Shot Length Distributions in the Chaplin Keystones

This week I have another draft of a Cinemetrics paper, this time looking at shot length distributions in Keystone films starring Charles Chaplin and directed by Chaplin, Mack Sennett, Mabel Normand, George Nichols, and Henry Lehrman. You can download the pdf here: Nick Redfern – Shot Length Distributions in the Chaplin Keystones, and the abstract is given below.

Cinemetrics provides an objective method by which the stylistic characteristics of a filmmaker may be identified. This study uses shot length distributions as an element of film style in order to analyse the films by five directors featuring Charles Chaplin for the Keystone Film Company. A total of 17 Keystone films are analysed – six directed by Chaplin himself, along with others directed by Henry Lehrman, George Nichols, Mabel Normand, and Mack Sennett. Shot length data was collected for each film and then combined to create data sets based on the studio style and for each director. The results show that for the distribution of shot lengths in Keystone films starring Chaplin (1) there is no significant difference between films directed Chaplin and the overall Keystone model; (2) there is no significant difference between Chaplin’s films and those of Lehrman, Nichols, and Sennett; (3) there is a significant difference between the films of Normand and the Keystone model but the effect size is small; and (4) there is a significant difference between Normand and the other Keystone filmmakers but the effect size of these differences is again small. This study shows that the distribution of shot lengths can be used to identify how the style of an individual filmmaker relates to a larger group style; and that, in the specific case of the Keystone Film Company, it is the studio style of fast-paced, slapstick comedy that determines the distribution of shot lengths with little variation present in the films of individual filmmakers.

As before, any comments and suggestions are welcome (as is the pointing out of glaring errors).

The raw data was collectde by examining the films frame by frame in my editing software, and can be accessed in a Microsoft Word Document here:

For Microsoft Word 97-2003 (x.doc): Nick Redfern – Shot length distributions in the Chaplin Keystones – data

For Microsfoft Word 2007 (x.docx): Nick Redfern – Shot length distributions in the Chaplin Keystones – data

About Nick Redfern

I am an independent academic with over 15 years experience teaching film in higher education in the UK. I have taught film analysis, film industries, film theories, film history, science fiction at Manchester Metropolitan University, the University of Central Lancashire, and Leeds Trinity University, where I was programme leader for film from 2016 to 2020. My research interests include computational film analysis, horror cinema, sound design, science fiction, film trailers, British cinema, and regional film cultures.

Posted on September 3, 2009, in Charles Chaplin, charlie chaplin, Cinemetrics, Film History, Film Studies, Film Style, Film Theory, George Nichols, Henry Lehrman, Hollywood, Keystone Film Company, Mabel Normand, Mack Sennett, Silent cinema and tagged , , , , , , , , , , , , . Bookmark the permalink. 6 Comments.

  1. A quick footnote to this remarkable study. If we add to the mix what we happen to know about the relationships between Chaplin and such major figures at Keystone as Lehrman or Normand we may still argue in favor of those who claim that Chaplin was, early on, in favor of slowing down the Keystone tempo. This may be established, perhaps, if we do not stick blindly to who is the credited director of this or that film, but take into account a broader “director factor” so to speak. Regardless of who is the credited figure standing behind the camera it was Chaplin who always invented the Chaplin gags, even at Keystone, and if he needed more screen time for them he took it. It makes sense, for instance, to do a comparative research of a similar kind as Redfern did but looking at Keystone films with Chaplin as a star against films without Chaplin acting. Or even, using the Mohsen Nasrin approach, to single out “shots with Chaplin” as a separate category, and then see if respective ASLs will show a significant difference.
    Here is an example. “Making a Living” is Chaplin’s first Keystone film. Here he is not waring the Charlie outfit yet and plays a conventional villain, much in line with what he was hired for. The film was directed by Lehrman in the Lehrman-Keystone fast style. As measured on Cinemetrics by Torey Liepa “Making a Living” has an ASL 4.3 seconds (Lehrman’s pre-Chaplin “Bangville Police” is somewhat slower, 6.9, still withing the fast Keystone range, if I read the number correctly). But: the same year Lehrman directed Chaplin’s second picture, “Kid Auto Races at Venice” in which Chaplin wears his signature costume, mustache and all, and is clearly in control of what is going on in front of the camera, it’s a semi-improvised film, Chaplin reacts to the camera and to passing cars. According to Charles O’Brien “Kid Auto Races” has an ASL of 19.5 seconds. This is quite slow, and I believe Chaplin’s philosophy according to which acting comes before editing (and, for that matter, before everything) may be at work here. Chaplin did quarrel with Lehrman on the set, and he fell out with Normand (whose acting style was more slapsticky than Chaplin’s and whose films, according to Redfern’s study are “marginally quicker than films by other filmmakers”). If the tendency to win more shot time for better acting is indeed what Chaplin (with his stage experience and training) was striving towards we may predict that Redferns’s forthcoming study of Chaplin at Essenay or Mutual (where Chaplin was more in control) will yield a longer overall ASL. At least so it seems looking at some of Cinemetrics submissions.

  2. The summary of the statistical measures for these Keystone films show values all over the place. The Lehrman films have medians of 2.6 seconds, 4.9 and 6.5 seconds, the Mack Sennett films George Nichols fillms have medians of 7.3 and 3,3 seconds, the Mabel Normand films have 2.5, 2.8 and 3.4 seconds, which is a little better, but still a difference of 50% between fastest and slowest, and only the Chaplins have any consistency, being nearly all in the 3.4 – 3.9 range, with one exception (A Busy Day) at 4.9 seconds. The interquartile range shows the same kind of variability. On this stylistic measure, it would appear that only Chaplin had a style at Keystone, and there wasn’t a general house style — in this respect.
    From experience. I would say the same kind of variability is general with one-reelers, which is why I have ignored them in my statistical style analysis. Even two-reelers are a bit dubious. The point is that in 10 minute long early films, the lack of variety in the script can have an overhwelming influence on the shot lengths, which is ironed out in feature films by the varied dramatic content down the length of the film.
    By the way, I could have talked in terms of ASLs and Standard deviations above, because the two ways of looking at distributions are fairly strongly connected in the case of shot lengths.

  3. Some more points. Strictly speaking, the distributions you are working with ARE censored, since you are leaving out the intertitles. This tactic could do with a bit more justification, as it is common in working with ASLs for researchers (not just me) to count intertitles as shots.
    Also, I think you should indicate which DVDs you are using to get the data, as many of these films are preserved in the first place with different copies.
    A very minor point is that you are assuming a binomial distribution for the Confidence Intervals, and these distributions are pretty certainly NOT binomial.
    By the way, have you looked at plots of the actual distributions you have got for these films?
    Delahal??? et al. or whoever they were you were recommending as a tutor text to the Cinemetrics gang says you should.
    And maybe you should do a Mann-Whitney for each director against each other, and not just for each against the whole studio.

  4. Dear Mr. Redfern,
    I was wondering, why you used the Mann-Whitney U test — which is for data on an ordinal scale — and not another nonparametric test for data on an interval scale.
    Thank you. Best,
    Veronika Koch

    • Parametric tests such as the t-test require that data is at least interval, but the Mann-Whitney U test can be applied when the data is at least ordinal and can also be used when the data is on an interval scale. The U test is not restricted to situations when the data are ordinal, but you do need to be able to say which of two values is greater than the other. (It could not be used for nominal data).

      The U test also has advantages of robustness, especially when dealing with non-normality and outliers, which is very useful for the data you get in dealing with film style.

      Many textbooks recommend the U test as the appropriate alternative to parameteric tests for interval data, and the alternatives aren’t up to much: the median test is limited to inferences about location only, and you can use the U test to do the same thing only better anyway.

      Other rank-based tests could be used, and there’s an interesting paper on this subject here:

  1. Pingback: Estimating shot length distributions « Research into film

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: