7.6 Comments on Bertin (with hindsight)
Bertin’s “Semiology of Graphics” contains many important ideas and was an attempt to develop a theory of graphics that could be built on and extended. Reviewing the dataset he considers in most detail does not reflect this. There are too many unsatisfactory graphics and there are few conclusions drawn from the data. The emphasis is more on presenting the available data in a broad range of different graphics than on deriving information from the data using graphics. Given the amount of work that was necessary to produce the graphics—some are beautifully drawn—this is disappointing. As Bertin himself says on p. 419 `Graphics has two objectives: 1. Data processing to understand the data and derive information from them. 2. Communication of this information.’
Bertin is worthy of serious study and applying his ideas to new datasets may be more effective than studying his examples. There are no reproductions of Bertin’s graphics here, hopefully encouraging readers to look at “Semiology of Graphics” for themselves.
Answers Each of the three sectors employed between 5 and 7 million. There were relatively more agricultural workers in the West of France. Numbers working in Industry and Commerce were positively related.
Further questions How have the numbers changed since 1954? Are there other sectors which have become important?