact that they do not serve the purpose of attracting the viewer.

 

 

Anotace: Příspěvek rozebírá vybraný druh titulku užívaný v televizní stanici CNN International. Jde o titulek zobrazovaný v permanentní textové oblasti, který shrnuje právě hlášenou zprávu. Článek je založen na korpusu 263 titulků zaznamenaných během pěti hodin vysílacího času ve dnech 28. – 30. 3. 2003 (při zpravodajství z války v Iráku).

Ještě před vlastním rozborem titulků je krátce shrnuta role psaného textu v různých typech zpravodajských médií. Autor upozorňuje, že právě ve třech nejnovější typech médií (teletext, zpravodajské TV kanály, internet) došlo ke zvýšení podílu psaného textu na sdělování zpráv, což je v rozporu s obecně rozšířenou představou o ústupu psaného textu.

To potvrzuje i zkoumaný typ titulků, jenž se zjevně liší od tradičních titulků ve zpravodajství: výrazná většina (74,5 %) jich je schopna fungovat jako samostatný text, kdežto tradiční titulky sloužily jen k informaci o jménu zobrazované osoby nebo místa (tedy vlastně jako predikát k subjektu zobrazenému na obrazovce).

Vlastní rozbor pak ukazuje, že zkoumaný typ TV titulků má některé rysy společné s jinými typy anglických krátkých textů – především vysoký podíl vět neobsahujících sloveso v určitém tvaru; těch je 43,4 % (v novinových titulcích 40,5 %); v rámci tohoto větného typu vykazují TV titulky poněkud větší podíl vět eliptických (kde je určité sloveso vynecháno; 82,4 %) a menší podíl vět jmenných (17,6 %) oproti situaci u novinových titulků (kde je poměr 61,2 % ku 38,8 %).

Jiné rysy titulků jsou naopak specifické; po gramatické stránce je to neobvyklé použití čárky jako náhrady za spojku and, a to i v případech, kdy není vynuceno potřebou úspory místa. Po stránce významové je to pak vysoký podíl údajů obsahujících informaci o zdroji dané zprávy (jde zřejmě o specifický rys zpravodajství konaného takřka v reálném čase, daný potřebou distancovat se od zpráv, které před zařazením do vysílání nebylo možno ověřit). Stylisticky se TV titulky liší od novinových naprostou absencí jazykové invence a stylistických figur, které jsou pro anglické novinové titulky typické. I tento rozdíl lze připsat rychlosti produkce titulků, ale také tomu, že jejich funkcí není přilákat pozornost diváka.

 

 

TV news subtitles: yet another instance of “little texts”

 

In this paper I will describe a specific class of TV subtitle—a form of “little text”[1]—and will examine the similarities and differences between such subtitles and newspaper headlines. Before I proceed to the analysis, I will briefly summarize the role that text—as opposed to other channels for relaying information—has played in the history of news announcements.

Table 1 lists different media “in order of appearance”, i.e. according to when they appeared and were used in reporting news, and assigns to them the channels they employ to convey information.

 

 

            MEDIUM

CHANNEL

Printed text

Still
picture

Moving
picture

Spoken text
(sound)

1

Newspapers and magazines

MAIN

x

 

 

2

Silent movie

XX

 

MAIN

 

3

Radio

 

 

 

MAIN

4

Sound movie (news-reels)

x

 

MAIN

XX

5

TV news

x

 

MAIN

XXX

6

Teletext

MAIN

 

 

 

7

TV news on news channels

XX

 

MAIN

XXX

8

Internet

MAIN

XX

X

x

 

Table 1 Media types

(The main channel characteristic of a particular media type is marked MAIN; the other channels involved are marked 'x' to 'XXX', expressing the importance the respective channel has in supporting the main channel.[2])

 

I would like to argue that the situation has changed radically in favour of printed text over the last few decades, and that this change is not always fully appreciated. Most analyses that were carried out in the 1960s and 70s did not foresee it—on the contrary, they spoke of the end of written culture, of new technologies “bypassing verbalizing” (McLuhan[3] 1995: 274), of picture-oriented culture, of images replacing the linear sequence of symbols. Lines 1-5 of Table 1 show clearly why this was, and many of these observations are still valid. However, the remaining three lines of Table 1 dealing with the three most recent media formats (i.e., teletext, dedicated TV news channels, and the Internet) suggest a different tendency: the comeback of written/printed text. Admittedly, news reporting is only a small segment of public media activity, and one especially suited to being relayed by printed text. On the other hand, the tendency seems to extend into other areas as well.[4] Furthermore, all instances of “printed text” in Table 1 fall into the category of “little texts”, with only two exceptions: newspapers, and the Internet. So only the oldest and the latest of all media forms offer the freedom to write a text as long as is necessary, and the Internet goes even further by removing the limitations present in the newspapers.

The increasing importance of text as a component of TV news is illustrated by the way it is used on dedicated TV news channels.[5] At the dawn of TV news reporting, only very simple subtitles were used, mainly to indicate the name of a speaker or the location of an event. Later, short captions were added that characterized each piece of news (they are called headlines but lack many of the characteristic features of newspaper headlines from which they take their name). Dedicated TV news channels introduced several new text elements: a permanent bar that sums up the news as it is being reported, and even one or more text areas that present news or data unrelated to the news just reported; these have the form of a running text in a news-bar (on CNN, different terms are used for it depending on the category of news being dealt with at any given moment: newsbar, bizbar, and sportbar), or a permanent text area on which sports results, quotations and business statistics etc. are posted.

If we compare dedicated news channels with traditional TV news output, the difference in the amount of text displayed and the overall layout of the screen is so prominent that it encourages us to consider such channels a special type of media: the actual area for the moving picture is sometimes restricted to less than half of the total screen area; also, presenting two totally independent pieces of news at the same time is a completely new feature. It should be stressed that until these news channels appeared, there was hardly any medium that combined moving pictures with spoken and written text in such a balanced way.[6]

In this paper I will look more closely at the first type of text bar mentioned above—the permanent bar summing up the news. In the example below (Fig. 1), it contains the text FIGHTING OIL FIRES.

 

Fig. 1 A sample screen

 

My corpus consists of 263 subtitles taken from the permanent subtitle area on CNN International, shown during 5 hours of broadcasting between 28th and 30th March, 2003, i.e. from live coverage of the war in Iraq.[7] This type of real-time reporting is characteristic of CNN (and most other news channels). It should be noted that the use of subtitles in this kind of reporting differs from the way in which they are employed in standard news broadcasts. This also explains why there are very few subtitles present in the corpus that deal with any topic other than the war, and even these are connected to the war in some way (e.g. bombing in Israel).

The subtitles can be divided into two main types according to the functions they serve:

(1) Caption

This type corresponds to the traditional use of captions in newsreels, TV news and newspapers. They can be divided into the following sub-types:

 

1a. Photograph caption type                12 instances                 4.6 %

This sub-type corresponds to the way captions are used with photographs in newspapers, e.g.:[8]

23

29.03.2003

00:06

BAGHDAD EXPLOSION

 

1b. Name of a speaker or location       36 instances                 13.7 %

This sub-type was most common in traditional TV news (that took it over from film newsreels):

5

28.03.2003

23:45

BAGHDAD

10

29.03.2003

00:02

UNITED NATIONS

12

29.03.2003

00:02

SERGEY LAVROV
RUSSIAN AMBASSADOR TO U.N.

 

1c. General topic                                             19 instances                 7.2 %

This type of caption gives the general idea of what is being discussed, without relating directly to the picture. It has no direct equivalent in other media; the closest would probably be subheads or sideheads used to make orientation in a printed text easier.

41

29.03.2003

17:41

WAR PERSPECTIVES

46

29.03.2003

17:48

EDITORIAL JUDGEMENT

69

30.03.2003

00:29

BAGHDAD POLITICS

86

30.03.2003

00:43

BRITISH ATTITUDES
TOWARD THE WAR

 

The “caption” type subtitles amount to 67 instances, i.e. 25.5 % of all subtitles.

 

(2) Summing up the news

In the examined corpus, 196 subtitles fell into this category, i.e. 74.5 %. These could be called “self-sustaining” subtitles: they would be able to convey the essence of the news by themselves. However, they are not intended to be read together as a continuous text, and if we read them in this way, we find their coherence clumsy without the surrounding context of picture and speech:

 

1

28.03.2003

23:42, 44

EXPLOSION JUST HEARD IN KUWAIT CITY

2

28.03.2003

23:42

PLUMES OF SMOKE FLOATING
ACROSS
KUWAIT CITY AFTER EXPLOSION

3

28.03.2003

23:44

U.S. RECENTLY SAID IT KNOCKED OUT
IRAQI MISSILE POSITIONS NEAR
KUWAIT

4

28.03.2003

23:44

REPORT: PATRIOT MISSILES INTERCEPTED
IRAQI MISSILE ATTACKS ON
KUWAIT CITY

6

28.03.2003

23:50, 51, 53

EXPLOSION HEARD IN KUWAIT CITY

7

28.03.2003

23:53

KUWAIT OFFICIALS CONFIRM
MISSILE HIT
KUWAIT

8

28.03.2003,
29.03.2003

23:56, 0:08

FLAMES AND SMOKE SEEN
RISING OVER KUWAIT CITY

 

The coherence is further diminished by the occasional inclusion of “caption type” subtitles: e.g., subtitle 5 has been omitted from the above sequence:

 

5

28.03.2003

23:45

BAGHDAD

 

 

Characteristics of the text-type

 

The following analysis is based on the 196 subtitles that form type (2) above.[9]  I shall compare the proportions of various clause-types with those found in newspaper headlines (the most frequently analyzed representatives of little texts). I will also point out one specific feature of punctuation that does not appear in other types of texts.

Two prominent syntactic features are often stressed in descriptions of newspaper headlines. They are the high frequencies of elipted clauses and nominal clauses, resulting in a high proportion of non-finite clauses. In the present material, the sentence types are distributed as shown in Tables 2 and 3.

 

 

Subtitles

 

Newspaper
headlines
(Chovanec, 2003)

 

occurrences

%

%

Finite clause

111

56.6

59.5    

Non-finite clause

85

43.4

40.5    

Total

196

100.0

 

 

Table 2 Distribution of finite and non-finite clauses

 

 

Subtitles

 

Newspaper
headlines
(Chovanec, 2003)

 

occurrences

%

%

Elipted clause

70

82.4

            61.2

Nominal clause

15

17.6

            38.8

Total

85

100

 

 

Table 3 Distribution of elipted and nominal clauses

 

Table 2 shows that the proportion of non-finite clauses in the corpus almost equals that in newspaper headlines. It is important to note that the “caption” category, if included in the analysis, would slant the results dramatically as it consists almost exclusively of noun phrases. It was left out for the reason already mentioned: the captions do not constitute sentences proper as they do not convey complete information. They can be seen as nominal predicates if considered in the context of the screen; e.g., the caption KEN POLLACK / CNN ANALYST could be read

“[The man on the screen is] CNN ANALYST KEN POLACK”.

Of the subtitles consisting of a nominal phrase, only those are included that can exist independently, e.g. ANTI-WAR PROTEST IN INDONESIA'S CAPITAL. Table 3 shows that they are less frequent than in newspaper headlines.

One more adjustment was made to the data: only clauses that carry the actual information were analyzed. As is mentioned below, many news subtitles begin by stating the source of the news (“IRAQ SAYS...” or “BUSH: ...”). These statements take the form of superordinate clauses. There are 17 instances with a finite verb, and 62 with a colon. These special instances cannot meaningfully be added to the statistics of the regular clauses. However, they seem to confirm the tendency toward non-finite clauses.

 

Punctuation

There is one feature that appears repeatedly in the corpus and this is the novel way in which the comma is used. The comma is rare in the subtitles—it only appears 12 times throughout the 196 samples (one subtitle actually contains two commas).[10] In no more than two instances it has the standard syntactic function of connecting two propositions:

 

29.03.2003

18:00

ANTI-AIRCRAFT FIRE OVER BAGHDAD,
PLANES HEARD OVERHEAD

30.03.2003

00:51

DESPITE QUARANTINES, MYSTERY
DISEASE REPORTS ON THE RISE

 

In most cases (the remaining 10 occurrences) it replaces the conjunction and:

 

29.03.2003

17:47

RUBIN: RUMSFELD, ADVISERS
MISJUDGED IRAQI CAPABILITIES

30.03.2003

00:41

POLLACK: SOMALIS IN MOGADISHU
USED HUMAN SHIELDS, CIVILIAN DISGUISE

30.03.2003

00:48

U.S., IRAQ EXCHANGE ACCUSATIONS,
CLAIMS ABOUT CIVILIAN CASUALTIES

30.03.2003

01:30

IRAQI'S IN THE SOUTH NEED
FOOD, WATER

30.03.2003

12:35

MARINES: IRAQIS MAY HAVE STORED
WEAPONS, MUNITIONS IN MOSQUE

30.03.2003

13:12

IRAQI GEN: COALITION “CRIMINALS
AGGRESSORS, INFIDELS”

30.03.2003

13:16

IRAQI GEN. HONORS IRAQI SUICIDE
BOMBER AS “HERO, MARTYR”

06.04.2003

00:31

U.S., KURDISH FORCES
ON THE MOVE IN
NORTHERN IRAQ

06.04.2003

00:43

WORLD HEALTH ORG.: HOSPITALS
LACK ANTIBIOTICS, ANESTHETICS

 

It could be argued that in some of these subtitles the comma suggests enumeration and does not necessarily stand for and; however, and would be more appropriate in all of them, and is obligatory in some (“U.S., IRAQ EXCHANGE ACCUSATIONS”).

When searching for the conjunction and, only 7 instances are localized in the corpus (so that substitution by a comma is more frequent than leaving and in the text):

 

28.03.2003

23:56

FLAMES AND SMOKE SEEN
RISING OVER KUWAIT CITY

29.03.2003

00:04

CNN TURK: FORMER AND CURRENT
GOVT. MINISTERS ON BOARD PLANE

29.03.2003

17:42

U.S. ISSUES WARNING TO
SYRIA AND IRAN

30.03.2003

00:36

MCCHRYSTAL: “IT LOOKS AND
FEELS LIKE TERRORISM”

06.04.2003

00:44

NUMBERS OF CIVILIAN DEAD AND
WOUNDED NOT VERIFIED

06.04.2003

01:01

IRAQI TV AIRS VIDEO SAID
TO BE OF HUSSEIN AND SONS

06.04.2003

01:35

BUSH AND BLAIR TO DISCUSS
POSSIBLE U.N. ROLE IN POST-WAR
IRAQ

 

Only one rule seems to hold concerning these substitutions: and is only replaced when it connects sentence elements, not when it connects clauses.[11] Otherwise, no logic can be found in the distribution of the two options: both of them seem to appear in all types of connections, cf. RUMSFELD, ADVISERS x HUSSEIN AND SONS (where “and his” would probably be used in longer texts).

The most logical expectation would be that the usage is governed by the need to save space. But the substitution occurs even in very short lines, e.g. “FOOD, WATER”. So it seems that this substitution has become rather habitual, although it is very likely that it was originally intended to save space.

A very frequent punctuation mark is the colon. Its function will be discussed in the following section.

 

Differences in style and contents

 

The main difference between the subtitles and newspaper headlines lies in style rather than in syntax or punctuation. They completely miss the one feature that makes newspaper headlines unique (and attractive for linguists): inventiveness and playfulness. No puns, alliterations, allusions or other embellishments can be found in TV news subtitles; they are just plain informative sentences. This is easily explained: first, there is no need for such effects, as these texts do not serve the purpose of attracting readers; TV channels rely mainly on the power of the moving image. Second, there would be no time to formulate them, as subtitles often have to be created and displayed in seconds.

Another common feature of the subtitles is the stating of information sources. Stating sources is generally important in news reporting. The sources become especially prominent in the subtitles where they can sometimes take up a very high proportion of the small space available:

22

29.03.2003

00:05

COALITION COMMAND SPOKESMAN:
CANNOT CONFIRM REPORT

 

But this is probably the price that has to be paid for the real-time character of the coverage: there is often not enough time to verify the news, and stating the source is necessary for the news channel to avoid responsibility if the news later proves to be inaccurate.

Several examples of stating sources are found in the tables above. The source is always given at the beginning of the subtitle as the subject of a complex sentence containing an object clause which conveys the main message; the object clause is linked with the subject by a finite verb (most commonly “SAY/SAYS/SAID”—14 instances), or—more frequently—by a colon (62 instances). The source may be approximate (“ARAB MEDIA:”) or exact (“U.S. MARINES:”); interestingly, “CNN:” sometimes appears as the source (7 times, plus once in the form “CNN TURK:”). Often the name of a person appears in this position (“BUSH:”).

Personal names often serve as sources in several successive subtitles when a longer speech or analysis is presented:

 

77

30.03.2003

00:39

KEN POLLACK
CNN ANALYST

78

30.03.2003

00:40

POLLACK: IRAQIS APPEAR TO BE
USING SOMALI TACTICS

79

30.03.2003

00:40

POLLACK: USE OF IRREGULAR
FORCES AN EXAMPLE OF SOMALI TACTICS

80

30.03.2003

00:41

POLLACK: SOMALIS IN MOGADISHU
USED HUMAN SHIELDS, CIVILIAN DISGUISE

81

30.03.2003

00:42

POLLACK: SADDAM TRYING TO
CONVINCE
U.S. TO STOP WAR

82

30.03.2003

00:42

POLLACK: SADDAM COUNTING ON
U.S. PUBLIC OPINION TO END WAR

 

Even though this usage still falls under “giving sources”, it clearly differs from the instances when the source is given to distance the news channel from the contents of the news; it could rather be seen as the combination of a caption (giving the name of the speaker), and a resume of the speech. Stating the sources of the news—and the means employed to signal the reliability of the news in general—would certainly deserve further attention.

 

Conclusion

News-bars and subtitles of all types represent a specific type of text. They share some characteristics with other “little texts”, especially certain syntactic features, and differ from them in others: in addition to the innovation in the use of the comma, they bring interesting changes to the use of on-screen text in general. A larger corpus comprising data from other news channels, as well as from news on regular TV channels would be necessary to draw more definite conclusions. Other interesting areas for further research are the relationship between subtitles and the context provided by the picture and speech, as well as the use of these texts in other languages than English.

 

References:

 

Halliday, M.A.K. 1994. An Introduction to Functional Grammar. London: Edward Arnold.

McLuhan, Eric and Frank Zindrone (eds.). 1995. Essential McLuhan. Ontario: Anansi Press.

Chovanec, Jan. 2003. “The Uses of the Present Tense in Headlines.” Theory and Practice of English Studies. Volume I. Proceedings from the 7th Conference of British, American and Canadian Studies in Brno, September 2002. Brno: Masaryk University: 83-92.

 

 



[1] The term comes from Halliday (1994)

[2] E.g., TV news get one X more that news-reels as TV uses spoken text alone to relay some pieces of news

[3] Unfortunately, this was not the only time that prognoses did not come true. McLuhan also claimed in 1973 (1995: 294-5): "People will not accept war on TV. They will accept war in movies. They will accept it in newspapers. Nobody will accept war on TV. It is too close."

[4] The Internet encompasses all areas of interest, and however flexible it is in the use of various files, including pictures, sound, and even video sequences and films, it still remains a text-based environment. TV commercials could also serve to support this view: they seem to rely on printed text more than ever. In the old days Brand names and slogans used to be the only printed texts in commercials; today we find many texts that are spoken and appear on the screen in print at the same time; indeed, the letters themselves often become the main visual effect present. This tendency is particularly strong in the U.S., which is traditionally seen as a typical TV (i.e. picture-oriented) culture.

[5] These channels (e.g. CNN, or TA3 in Slovak) have their own entry in Table 1 because their way of presenting the news differ substantially from general TV channels in many respects.

[6] Technically speaking, a balanced combination of these three channels appeared long before, in subtitling foreign movies. But these subtitles were never part of the original work: they did not bring new information and were added to the original format to compensate for an insufficiency on the viewer's part.

[7] The subtitles were photographed from the TV screen in 9 sessions of various length (from 8 to 52 minutes, with the exception of two sessions that only consisted of one screenshot due to technical problems). The total monitoring time was 302 minutes. In the monitored periods of time, all subtitles were documented as they appeared, although a few may have passed unnoticed. Several subtitles were re-used 2 to 3 times by the editor; in those cases, the duplicates were not included in the statistics.

[8] The examples include the serial number, the date and time of broadcasting, and the actual text displyed on the screen. are given with theirin the form o this

[9] The reason for limiting the analysis to this category is discussed below.

[10] The comma is quite frequent in the caption-type subtitles which were left out of the analysis. It is used there as a separator in a location specification ("NEAR BASRA, IRAQ")—7 instances, or between location and time ("HAIFA, MARCH 5TH")—1 instance.

[11] However, as has been mentioned before, complex sentences are very rare so that a larger corpus would be necessary to confirm this.