This article in the series covering the Eighteenth Century London Concerts dataset looks at composers. As previously discussed, composers can be identified as the names preceding a “Genre” code in the list of entries in the dataset’s “Programme” field. In most cases they can also be associated with the genre of the work in question (and sometimes the precise work can be identified, although this information is quite patchy).

The most common composer was Handel, who was mentioned in 1,236 concerts – over 30% of all concerts. Next were Haydn (739), J C Bach (332), Ignaz Pleyel (219) and Thomas Arne (187). Although this looks like an excessively Handel-heavy programme, the distribution of concerts per composer actually follows a power law with exponent equal to about 2, as predicted by the Law of Evenly Distributed Log Success, as shown on the following chart:¹

The chart shows the number of composers (vertical axis) mentioned in at least the number of concerts on the horizontal axis. Both axes have logarithmic scales, and the roughly straight line indicates an approximate power law. The dotted blue line shows the slope corresponding to an exponent of 1.9, for comparison. Thus all 441 composers were mentioned for at least one concert (the point at top left), 275 appeared at least twice, and so on down to Handel in the bottom right.

Handel lived in London and died in 1759. He was the top composer until the 1780s, when he was overtaken by Josef Haydn. The following chart shows the top ten composers by decade, where the text size reflects the number of concerts relative to the most popular.²

There was a lot of change over the period. Handel was the only composer making the top ten in each decade, and only two others (Thomas Arne and J C Bach) managed three or more appearances. Handel was very dominant in the 1750s-60s, as evidenced by the small font of the other names on the list, but by the end of the period there was more diversity, with the big names being relatively less dominant.

We should remember that the dataset is largely drawn from advertisements in the press. There is inevitably a tendency in such advertisements to focus on the big names. Many concerts list, for example, a work by Handel, plus a number of other pieces for which no composers are mentioned (performers are more likely to be credited). Among the concerts mentioning a single composer, Handel is that composer in 731 cases. Second is Thomas Arne, with just 65 single-composer concerts. Such a large differential between first and second place is surely evidence that Handel’s name was often the only one felt to be worth mentioning in an advertisement.

We can also look at pairs of composers, to see the combinations that were most or least likely to appear on the same programme. I have done this analysis, and considered the network graph of names that tend to appear together, but it is not particularly enlightening, so I won’t say more about it here.³

As mentioned above, in most cases, composers can be associated with at least the genre of the work performed. The following chart shows the links between the top few composers and the most common genres.⁴

This kind of chart is sometimes called an “alluvial plot”.⁵ On the left we have composers, occupying a space proportional to their share of works, and on the right we have genre categories on the same basis. The coloured bands indicate the number of works of each composer in each genre. So Handel was particularly prominent for his oratorios, plus a few overtures, songs and concertos. Haydn, on the other hand, was mainly represented by overtures and symphonies. Songs, apart from those by Handel, were largely accounted for by a handful of lesser-known composers, such as Sacchini, Sarti and Cimarosa. Pleyel seems to have cornered much of the quartet market, as Webbe has with glees.

There is more than can be done on composers, along the lines of some of the other analyses on this site. It is, however, somewhat hampered by the gaps in the data, making it difficult to draw firm conclusions.

1. For more detail of the Law of Evenly Distributed Log Success, see my article in Empirical Musicology Review.
2. The names here are the standard abbreviations used in the dataset. Most are self-explanatory, except that “BACH” is Johann Christian (London resident and organiser, with Carl Friedrich Abel, of a popular concert series), rather than his (today) better-known father, who was little-known here until the nineteenth century.
3. I used a similar approach to that used in this article looking at teachers and students.
4. Note that we are now considering the individual works, rather than simply whether a composer is mentioned for a concert.
5. It was produced using the ggalluvial package in R.