DOCUMENT RESUME

ED 408 944 IR 018 357

AUTHOR Mokhtar, Fattawi B. TITLE The Effects of Dubbing Versus Subtitling of Television Program. PUB DATE Jan 97 NOTE 11p.; In: VisionQuest: Journeys toward Visual Literacy. Selected Readings from the Annual Conference of the International Visual Literacy Association (28th, Cheyenne, Wyoming, October, 1996); see IR 018 353. PUB TYPE Reports Research (143) Speeches/Meeting Papers (150) EDRS PRICE MF01/PC01 Plus Postage. DESCRIPTORS Audiovisual Aids; Audiovisual Instruction; College Students; Comparative Analysis; Comprehension; *Educational Television; Foreign Countries; Higher Education; *Language Processing; Programming (Broadcast); Second Languages; *Television Viewing; *Translation; Use Studies IDENTIFIERS *Dubbing; *Subtitles (Television); Universiti Sains Malaysia

ABSTRACT The purpose of this study was to investigate viewers' knowledge of program content under various television translation modes and viewing experiences. Subjects were 176 students from the Center for Matriculation Program, Universiti Sains Malaysia in Penang, Malaysia. The Spanish version of an instructional television program was used; the program was translated into Malay using dubbing and subtitling. Subjects were randomly assigned into 4 groups of 44 subjects each. The study examined differences in viewers' knowledge of content in programs with and without translation, and for the translated program, differences in content knowledge among viewers who watched the programs under single viewing and repeated viewing conditions. Findings indicated the following:(1) students who viewed the program without translation had significantly lower scores on a multiple-choice test compared to those who viewed the same program in translations, both under single and repeated viewing conditions;(2) under single viewings of translated instructional television, translation modes did not contribute to any differences in knowledge of program content, however, under repeated viewings, translation in dubbing and subtitles without sound helped viewers to acquire more knowledge on content than translation in subtitles with sound;(3) repeated viewings appeared to significantly help viewers to acquire more knowledge on program content than single viewings, however, effects of repeated viewings were not consistent in all translation modes; and (4) interactions between translation modes and viewing experiences were statistically significant. (Contains 27 references.) (AEF)

******************************************************************************** Reproductions supplied by EDRS are the best that can be made from the original document. ******************************************************************************** The Effects Of Dubbing Versus SubtitlingOf Television Program by Fattawi B. Mokhtar Abstract The knowledge of program contentamong viewers who watched an instructional television program translated into several translation modes and under different viewingfashions was investigated. One hundred seventy six college studentswere randomly selected and randomly assigned into eight experimental groups. A 4x 2 factorial design was employed.The dependent variable was the viewers' scoreson multiple-choice comprehension test questions on the program content administered immediately after treatments. Severalsignificant differences among variables were found and discussed.

Introduction process point of view. When a program is When language becomes a barrier to translated insubtitles,it automatically communication, the classicalternative changestheconveyanceof information solution to this problem is translation. fromauditorydominancetovisual There are twomodes of translating dominance (d'Ydewalle et al. 1991). With television programs. First is a translation translation by dubbing, the translation process known as "dubbing." With this maintainstheoriginalmethodof method, every spoken word, including communication used by television. voice-over narration as usually employed in With translationinsubtitles and in instructional programs, is translated into dubbing then, the television medium tends the target language. to communicate in two different fashions: The second mode of translationis a with visual dominance in the former and process known as "subtitling."With this auditory dominance in the latter. Whether method, spoken words, either in dialogues, thesedifferencesaffectviewers' monologues (on-camera voices) or voice- comprehension of the program contentis overnarrations,aretranslatedand unknown at this point. presented in the form of text superimposed Technically, translation by subtitles will on the visual, typically at the bottom part preserve all audio signals in the original of the screen. Hence, with this method the soundtrack (including music, sound effects, dominant meaningful symbolsare and spoken words) exactly as in the original presented in visual form while thenon- production. Aesthetically, this isone of the meaningful spoken words remainas major advantages of subtitlesover dubbing auditory symbols. (Voge 1977). However, when considering In many non-English speaking countries, the communication process,the spoken translated broadcast materials, either for words (in the original language)are turned entertainment or instructionalpurposes, into non-meaningful symbols. The role of constitute a great if not dominant portion these symbols during viewing is notyet of the overall number of mediatedresources understood. in those particular countries. Translation of If Schram's Communication Model is used imported programs isa viable effort to asatheoreticalground,thosenon- overcome the problem of high cost versus meaningful spoken words will be categorized limited audience for localproduction as "noise," which according to the model (Kilborn 1989). will impede the communicationprocess (Schram 1954 in Heinich et al. 1989). An Statement of the Problem investigationinvolvingretentionand Translation by dubbing or subtitlingseems elimination of the soundtrack in subtitled to impose several new characteristics on television might answer this question. the product of this. process. There are Another formof writtentexton severalquestionsrelatedtothese television is the "caption."Captions and characteristics which need to be addressed, subtitlesare identical in physical format particularlyfrom the communication and in the way they are presentedon the U.S. DEPARTMENT OF EDUCATION Office of Educational Research and Improvement "PERMISSION TO REPRODUCE THIS EDUCATIONAL RESOURCES INFORMATION MATERIAL HAS BEEN GRANTED BY CENTER (ERIC) This document has been reproduced as received from the person or organization 23 Alice D. Walker originating it. Minor changes have been made to TO THE EDUCATIONAL RESOURCES improve reproduction quality. 2 rEgir CCM' MAR INFORMATION CENTER (ERIC)." Points of view or opinions stated in this document do not necessarily represent nffirinl (WOO nnqitinn nr nnliry screen.They are different, though, in the Hypothesis 3.There is a significant sense that captions use the samelanguage as difference in the knowledge of program the soundtrack, whereassubtitlesare content between viewers who viewed a presented in a different language.Several television program in a single viewing studies on theeffectsofcaptioned experience and those who viewed the same television on learning among hearing- program in a repeated viewing experience. impaired and normal-hearing viewers Hypothesis4.There aresignificant indicate that captioned television improves interactions between modes of translation learningsignificantly (Neuman1991, (dubbing or subtitles) in an instructional Markham 1989, Koskinen1988, Bean television program and viewing experiences. 1989). Whether subtitled television affects comprehension and learning as consistently Methods as captioned televisionis unknown at this Research Design point. A 4 x2 factorial design with random selection and random assignment was Purpose of the Study employed in this study (Keppel 1991). The purpose of thisstudy was to Translationmodeswerethefirst investigate viewers' knowledge of program independent variable (Factor A), with four content under various televisiontranslation levels:dubbing (al); subtitles with sound modes and viewing experiences. The (a2); subtitles without sound (a3); and researcher was interested in investigating without translation (a4). whether there were any differences in the The second independent variable (Factor knowledge of program contentamong B) was the viewing experience with two and repeated viewers who watched an instructional levels:single viewing (b 1) television program that was translated in viewing (b2). several translation modes and variations,. The dependent variable was the viewers' those who watched the program translated, scores on multiple-choicecomprehension and those who saw the same program test questions on the program content.This without translation. test was administered immediately after The researcher was also interested to treatments. Each mean cell and the main test investigatewhether therewere any effectswerecomparedto the differences in the knowledge of program hypotheses.When there was significant content among viewers who watchedthose difference, post hoc tests were conducted to translated and not translated programs pinpoint where the difference was. under single viewing condition to those who saw the same, but underrepeated viewing Subjects condition. One hundred seventy sixstudents from the Center for Matriculation Program, Research Hypotheses UniversitiSains Malaysia in Penang, This study advanced thefollowing Malaysia, were selected as the subjects in hypotheses: this study. They were randomly selected and Hypothesis1.Thereisa significant randomly assigned into eight experimental difference in the knowledge of program groups. Each group consistedof twenty-two content between viewers whoviewed an subjects. instructionaltelevision programwith Because of the nature of selection criteria translation and those who saw the same to enroll in this program and therigidness program without translation. of the program itself, matriculation students Hypothesis 2. There is asignificant are highly homogenousin terms of their difference in the knowledge of program academic backgrounds and age levels. content among viewers whoviewed an instructionaltelevision program with Instrumentation different treatments of translation by Discussions on instrumentation are dubbing or subtitling. presented in two sections. The firstsection

24 is on the materials,inthis case the character generator machinewas used. translated instructional televisionprograms The lettering was Futura regular in 36point where the independent variableswere size. Kemp et al. (1985) notes that for manipulated. The second section is on the recognized legibility standards forease of comprehension test questions in which the reading, the letter size should beat least one dependent variable was measured. twenty-fifth the height of thescreen. On a 23-inch television monitor with 15 inches The materials of screen height, 36 point Futura satisfied The Spanish version of an instructional this requirement. The letteringwas in white televisionprogramentitledCardio and bordered in black. A maximum of2 PulmonaryResuscitation(CPR)For lines with 2 inches of margins on each side Bystanders (revised), produced by Pyramid and a minumum of 2 inches of bottom Film and Video (1992) was chosen for this space were uniformly employed. study. The English versioh of the program The exact translation used in the dubbing was not chosen because English is the was used in subtitles.The text was not second language in Malaysia.Narrations summarized asisusually done in the and other on-camera voices in the original subtitling of feature programs. soundtrack might to a certain degree have To see the placebo effects, the original been understood by the viewers. Hence, this version in Spanish without translationwas effect was a potentially confounding used. For the subtitles without sound variable for the subtitled version of the condition variable, the same subtitled instrument.Besides language, the choice version was used, but played in mute. was also based on the following criteria: All versions of the program, including the i) The length of the program was not original without translation, were recorded more than 30 minutes. on half-inch Sony high-grade VHS tapein ii) The program content was something PAL color system. of which the researcher believes the subjects have some basicentry-level The comprehension test questions knowledge, but no knowledge in detail. All questions were developed by the iii) The program was produced in a researcher based on the program content typical documentary fashion where verbal conveyed through both channels (audio and modalities such as voice-over narrations video) as a whole.To ensure content and on-camera voicesdominatethe validity and precise wordingson the conveyance of the messages. questions, the printed CPR procedures iv) The overall program contentwas written by the American National Red something believed to be meaningful or at Cross (1978),Feldmanet al. (1975) and least stimulating to a subject's interest. Seymour (1981) werereviewed.After a v)Subjects in this study were within the series of pilot test, the final version of the scope of the intended audience specified in 35 comprehension questions was translated the program. into the Malay language. The reliability of Upon receipt of permission of the thecomprehensionquestionswas copyrightowner,theprogramwas determined tobe .79 by the Cronbach duplicated and translated into Malay using Alpha technique computed bySPSS-X two modes of translation,dubbing and program. subtitling. The translation was based on the original soundtrack in English. Experimental Procedures For the program translated in dubbing, the All viewings were carried out in the voice-over was narrated by a semi- projectionroom. The 60-person capacity professional narrator.On-camera voices roomhas semi-permanent seats and was were spoken by amateur actors and equipped with two 23-inch television actresses.Original sound effects and music monitors, a VCR,and several other were reused whenever possible. facilities for projection.The acoustic and For the program translated in subtitles, a ambient light controls of the room were

25 BEST COPY AVAILABLE adequate. The temperature was controlled second viewing, the subjects were given by means of three window-typeair another 10 minutes to complete the test. conditioning systems. The exact same procedures were repeated in For the purpose of this study, only 44 every session.In the subtitles without seatswereallocated.Two23-inch sound condition, the same subtitled version television monitors were set up to ensure of the program was used, but shown in mute. that all viewers were seated no further than Allsessions were conducted by the the maximum distance suggested.Heinich researcher himself without any assistance. et al. suggested that one 23-inchtelevision monitor should serve no more than thirty Results students (Heinich et al. 1989).Hence, for Data were analyzed using Two-Way 44 viewers, two 23-inch monitors were Analysis of Variance (ANOVA), One-Way more than generous. Theseating distance Analysis of Variance (ANOVA) and t-Tests and the angle of viewing were also set in to test the hypotheses. The alpha level accordance to Heinich's suggestions. was set at .05. When there wassignificant The experiment sessions were held on difference between the means, Tukey four consecutive evenings. All seats were test procedures were conducted to numbered from 1to 44.All set-ups were pinpoint where the difference was. readied at least thirty minutes before sessions were to begin. Single Viewing The descriptive statisticsin Figure 1 Experiment Implementation revealed that under the single viewing Randomly chosen subjects were ramdomly experience condition, students who viewed assigned into 4 groups of 44 subjects each. the program without translation had the Minutes before the session was to begin on lowest score (Mean = 16.45, SD = 3.5). the day of the experiment a representative Analysis of variance (ANOVA) tests and of the group in that particular session drew statistical tests using Tukey test procedures to determine on which versionof the indicatedthatthistreatmentwas program the group would betested. significantly different than the other three Immediatelyafterviewing, the treatments at p = .000 level (ANOVA) and comprehension test was administered to all p <.05 level (Tukey). subjects in the group.The subjectswere Figure 1 instructed to answer by circling the correct DESCRIPTIVE STATISTICS responses on the question paperitself. FOR TRANSLATION MODES They were given 30 minutes to complete AND VIEWING EXPERIENCE At the end of the 30 minutes, all the test. TRANSLATION MODES subjectswereaskedtostop.One VIEWING representative of the group drew to EXP. Dubbing SubtitleSubtitle No Trans determine whether subjects seated in odd +Sound - Sound numbered or even numbered chairs would stay for a repeated viewing.Students seated in unselected numbers were asked to hand in Single Mean22.36 21.23 22.50 16.45 SD 3.1 3.2 3.0 3.5 their comprehension test answers and leave n 22 22 22 22 the room. The remaining subjects would be in a Rep. Mean29.68 25.77 27.91 18.55 repeated viewingcondition. A brief SD 1.6 2.4 1.9 3.1 explanation was given on this viewing n 22 22 22 22 condition.They were requested not to change their seats. The program was then shown for the second viewing. The subjects Mean26.02 23.50 25.20 17.50 TOT. SD 4.4 3.6 3.7 3.4 were told that theycould answer the 44 questions while in the second viewing if n 44 44 44 they wished.Upon completion of the The students in subtitles without sound SD = 1.9)was not significantly different treatmenthad the highest mean score than dubbing treatment(Mean= 29.68, (Mean = 22.50, SD = 3.0).Statistical tests SD =1.6).Pairsof treatments under however, indicatedthatthere were repeatedviewingconditionswhich no significant differences between this significantly differed fromeach other are treatment, the dubbing treatment (Mean = shown in Figure 3. Data analysis using Two- 22.36, SD 3.1), and subtitles with sound Way Analysis of Variance (ANOVA) treatment (Mean = 21.36, SD = 3.2) at p indicatedthatinteractionbetween <.05level.Pairs of treatments which translation and viewing was significant at p significantly differed under single viewing = .000 level.Further examination of the conditions are shown inFigure 2. cell means revealed that the mean score of

Figure 2 Figure3 PAIRS OF MEANS UNDER PAIRS OFMEANS UNDER SINGLE VIEWING REPEATED VIEWING WHICH ARE SIGNIFICANTLY WHICH ARE SIGNIFICANTLY DIFFERENT DIFFERENT

Dub. SubtitleSubtitle No Trans. Dub.Subtitle SubtitleNo Trans. + Sound - Sound + Sound - Sound

Dub. * Dub. * *

Subtitle Subtitle + Sound * + Sound *

Subtitle * Subtitle * * - Sound - Sound

* p < .05 * p < .05

Repeated Viewing translation modes treatment under single Intherepeated viewingcondition, viewing conditions changed inconsistently students viewing the without translation under repeated viewing conditions. This treatment again had the lowest meanscore appeared to indicate that an interaction (Mean = 18.55, SD = 3.1). Dubbing between translation modes and viewing treatment (Mean = 29.68, SD = 1.6) experiences was truly present. appeared to be the best of the three translation modes.Statistical tests using Tests of Hypotheses Tukey test procedures indicated that under All hypotheses were tested using Two- this viewing condition, without translation Way Analysis of Variance (ANOVA), One- treatment was significantly different than Way Analysis of Variance (ANOVA), and t- all other treatments.Subtitles with sound Tests. The alpha level was set at p < .05. treatment (Mean = 25.77, SD = 2.4) was When there was significant difference significantlydifferent thansubtitles between themeans,the Tukeytest without sound treatment (Mean = 27.91, procedure was used to pinpoint where the SD = 1.9) . Also significantly different difference was. from subtitles with sound treatment was Hypothesis4.There aresignificant dubbing treatment (Mean = 29.68, SD = interactions between modes of translation 1.6).Surprisingly,however,subtitles in instructional television programs and without sound treatment (Mean = 27.91, viewing experiences. H4 was tested first

27 BEST COPY AVAILABLE because the outcome of the hypotheses on p < .05 level. Another one-way ANOVA to interaction will have a strong implication test the mean of translation modes under on the testing procedure of the other repeated viewing conditions was conducted. hypotheses (Keppel1991). Analysis of Again, the analysis indicated that at least Variance (ANOVA) revealedthatthe one translation mode was significantly interaction between translation and viewing different than the others F (3, 87) = 94.83, was significantF (3, 168) = 6.429, p = p = .0000.The outcome of Tukey test .000. H4 therefore was accepted. procedures indicated that the without- Hypothesis1.Thereisa significant translation mode was significantly different difference in the knowledge of program than all other translation modes at p < .05 content between students who viewed an level. Hence hypothesis 1 was accepted. instructional televisionprogram with any Hypothesis2. Thereisa significant modesorvariationsoftranslation difference intheknowledge of program compared to students who viewed the same content among students whoviewed an program without translation. instructional television programwith Figure 4 Figure 5 ANALYSIS OF VARIANCE ONE-WAY ANOVA SUMMARY TABLE (SINGLE VIEWING)

Source SS DF MS F Sig. F Source SS DF MS F Sig. F

Translation (T)1957.295 3 652.43281.08 .000 Between Group 534.45 3 178.15 16.89.000 Within Group 885.90 84 10.54 Viewing (V) 1031.114 1 1031.114128.144.000 T x V Interac. 152.205 3 51.735 6.429.000 Total 1420.36 87 Explained 3134.6147 449.08855.811 Residual 1351.818168 8.047 p <.05

Total 175 25.688 4495.432 translationindifferentmodesand p <.05 variations. The same analysis and test mentioned in Analysis of variance (ANOVA) revealed the testing of HI were used to test this that the translation's main effect was hypothesis. The conclusion, however, was significant F (3, 168) = 81.082, p = .000. rather open-ended.Under single viewing However, because the interaction between conditions, Tukey test procedures indicated translation modes and viewing experiences that none of the three translation modes, (H4) was significant (p = .000), a judgment (i.e. dubbing,subtitleswithsound and based on the translation's main effect alone subtitles without sound) were is not sufficient.All translation cell means significantly different than each other at p under both viewing experienceconditions < .05 level.Hence, under single viewing should be examined and tested. conditions, H2 was rejected. Two single-factor (one-way) analyses of variance (ANOVA) were conducted. The Figure 6 analysisindicatedthatatleastone ONE-WAY ANOVA translation mode is significantlydifferent (REPEATED VIEWING) thanthe others F (3, 87) = 16.89,p = .0000.Tukeytestprocedureswere conducted to pinpoint where the significant Source SS DF MS F Sig. F difference was. The outcome of the Tukey Between Group 1578.04 3 526.01 94.83.000 test procedure indicated that the without- Within Group 465.90 84 5.54 translation mode condition was significantly different than all other modes Total 2043.95 87 and variations of translation conditions at p <.05

28 7 A contrasting result was found under viewing conditions. This phenomenon can repeated viewing conditions.The Tukey be comprehended by understanding that test procedures indicated that subtitles with televisionis a dual-sensory medium. Part sound was significantly different than of the information is conveyed through dubbing and subtitles without sound at p < visual channels andthe rest are through .05 level.Hence, under repeated viewing auditory channels.Visual channels utilize condition, H2 was accepted. visualsymbols(inthiscase,scenes Hypothesis 3.There is a significant demonstrating the CPR procedures), while differenceinthe knowledge of program auditory channels utilize spoken words (in content between students who viewed a this case the narration and dialogues television program in a single viewing givingthe detailsabout CPR). In the experience and those who viewed the same process of interpreting each symbol in program in a repeated viewing experience. television, viewers switched back and forth Analysis of Variance (ANOVA) revealed between the two channels (Findhal 1971) to that theviewing main effect was decode the message.If the symbols were significant F (1, 168) = 128.144, p = .000. within both the sender's and the receiver's However, since interaction between viewing field of experience (Schramn 1954 in experiences and translation modes was Heinich et al. 1989), the symbols could be significant, further analysis was needed decoded.Whenthesymbolswere before a conclusion on this hypothesis successfully decoded, communication was could be drawn. successful. In other words, the message was Four t-Tests with Independent Samples understood.In the situation of viewing a were conducted.Group means of each programwhere the auditory channel was translation mode treatment under single full of symbols which were not within the viewing conditions were compared with the viewer's field of experience (i.e., unknown group means of the same translation mode words in the program without translation), treatmentunderrepeatedviewing the message in the auditory channels could conditions. Group means ofwithout not be decoded.Hence, communication translation under single viewing conditions through auditory channels did not occur. were also compared to means of without Viewers were forced to depend upon only translation under repeatedviewing one channel, the visual channel.Messages conditions. in words were not recieved by the viewers, Allt-Testresultssupportedthe meaning they missed part of the program hypothesis that repeated viewing conditions content. were indeed significantly differentthan It is concluded, therefore, thatviewers single viewing (p < .05) regardless of the acquired the least amount of knowledge on translation modes. H3therefore was the program content when viewing an accepted. instructional television program produced in aforeignlanguagewithout translation Discussion and Conclusion on the compared to when viewing the same Findings program with translation. Hypothesis 1 The first hypothesis was developed in Hypothesis 2 relation to the question of whether or not The second hypothesis was developed to translations would make any difference on test whether dubbing, subtitles with sound, viewers' comprehension of the program and subtitles without sound, affects viewers' content. comprehensionof programcontent As expected,students who viewed the differently. The mean scores of students programwithout translationobtained who viewed the program under those three significantly lower mean scores on the conditions were compared. Two-Way multiple-choice test compared to students Analysis of Variance procedures indicated who viewedthesameprogramin thattranslation'smaineffectswere translations, both under single and repeated significant.At the same time however,

29

EST COPY AVAILABLE 8 :yt

modes and repeated viewing,the differences between interaction between translation subtitleswithout sound viewing experience was alsosignificant. dubbing and remained insignificant. Subtitleswith sound, Further analysis wasconducted. Two than of Variance for however, was significantlydifferent separate One-Way Analyses dubbing and subtitles withoutsound. The translation modes under singleviewing and two different effects of second viewing seem tobe more repeated viewing yielded prominent on dubbing and subtitleswithout results. Under single viewing, nosignificant sound than on subtitles withsound. Under differences were foundbetween the three Under repeated single viewing, the mean scoresfor dubbing modes of translation. and subtitles without sound were22.36 and viewing, dubbing and subtitleswithout sound different. 22.50 respectively. Underrepeated viewing, were also notsignificantly 29.68, an repeatedviewing the mean score for dubbing was However,under increment of 7.32.The mean score for conditions, subtitles with sound was dubbing and subtitles without sound was27.91, an significantly different than single viewing, The mean scores incremenet of 5.41. Under subtitles without sound. with sound was with sound were the mean score for subtitles of the students in subtitles 21.23, and under repeatedviewing was significantly lower. 4.59. It was As mentioned in theStatement of the 25.82, an increment of by subtitles obvious that the incrementsof the mean Problem section, translation scores of studentsin dubbing and subtitles will preserve all audiosignals, including thanthe spoken words, without sound were more music, sound effects, and increment of the mean scoreof students in exactly asinthe original production. the subtitles with sound.Also, the standard However,when considering subtitles without communication process, thespoken words deviations for dubbing and turned into sound were 1.6 and 1.9respectively, while in the original language were the standard deviationfor subtitles with non-meaningful symbols. According to for the Schramn (1954 in Heinich etal.1989), sound was 2.4. The possible reasons symbols were differences in mean scoresand standard those non-meaningful in subtitles without soundand categorized as "noise" whichwould impede deviations subtitles with sound weredue to the absence the communication process.There was "noise" in these two that Schram's theory was and the presence of some indication modes of translation. true, as shown in thedescriptive statistics the effects of for subtitles with In subtitles without sound, where the mean score second viewing and theabsence of "noise" sound was 21.23, whilethe mean score for 22.50 (a probablyenabledthestudentsto subtitles without sound was concentrate more and tobetter comprehend difference of 1.27). Thestandard deviation The same effects for the subtitles withsound was 3.2, while the program content. for subtitles without could be noticed indubbing translation, the standard deviation where "noise" was alsoabsent. Although sound was 3.0, whichindicates that students employed visual and in subtitles without soundhad scores closer dubbing translation auditorychannelssimultaneouslyin to the mean thanthe students in subtitles without Subtitles with sound wasalso communication, while subtitles with sound. sound employed onlyvisual channels, both compared to dubbing to testthe prediction have transmitted in visual translation modes appearedto that the message equal effects undersingle channels and the messagetransmitted in statistically viewing and repeated viewing. audio channels wouldprobably affect the effects of The mean score for In subtitles with sound, viewers differently. second viewing were presentbut at the same dubbing was 22.36 (adifference of 1.13). "noise" were also indications that the time, the effects of Hence, there were present. The "noise"acted as a kind of an prediction was true.Unfortunately, not obstacle during viewing.As a result, the however, those predictions were scoresof test that showed increment ofthemean supported by the statistical translation mode was the significance of thedifferences. Under students in this

30 9 lower compared to the increment of the analyses indicate thatsingle viewings were mean score of students in subtitles without significantlydifferentthanrepeated sound. The "potential difference" between viewings, which appeared to be superior to subtitleswithsoundandtheother single viewing. translationmodes,whichwas not Further investigation of the interaction statistically significant under single viewing between viewings and translation modes, conditionsappeared to be statistically however, revealed that theeffects of significant under repeated viewing. repeatedviewings,wereparticularly The fact that there was no significant prominent onthe program translated in difference between dubbing and subtitles dubbing and subtitles without sound. Results without sound was indeed an unexpected indicating a significant difference between finding. In principle, comparison between single viewings and repeated viewings were dubbing and subtitles without sound could be not unexpected.Earlier studiesonthe equivalenttoa comparison between effectsofrepetition yieldedsimilar simultaneous audioand visual stimulants results (Ash andJaspen1953,Kendler, versus visual stimulants only. Earlier Cook and Kendler 1956, Cook 1960, Garza studies in this domain (Hoban and Van 1991). Omer 1950, Travers 1964, Findhal 1971, The program used in this study was 30 Reese 1983, Baggett and Ehrenfeucht 1983, minutes long. After the firstviewing, the Pezdek and Steven 1984, Drew and Grime subjects were given the first chance to 1987) supported the belief that both visual answer the comprehension questions. Since and audio channels working together were they were not told previously that they much more effective than either one alone. would be given a second viewing, they In fact, this was part of the theoretical probably tried their best to complete the ground upon which the second hypothesis testafter the first viewing.When they inthisstudywasbuilt.Particular were given the second viewing they seemed investigations into those earlier studies, to pay even more attention to the program. however, revealed thatnone of the Saturationeffectsseem tohavenot conditions used in those studies which occurred with that sequence and length of represented visual stimulants only were viewing. identical to the subtitles without sound In conclusion, repeated viewings seemed conditionusedinthisstudy.Any to significantly help viewers to acquire discrepancy in results between earlier studies more knowledge on the program content and this study might be associated with this than single viewings. However, the effects aspect of dissimilarity. of repeated viewings were not consistent in Itis concluded, therefore, that under alltranslation, modes. It was equally single viewings of translated instructional prominent on the program translated in television,translation modes didnot dubbing and in subtitles without sound, less contributetoany differencesinthe prominent on the programtranslatedin knowledge of program content among subtitles with sound, and least prominent on viewers. However, under repeated viewings, the program without translation. this study indicated that translation in dubbing and subtitleswithout sound Hypothesis 4 helped viewers to acquire more knowledge Thefinal hypothesis was developed to see on the program content than translation in the interactions between translation modes subtitles with sound. andviewingexperiences.Statistical procedures indicated that the interactions Hypothesis3 between the two factors weresignificant. The third hypothesis was developed to Sincetherewere noearlierstudies test whether single (one-time) viewings mentioninginteraction between viewings would make anydifferenceinthe and translation modes, no references to knowledge of program content as compared earlier studies could be made. torepeated(two-time)viewings. Data Discussion and conclusions related to the

31 BEST COPYAVAILABLE 10 interactioneffects between translation Feldman, S. and Ellis, H. (1975). Principles of resuscitation, 2nd.ed. Oxford, Great Britain:Blackwell Scientific modes and viewing experiences were Publication. combined with the discussion on hypotheses Findahl, 0. (1971). The effect of visual illustrations upon 2 and 3. Hence,theyare not presented perception and retention of news programmes. ERIC Document Reproduction Service No. ED 054 631.7- 801. again here. Garza, T. J.(1991). Evaluating the use of captioned video Language barrier in the use of foreign- materialsin advanced foreign language learning. produced instructional media is a universal Foreign Language Annals. v 24. Heinich, R.Molenda, M.& Russel, J. D. (1989). problem, particularlyamong non English Instructional media and the new technologies of speaking countries.Instructional media instruction. (3rd. ed). New York: McMillan. practitioners, as well as scholars especially Hoban, C. andVan Ormer.(1950).Instructional film research, 1918 - 1950.Technical Report No. SDC 269 - inthecountriesconcerned,should 7 -19. Port Washington, NY: US Naval Special investigate and address this problem more Devices Center. seriously.Technologicaladvancesin Kemp. J. E and Dayton, D. K. (1985). Planning and Producing Audiovisual Instructional Media 5th ed. New electronics and atmospheric exploitation York: Harper and Row. only helps human to overcome the Kendler, T.S., Cook, J.0., and Kendler, H. H. (1953). geographicalbarriers and distancein An investigation of the interacting effects of repetition and audience participation on learning from training communication, but notto language.It films. AmericanPsychologist, v 8, 378 - 379. seems, due tothelanguagebarrier, Keppel, G.(1991).Design and analysis, a researcher's McLuhan's dreamed global village is yet a handbook (3rd ed). Englewood Cliffs, NJ :Prentice Hall. Kilbom, R. (1989). They don't speak proper English:a reality. new look at the dubbing and subtitling debate.Joumal of multilingual and multicultural development. v. 10, 5, References 421 - 435. Koskinen, P. (1988). Using captioned television in American National Red Cross. (1978). Respitory and classroom reading instruction.Teaching English to the circulatory emergencies. The American National Red deaf and second language students. v. 6. n 1. 15 - 19 Cross Publication. Markham, P. (1989). The effects of captioned television Ash, P. and Jaspen, N. (1953). The effects and videotapes on the listening comprehension of beginning, interactions of rate of development, repetition, participation, intermediate and advanced ESL students. Educational and room illumination on learning from a rear projected Technology. v 29. n. 10. 38-41. rum hrennrtS. Port Washington, McLuhan, M. (1989).The global village: transformation N.Y: U.S. Naval Special Devices Center. rnodia in the 21st century. New York: Baggett, P. and Ehrenfeucht, A. (1983). Encoding and Oxford University Press. retaining information inthe visuals and verbals of an Neuman, S. and Koskinen. P.(1991). Captioned educational movie. Educational and communication television as 'comprehen- sible input': effects of incidental technology journal. 31. 23 - 32. word learning from context for languageminority Bean, R. and Wilson, R. (1989). Using closed captioned student. ERIC Document Reproduction Service. ED television to teach reading to adults. Reading Research and 332538. Instruction. v 28. n 4. 27 - 37. Pezdek, K. and Stevens, E. (1984). Children's memory for Cook, J. 0. (1960). Research in audio- visual communication. auditoryand visualinformationontelevision. In J. Ball and F. C. Byrnes (eds). Research, principle Developmental psychology. 20. 212- 218. and practices in visual communication. EastLansing: Pyramid Film & Video.(1992).Cardio pulmonary MichiganStateUniversity, NationalProjectin resuscitation (CPR) for bystanders (revised). 30 minutes Agricultural Communication. video. Santa Monica, California: Pyramid Film & Video. d'Ydewalle, G. Praet, C. Verfaillie, K. and Rensbergen, J. Reese, S. D. (1983).Improving audience learning from (1991). Watching subtitled television, automatic reading television news through between-channel redundancy. behavior. Communication Research, v 18, n 5. 650 - 666. ERIC document reproduction service No. ED 229777. d'Ydewalle, G. Van Rensbergen, J. and Pollet, J. (1987). Seymour, J. (1981). The heart attack survival manual: a Reading a message when the same message is available guide to CPR in acrisis.Englewood Cliff, New Jersey: auditorily in another language: The case of subtitling. In J. Prentice Hall. K. 0' Regan and A. Levy Shoen (Eds.). Eye movements: Travers, R. (1964). Research and theoryrelated to from physiology tocognition. 313-321. Amsterdam: audiovisual information transmission.Salt Lake City: Elsevier Science Publishers B. V. University of Utah, Bureau of Educational Research. Drew, D.G. and Grimes, T.(1987).Audio-visual Voge, H. (1977). The translation offilm:Subtitle versus redundancy and tv news recall. Communication research. dubbing. Babel , 23, 120 -125. 14. 452-461.

32 11 BEST COPY AVAILA:LE U.S. DEPARTMENT OF EDUCATION Office of Educational Research and Improvement (OEM) Educational Resources Information Center (ERIC) !ERIC'

NOTICE

REPRODUCTION BASIS

This document is covered by a signed "Reproduction Release tv- (Blanket)" form (on file within the ERIC system), encompassing all or classes of documents from its source organizationand, therefore, does not require a "Specific Document" Release form.

This document is Federally-funded, or carries its own permission to reproduce, or is otherwise in the public domain and, therefore, may be reproduced by ERIC without a signed Reproduction Release form (either "Specific Document" or "Blanket")..

(9/92)