Methods for Analyzing the Content of Motion Pictures
Edgar Dale
The purpose of the study described in this. report is two-fold: First, a method was devised for analyzing the content of motion pictures, and second, this method was used to analyze the content of typical motion pictures. It is the specific purpose of this article to describe the methods used for discovering the content of motion pictures.
A series of criticisms of and claims for theatrical motion pictures have been made which cannot be answered until studies have been made of motion-picture content. There is the charge, for example, that certain fundamental areas of human concern are not treated at all in motion pictures. It is further charged that there is preoccupation with certain areas of human living—a preoccupation which is wholly unjustified and sometimes harmful. A second type of charge is leveled at specific content within the motion picture. Some maintain that certain fine ideals of human living are consistently portrayed by current motion pictures. Others declare that the motion pictures are almost entirely preoccupied with the depiction of crime; the approval of race prejudice; the covert and sometimes explicit approval of sexual impropriety; and frequent display of vulgarity. These same persons maintain that, in general, the content of such motion pictures not only has a harmful effect upon Americans but also puts us in an unfavorable light abroad. A study of the content of motion pictures makes it possible to secure evidence on these disputed questions.
The only way we can know the effective content of a motion picture is through the responses that individuals make to it. Because of a common background of experience most individuals will react very similarly to certain images which they see on the screen. This agreement among individuals as to what they see on the screen represents the common denominator of communication. So, within cer-
( 245) -tain limits, there will be a series of reactions to a screen story which differ very little among individuals. We have used the word content, therefore, to cover the common reactions which we should expect typical individuals to get from a motion picture. It is true that if we wished adequately to analyze all the effective content of motion pictures it would be necessary to sample progressively the reactions of all possible viewers until we had reached a point where no significantly new reactions occurred. This it was manifestly impossible to do. Highly specialized reactions to motion pictures such as might be made by a specialist in the field of photographic art are therefore not included in our study.
It is evident from the nature of the charges mentioned above that two types of analyses of content are necessary. The first is a study of the general themes or the areas in which motion pictures have been developed. The second is a type of analysis which describes verbally, with much precision and detail, the content of a motion picture. A survey of the literature quickly disclosed that this evidence had not yet been secured and that analyses of the general and specific type were necessary.
The methods used to analyze films for their general themes must depend, of course, upon the type of evidence available regarding such content. The ephemeral nature of the motion-picture film makes it impossible to view the motion pictures of past years to discover their content. In many cases the films are not available and positives would have to be printed at a cost that would be prohibitive for the purposes of this investigation.
Our source of information concerning the pictures which had been produced during these years was Harrison's Reports,[1] a reviewing service to exhibitors, which furnishes a short account of the story of the film and a statement of its probable box-office value. The accuracy of these stories was validated by comparing them with other writ-
( 246) -ten accounts and by verifying those accounts of motion pictures which the investigators had viewed. We decided to make our study one of the general content of 500 feature pictures produced in 1920, 1925, and 1930. This represents the total output of feature pictures released in these years by the major producing organizations.
Our next problem was to discover the classes into which these pictures might logically fall. We adopted for this purpose what might be termed a common-sense classification; in other words, a classification which is similar to that which lay adults commonly use for the description of motion pictures. Our tentative examination of the stories showed that they grouped themselves into the following classes: crime, sex, love, mystery, war, children, history, travel, comedy, and social propaganda. Subclassifications were drawn up under each of these categories; first, in order to assist the classifier; and second, in order that further data might be gathered concerning the content of the motion picture. These subclassifications were given a number and were checked in the appropriate column of the data sheet. The symbols A and B were used to designate those films where the reviewer felt that there was not only a major theme but also a strong minor theme. No attempt was made to discover the objectivity of these subclassifications.
Does this method of classification yield uniform results when utilized by trained workers? To test this out we took 100 sample reports at regular intervals from each of the three groups of 500 pictures. The reviewers were asked to use the instructions prepared and classify them according to their best judgment. We discovered that in the 1920 movies there was perfect agreement among the three readers in 87 out of 100 pictures when classified as to type; e. g., crime, sex, love, and so on. For the 1925 pictures there was perfect agreement in 86 cases out of 100, and for the 1930 pictures there was perfect agreement in 88 out of 100 pictures. This is a perfect agree-
( 247) -ment of approximately nine cases in every ten. The technique was therefore considered satisfactory for our purpose; namely, to classify motion pictures according to the main types set up by us. The evidence as to the number of pictures of each type is, therefore, indisputable within the limits given.
The analysis just described is valid for presenting the major themes or leading ideas with which motion pictures are concerned. It is not valid for answering many of the critical questions which sociologists and others are asking concerning the content of motion pictures. For this detailed analysis we viewed one hundred and fifteen motion pictures at the theater. The steps followed in this analysis were these :
1. A canvass was first made of the safeguards which are necessary to ensure fidelity of report when observers are used. Whipple's suggestions for such safeguards[2] were carefully heeded. He states [3] that "if the expectant attention is properly directed, however, the efficiency of observation is greatly increased." This precaution was observed in this fashion: First of all, the observers familiarized themselves with the story before they went to the theater. The motion-picture reviews in the daily papers usually gave such an account. Reading the story before reviewing the picture gave the investigators P. frame of reference, a schematized outline which made it possible for them to grasp easily what occurred on the screen. Second, each observer carried a schedule of points on which to secure information. This schedule included the critical areas in which we desired information and had been worked out in coöperation with the observers. Further, three observers were used on 75 of the 115 films reviewed by this schedule.
The schedule was developed in this fashion:
All available literature dealing with favorable and unfavorable criticisms of theatrical motion pictures was read with a view of determining the positive and negative values
( 248) which have been stated for such motion pictures. An analysis schedule was developed based on a classification of these possible values and detriments. The major headings in the final form of this schedule sheet are as follows:
SOCIAL VALUES IN MOTION PICTURES
I.Nature of American Life and Characters
II. Nature of Foreign Life and Characters
III. Motivation of Characters
IV. Emotion Appeals to Audience and Methods of Making Them—The "Kick" of the Movies
V. Crime, Delinquency, and Violence
VI. Relations of Sexes
VII. Military Situations
VIII. Depiction of Underprivileged Peoples
IX. Deportment, Language, Manner and Tone of Voice, Type of Dialogue and Song
Each of these categories was further subdivided. The subdivisions used for Category No. 1 follow:
1. Nature of American Life and Characters
A. Home
B. Education
C. Religion
D. Economies
E. Agriculture
F. Industry and commerce
G. Civic life
H. Recreation
I.Social conventions
J. Clothing conventions
K. Narcotics and stimulants
L. Law enforcement
M. American men
N. American women
O. American youth
P. American children
Each of these subcategories was further divided by a series of points; e. g.:
Industry and Commerce
Pay special attention to the following points:
1. The nature of the portrayal of industrial and commercial activity
2. Goals of characters engaged in industrial activity
3. Methods of distribution of goods
4. Nature of portrayal of owners and workers
5. Nature of the management of industry
( 249)
The reviewer was expected to note descriptive details in the picture which dealt with these points. It is evident that from these data we shall be able to draw inferences concerning critical questions such as these: Do theatrical motion pictures acquaint the viewers with the major problems of industry and commerce? Do they show industry as democratically or autocratically managed? Are workers shown as thoughtful, independent, and self-respecting, or as thoughtless, dependent, and obsequious? Are the problems of the coal, cotton, and wheat industry realistically portrayed or are motion pictures entirely free from problems that beset American industrial civilization?
2. Accuracy of report was further ensured by following a second warning of Whipple's; e. g.: "Whenever any interval of time elapses between the actual carrying out of observation and the recording of it by word or gesture or pen, the accuracy and completeness of the record tends to be reduced by errors of memory."[4] Each observer recorded at the theater the pertinent material which he was seeing on the screen. He occupied a seat near a light and it was possible in this way to make satisfactory notes. These notes were written up either that day or the next. Even with these precautions, minor errors were discovered. This situation was met, in part, by observing a third canon set up by Whipple: "When a number of persons report upon the same matter, those details upon which agreement appears may in general be considered as correct."[5]
An analysis of this type makes possible the answering of many important questions concerning motion-picture content. Its deficiency lies in the fact that it does not make available the total context in which each of the situations occurred. We felt, further, that we needed a number of accounts which would present almost completely the entire range of content in a motion picture in the context of the narrative itself. To that end, we secured from the
( 250) producers dialogue scripts and used them in our analysis of 40 motion pictures. The script contains all the dialogue and enough of the settings and action to give each bit of dialogue its proper chronological order. The observers for these 40 motion pictures were all trained stenographers and the schedules were used as before. What the observers now did was to:
1. Familiarize themselves with the dialogue script before attending the motion picture.
2. Attend the film and take stenographic notes of all materials not included in the dialogue script. This consisted of detailed descriptions of settings, clothing worn, gestures, intonations and facial expressions of characters, approximate age, economic levels, and so on.
3. Immediately write up the picture in the form of a running narrative based upon a combination of the dialogue script and stenographic notes, every change of scene being carefully indicated. These reviews will average approximately 40 double-spaced typewritten pages each.
Of the 40 pictures thus reviewed, 27 were viewed by two or more trained observers, the remaining 13 being viewed by one trained observer who had been the research assistant throughout the entire experiment.
The final results of this investigation as far as methodology are: (1)a reliable technique for the classification of motion pictures according to major theme, (2)a schedule sheet by means of which critical information about motion pictures can be secured by trained observers, and (3)a technique for highly detailed film analysis.
The technique for evaluating motion pictures according to major theme was applied to 500 feature pictures in each of the years 1920, 1925, and 1930. The schedule sheet was applied to 75 motion pictures and their content deter-mined through this method. And finally, a highly detailed narrative account was secured through the application of this schedule sheet to 40 additional motion pictures.
