Paradise lost nor regained: social composition of theatre audiences in the long nineteenth century.

This essay criticises on empirical grounds the assumptions about the social composition of theatre audiences from the late eighteenth century to the Great War as they were made in theatre historiography from the late nineteenth century till well into the twentieth century (even till the present day). We focus on Dutch historiography, since the empirical evidence presented comes from the Rotterdam theatre archives. The grand narrative it criticizes, however, was international. We will show that the traditional assumptions about the social composition of theatre audiences in the period 1773-1914 are debatable, particularly with respect to the elite.

Two essays on the average rank occupation rates as related to the supply in the Rotterdam Grand Theatre, 1802-1916, proved, by way of time series analyses, that traditional assumptions in theatre historiography with respect to audience preferences were untenable, at least for the theatre in Rotterdam over the period 1802-1916. Little evidence was found for the supposed decline of the stage due to melodrama, nor for a supposed recovery of the stage after 1870. (1) These essays stressed that time series analyses of ticket sales, though much more reliable than casual remarks in reviews or gross lists of receipts, still dealt with aggregated anonymous spectators, so that the social composition of the spectators was still unspecified. Different outcomes of supply-occupation rates per rank, however, invited hypotheses with respect to differences in social composition per rank. (2)

The present essay deals with the social composition of non-anonymous patrons of drama and opera. It offers an analysis of the subscription lists for season tickets and coupon booklets preserved in the Rotterdam Grand Theatre in the period 1773-1912. We research whether the social composition per rank indeed differed as was supposed in our analysis of ticket sales. This research, however, will be set in a larger framework, to wit, the tenability of the traditional theatre-historical assumptions of audience composition in the period mentioned.

The order of the argument is as follows: Section 1 discusses the traditional theatre-historical view of audience composition in the long nineteenth century, and specifies its often implicit assumptions and hypotheses. It also presents the data we used to test the given knowledge of theatre historiography. Section 2 presents the results of the analyses of the subscription lists. Section 3 concludes that the theatre in many ways is a prison of longue duree in the sense of Braudel and Vovelle. (3)

1. Introducing the Case

This section overviews the traditional notions of audience composition in the long nineteenth century. It specifies the markers involved and the hypotheses assumed, and discusses the data with which the traditional view will be tested.

The Traditional View of Audience Composition in the Nineteenth Century

Traditional theatre history assumes that theatre audiences were 'elite' in the late eighteenth century, but that the rise of melodrama indicated a change in the social background of the audience changed for the worse. In the early nineteenth century, the lower middle class and the working class took over the theatre, even in the best ranks. The elite fled to opera. Only in the late nineteenth century could it be persuaded to return to drama. About 1870 a movement set in for the reformation of theatrical amusements. (4) As a result, the social composition of the audience changed for the better. 1870 was the founding-year of the Dutch Stage League, which, under the leadership of the banker-poet Schimmel, aimed at elevating actors, acting, and reviewing. According to De Leeuwe the literate and rich bourgeoisie organised itself to reconquer the theatre for their own class. (5) The fact that Queen Sophie visited The Danicheffs in 1876 was the ultimate proof that the balance had changed in favour of the 'civilized classes'. This is a narrative of Paradise Lost and Regained, which focuses on the city elites and the working classes. The leading Dutch theatre historians of the middle twentieth century, Hunningher, Albach, Koster, and De Leeuwe, adhered to this view.

The dominant narrative of Dutch theatre historiography meshes with the European historiographical tradition. Davis and Emeljanow sketched the British traditional narrative, supported amongst others by the authority of Allardyce Nicoll and George Rowell, as "a progression towards the restoration of literary drama, improved standards of production, and greater social respectability, both on and off stage." (6) Here, too, the repertoire of melodrama was related to the social composition of the audience, and here, too, the power of the gallery audience caused the decline of the drama (p. 97). They quote Nicoll:

  The stage in the early part of the century was largely a "popular"
  affair, and for the most part bourgeois opinion regarded its delights
  with cringing disapproval. Typical audiences were composed mainly of
  lower-class citizens with a sprinkling of representatives from the
  gayer or more libertine section of the aristocracy. The staid middle
  class and the respectable, dignified nobility tended to look upon the
  stage as a thing not to be supported in an active manner. (7)

Rowell is quoted for his conviction that in the early nineteenth century "butchers and barbers had driven (...) fashionable clients from the pit to the boxes, or in many cases out of the playhouse altogether". Nicoll and Rowell also saw things improve in the later nineteenth century, because of the good taste of some actor-managers (Macready, Phelps, the Bancrofts, Irving, Gilbert and Sullivan), which brought back a better-mannered audience. Queen Victoria interested herself in this process of refashioning manners in theatre audiences and repertoire. (8) What would have become of the theatre without queens?

Of France and Germany comparable stories are told. With respect to the latter (divided) country, the (probably apocryphal) instance of Goethe giving up his position as intendant in Weimar because of the staging of Pixerecourt's Le Chien de Montargis (in which a dog played the lead), has been given symbolic value. (9)

Somewhat surprisingly, this traditional narrative was recently defended by the leading American theatre historian Bruce A. McConachie in Melodramatic Formations. (10) This book focusses on the social composition of the audience and is presented as a long-term analysis of theatre history ("Employing the concept of theatrical formation [...] requires the historian to dive into the apparent chaos of theatrical events and to emerge with regularities of production, genre, and audience over a significant stretch of time"). (11) McConachie envisions a predominantly 'elite' audience in the stock companies in the Northeast of the USA in the first decennia of the nineteenth century, who abandoned the theatres for opera in the 1830s because of the star system. This system brought the working classes into the houses. About 1850, however, new forms of theatre operation, linked with liberal capitalism, resulted in spectacular theatre for the well-to-do bourgeois classes. In essence, this is the traditional story told for Europe. This relatively recent re-invention of tradition, however, is unconvincing in its application of Tony Bennett's deterministic aesthetic-reception theory, in its use of undetermined terms to define social class differences, and in its anecdotal, unrepresentative dealing with the repertoire. (12) Theoretical literary models are tricky ways to empirically establish the composition of a theatre audience and the audience-performance relation. Audience research needs data on spectators. (13)

Assumptions and Hypotheses Implied in the Traditional Narrative

Let us begin with specifying the often implicit assumptions of theatre historians with respect to the social characteristics of the audience during the long nineteenth century. These provide clues of what to expect and give shape to the analysis of the subscription lists. Occupation and wealth are highly relevant. Hunningher conjectured that an 'elite' patronage in the late eighteenth century changed from about 1805-1870, at least for drama, to a patronage of the lower middle classes and the uneducated labourers. After 1870 the 'elite' took possession of the theatre once more. Albach, Koster, and De Leeuwe largely followed this view. Thus, we must expect that the dominant audience in the late eighteenth century had occupations and wealth above the middle class job categories, which brings us to the aristocracy of regents, and to merchants, bankers, and academic occupations. For the social characteristics of subscribers in the 'melodrama years' (ca. 1805-1870) we must expect a considerable drop in occupation and wealth categories for Dutch-spoken drama performances (artisans, small shopkeepers, working class persons like dock-hands, and so on). Occupation and wealth categories only rose to their former glory after 1870. With respect to occupation, the first half of the nineteenth century must display a sharp division between opera and drama as well. This also implies that subscribers for season tickets and coupons in the period ca. 1805-1870 must have been significantly poorer than those before and after, at least for drama. Job category and wealth, however, are not the only variables to establish an elite or non-elite audience. Most likely social and political functions will relate to occupation and wealth, since voting rights (relevant since 1795) depended on wealth (census). Thus, we must expect that subscribers in the period 1805-1870 held considerably fewer functions than drama audiences before and after these dates, whereas all drama subscription cohorts differed sharply from opera cohorts. In the years in which the elite was supposed to have ruled the theatre, we must also expect a good deal of organized political activity since only voters partook in election clubs. (14) Both Hunningher and De Leeuwe referred to differences in education (Hunningher: "completely uneducated labourers"; De Leeuwe: "petty bourgeois with little comprehension"). (15) This leads us to expect differences in informal education shown in memberships of cultural societies (sociability), and formal school education.

There are also relevant variables which the traditional narrative strangely neglects. The first of them is religion as a social-cultural variable. This variable is relevant in view of the wide-spread theatrical prejudice in the Christian tradition, especially Calvinism, the official religion in the Netherlands. We must expect religion to have influenced the composition of theatre audiences, the more orthodox denominations being under-represented. Also, the older literature does not consider age. McConachie envisioned 'waning patriarchs' at the melodramas in the 1820s and 1830s, which implies an elderly group of frustrated upper bourgeois. Of course, attention must also be paid to the presence of women in the audience. Many theatre historians have stressed that few women attended drama performances in the late eighteenth and early nineteenth century. A final aspect, hardly mentioned in theatre history literature, but featuring in the sociology of culture is social mobility. This aspect is implied in the dominant narrative, since it assumes that elites were hereditary theatre-goers. The late eighteenth-century 'elite' drama patrons, cannot have been the same 'elite' persons, who flocked to opera in the 1840s. The assumption is also one of variance in social background: melodrama audiences supposedly were predominantly lower middle or working class persons, who conquered the theatre due to upward mobility in the Revolutionary era. We must then expect a similar, or an even lower status for their parents. That leaves us with the following markers: occupation, wealth, social and political functions, sociability, education, political allegiance, age, gender, religion, and a genealogical component that must give us insight in social mobility.

With respect to the larger typification of the audience as elite we must pay respect to the results of Pierre Bourdieu's empirical research into participation in culture as related to (many) other social variables. (16) Bourdieu constructed three types of taste (high brow, middle brow, low brow) and also two rival social formations, those who mint social prestige from cultural capital and those who mint it from economic capital. This is to say, a high educated, often state-employed, well-paid class sets itself off, by way of a predilection for avant-garde art forms, from a class of money-makers, not necessarily uneducated, whose taste tends generally to more traditional art forms. In terms of mentality, the carriers of cultural capital stress a cognitive, intellectual approach to 'art', whereas for the carriers of economic capital 'art' is a form of amusement which works on the emotions (including basic pleasure, such as laughter). With that comes an aesthetic predilection for abstract art versus a liking of beautiful things well imitated.

New Empirical Data for Audience Composition

The Rotterdam Theatre Archives contain a number of lists of subscribers for season tickets and coupons (booklets with at the minimum five or six tickets). The lists are specified in appendix 1. For sake of space and clarity, we will utilize here the following aggregated cohorts:

                                 No. of
                                 Persons

 1. Dutch-spoken Drama, 1773-92   157
 2. French Opera, 1789-92         120
 3. Dutch-spoken Drama, 1824-30   253
 4. French Opera, 1817-44         276
 5. Dutch-spoken Drama, 1867-68   115
 6. 'French Opera', 1852-59       134
 7. Dutch-spoken Drama, 1890-95   582
 8. German Opera, 1860-82        1150
 9. Dutch-spoken Drama, 1902-12  1476
10. Maifestspiele, 1909-11        677

In terms of the traditional narrative, these cohorts cover four periods: (a) The later years of the ancien regime, when the comedie larmoyante split, and the opera comique pleased a presumably elite audience (cohorts 1-2). (b) The 'melodrama years', first half of the nineteenth century, during which the elite is believed to have been absent from drama performances, so that a huge social cleft existed between drama and opera subscribers (cohorts 3-4). (c) A transition period, in the 1860s, when slowly the drama-repertoire changed to realism, and the Royal French Opera was replaced by the Rotterdam German Opera, introducing Wagner. We included here the first shareholders of the 1851-1887 Grand Theatre as representing the French Opera, 1852-1859. (17) Especially German opera was considered to attract the elite, but drama was still supposed to cater to a middle class audience (cohorts 5, 6, 8). (d) The period of the recovery of drama, when acting improved and the repertoire made the better sort of people is believed to have returned to the theatre (cohorts 7, 9, 10). For opera after 1900, we have only data for part of the Maifestspiele. Originally during the Maifestspiele the Ring des Nibelungen was twice performed, but the last two festivals, in 1909 and 1911, had a mixed repertoire. The performances were given by German troupes. An assessment of the data is given in appendix 2. Except for the first two seasons of the Rotterdam German Opera, shareholders never got free entrance.

The subscribers have been studied, leading to a database now containing some 14,000 person files. We sought data on name, sex, noble title, address, date and place of birth, date of marriage, date of death, the place from which they migrated or to which they departed, occupation, wealth, religion, social and political functions, sociability, theatre seasons, subscription for or against the Rotterdam fair, education, political allegiance. Also we collected data for wives (or in a minority of cases, of husbands), father, mother, grandfather and grandmother of the shareholder or subscriber, provided they were born in Rotterdam or the towns of which Rotterdam now stores the population registers. Appendix 2 gives information on the sources for the data and the construction of the wealth classes.

Season tickets and coupons were not available for the cheep ranks. In all three consecutive Grand Theatres subscription for the upper tier was impossible. Also, considerably fewer subscriptions were made for ranks lower than the first (stalls, boxes, balconies, and baignoires). Since the issue with traditional theatre historiography is particularly the question of the participation of the upper classes, the skewness in the data, due to the distribution over ranks and income is not a hindrance for the historical part of the analysis, but a major obstacle for sophisticated statistical tests: only non-parametric tests can be applied to this database.

2. Analyses of the variables

This section presents the analyses of the social variables of the subscribers. We will first address the question whether melodrama audiences in the first half of the nineteenth century indeed were predominantly lower middle class or even working class persons and whether the elite indeed massively preferred opera over drama. Then the social characteristics of the subscribers over the whole period are surveyed, distinguishing between the social variables explicitly addressed in the traditional historiography and the social variables implicit in it, especially religion, the issue of women in the subscribing audience, and some brief remarks on age. The analysis ends with specifying differentiations per rank, and the background of the subscribers. Appendix 3 gives the basic cross-tables of the variables over the period 1773-1912. We will refer to these cross-tables in the following analyses. Tables are thematically numbered after the order of the basic tables in appendix 3. So, all tables having to do with occupation and wealth are numbered 1a, 1b, etc., and all tables having to do with religion are numbered 3a, 3b, etc.

A New Audience at Melodrama?

With respect to occupations, the traditional historiography assumed a (lower) middle class and working class audience. From table 1a (appendix 3) it is clear that most of the melodrama subscribers (cohort 1824-1830) fall in the category 'trade' (62 per cent), middle class occupations and professionals limped far behind (15 and 13 per cent respectively). This distribution refutes the assumption of the traditional view. With respect to wealth, the melodrama cohort indeed stands out, not only with respect to other drama cohorts, but also to the parallel opera cohort, the French Opera (1817-1844). A distribution which decreases by about 50 per cent per wealth class (from the third class onwards) is only weakly repeated in the eighteenth-century drama cohort (1773-1792), but there we found data on rateable value for only 49 out of 157 subscribers, and those were at best twelve years after the subscription (1803 register). Also, the fact that we deal with rateable values for the melodrama cohort hinders a precise comparison with other drama cohorts, since these are based on different wealth data. To give more context to these figures, we also analysed the so-called taxable income based on real estate, that is, the sum rateable value of a person's realty. It turned out that 83 of 132 melodrama subscribers who were taxed for real estate, possessed realty over fl. 1000.-rateable value (63 per cent). Thus, while 32 persons in the melodrama cohort lived in a house with a rateable value over fl. 1000.-, fifty-one subscribers possessed realty larger than that sum (twenty-two persons, 17 per cent, held realty with a sum rateable value over fl. 3000.-, of whom six very rich peaked over fl. 5000.-). Though we would not sustain the assumptions of the traditional narrative on the basis of the rateable values (the 103 values between fl. 200.- and 600.- can only be regarded as a solid wealth position), the additional data on taxable income clearly falsify the assumptions with respect to the wealth of melodrama subscribers.

As explained in Appendix 2, the subscribers for season tickets in 1824-1830 formed an average of some 50 per cent of the pit audience per season. Against a downgrading of not-subscribing audiences in the pit during the melodrama years we may call attention to class sensitivity. We regard it unlikely that prestigious families like the Reepmakers or the Van Rijckevorsels would be prepared to rub shoulders with butchers' boys or hawkers in the pit. It is even doubtful that these latter would have been the regular audience of the gallery, for a ticket there cost 60 cents. That was not the sort of money for many an artisan, let alone workmen in the docks or construction to spend on a play. Likely it was even too much for many a small retailer, independent artisan, clerk, or employee to spend regularly. (18) Possibly these groups caused the high peaks in occupation rates in Fair-time and benefit performances, or the yearly New-year's performance of Vondel's Gijsbrecht van Aemstel. There is no reason to assume radical changes in the composition of this audience in the period ca. 1770-1860 which are not related with changes in the social structure according to occupation in itself (such as e.g. the increase of employees after 1815).

We speculated, because of the census voting right, that power, expressed in political and social functions, would go along with wealth. Table 4a (appendix 3) confirmed this. The number of functions, to be sure, is a gross list of the functions a person held throughout his career. We calculated the average number of functions per function-holding subscriber, and also the average number of functions in the cohort. This latter figure we defined as the functional weight of a cohort. With an average of three functions per function holder, and a functional weight of 1.2 the assumption that melodrama patrons had a low status is falsified. Function-holding in the melodrama cohort differed not much from that in the late ancien regime, where the traditional narrative assumed an educated elite, and it was larger than that in the drama cohorts around 1900.

Traditional theatre historians stressed the low level of education of melodrama audiences. We consider education in the form of sociability and as formal school training. Table 5a (appendix 3) gives the distribution of memberships of societies. Here, too, it concerns a gross list of all collected memberships held in life. Again, the melodrama cohort does better than the late eighteenth-century cohort, but worse than the later drama cohorts (sociable subscribers had an average membership of 2.4 per person, and the sociable weight of the cohort is 1.0). Figures on sociability, however, depend more strongly on available archives than political and social functions, and the archival situation at the end of the nineteenth century is considerably better than in the melodrama years or the late eighteenth century. (19)

Data on formal education are very hard to find before the 1860s. The only data available concern higher education (university and the medical school) and the Latin school. (20) Against the view of a completely uneducated audience or one with little comprehension argue the fourteen melodrama subscribers with higher education and the one grammar-school pupil. The drama cohort of the ancien regime counted only two university-trained subscribers. The assumption of a generally illiterate melodrama audience is, to our view, to be rejected.

Due to the political system and poor records, little is known about the political stand of the melodrama subscribers. We detected two conservative subscribers, one conservative liberal and three liberals. Melodrama subscribers at least counted twelve former democrat Patriots, but no former Orangists. (21) The majority of the merchant elite families subscribing in the melodrama cohort had Patriot backgrounds, either of themselves, or via their fathers. We counted 26 Patriots and two Orangists in the father generation (150 fathers), and ten Patriots against three Orangists in the grandfather generation (65 grandfathers).

In comparison, the larger French Opera cohort (1817-1844), counted only six former Patriots and two Orangists. This cohort, of course, counts more persons, who were too young to participate in the party struggles of the 1780s. In the father generation indeed more data were found: 37 fathers had a Patriot background, against ten Orangists (178 fathers). In the grandfather generation these figures are fourteen Patriots and seven Orangists (99 grandfathers). Here, too, the tendency is towards democratic Patriotism, but Orangism manifested itself more strongly at the Opera than with melodrama. Perhaps the managers of the drama company were right to stress that subscribing for drama was an act of patriotism in the sense of 'nationalism'.

This comparison has taken us to the second issue, to wit the supposed cleft between drama and opera-goers in the first half of the nineteenth century. Table 1a (appendix 3) suggests just minor shifts in the distribution of occupations. Remarkable, but still relatively small, is the difference between middle class drama subscribers and opera subscribers (10 per cent less middle class at the opera), which equals the difference of academics in both forms of theatre (10 per cent less academics at melodrama). Wealth stated in rateable values, is far more equally spread in the French Opera cohort (1817-1844) than in the melodrama cohort. Of the opera-goers, 121 persons (77 per cent) owned realty yielding more than fl. 1000.-rateable value, and of those 121 persons, 32 persons (21.5 per cent) owned realty yielding a rateable value over fl. 3000.-, twelve persons peaking over fl. 5000.-. This is better than the melodrama cohort, but it does not at all sustain the conjecture that the city elite massively turned its back on melodrama and went to the opera. Of the melodrama subscribers 69 also subscribed for the French Opera (1817-1844), which is 27 per cent of the melodrama cohort, contributing 25 per cent of the Opera cohort. Some of the melodrama subscribers still subscribed for the German Opera in the 1860s.

The better wealth position of the opera cohort mirrors itself in a larger participation in political and social functions (table 4a, appendix 3). The functional weight of the French Opera, 1817-1844 is 2.6, thus more than twice that of the melodrama cohort (average number of functions per function holder is 4.3). The political stand of the opera-subscribers was also somewhat easier to trace than those for melodrama: we detected seventeen liberals, nine conservatives, and two conservative liberals, which is ten per cent of the subscribers. For melodrama the political stand could be traced for only two per cent of the subscribers. This difference is due to the different measures in which subscribers were active in the election clubs formed after 1848, and it must be kept in mind that the melodrama cohort was about 20 years removed from that effect of the revolution. Yet, also considering the relatively fewer functions, it is reasonable to say that more opera-subscribers than melodrama subscribers became part of the post-1848 political elite.

As might be expected, sociability and formal education in the opera scored better than in the melodrama cohort (tables 5a and 6a, appendix 3): the sociability weight of the melodrama was 1.0 against 2.1 for the opera. Considerably more subscribers for opera partook in cultural and social societies (71 against 41 per cent). Yet, the melodrama subscribers were members of the same elite societies, both cultural and social, as those for the opera. Only four participated in typical middle class literary societies. The French opera-subscribers scored better on formal education, too, and counted 41 persons with higher education and two with grammar school. This is, of course, mainly due to the larger portion of academics participating in the opera cohort.

All in all both tenacious assumptions of traditional theatre historiography cannot be maintained: in the first half of the nineteenth century, melodrama did not just draw lower middle class and working class patrons. At least half of the pit audience consisted of solid citizens and the merchant elite. This merchant elite also visited the opera, and, thus, falsifies the conjecture that the elite massively left drama for opera.

An Overview of the Explicit Social Variables of the Subscribers, 1773-1912

This section surveys for the whole period, 1773-1912, the variables which explicitly played a role in the traditional narrative of audience composition, job category, wealth, functions, sociability, education, and political allegiance. We argue that not only the conjectures about melodrama patrons and the difference of patronage at the opera and drama are untenable, but that in fact so is the narrative of Paradise lost (in the French Revolution), and regained (after the Paris Commune), cannot be maintained either.

Table 1a (appendix 3) presents the aggregated data on job categories and wealth classes. It appears that the distribution of wealth and job categories per cohort remains stable. 'Trade' dominated, and all other categories limped far behind. For nineteenth-century Dutch-spoken drama in Rotterdam, 'middle class' occupations normally take the second place, followed by the 'professionals', who, however, take over the second place after 1900. This has likely to do with the trade organizations of office clerks, who often bargained for special prices. For opera, this picture is less rigid. The overall skewness tending to high income is noticeable. Not only is trade by far the largest category, but in most of the cohorts it has a lion's share in wealth. (22) The difference between opera and drama remains relative. (23)

The data for both the late eighteenth-century cohorts are a bit unsatisfactory, because for drama (1773-1792) the number of missing items is large (108 of 157). For this cohort the Marriage and Burial Tax (MTB) gives a better result, but is as a measure not very satisfying. (24) We must not suppose a slow decrease of the category 'trade' over time, as table 1a might suggest. The changes are relative to subscription opportunities: from the middle of the century the coupon booklet brought in many more subscribers, and by the end of the century more ranks were open for subscription. The basis for subscription, thus, broadened.

The relation between occupation and wealth categories was, for those cohorts for which it made sense, strong. (25) Generally the cohorts also show the same tendencies in under- and over-representation of job categories by wealth class. For 'trade' and 'academics' there was a strong under-representation in the lower wealth classes and a strong over-representation in at least the fifth wealth class. For 'middle class' and 'professional' occupations this under- and over-representation worked the other way round.

The last cohort, Maifestspiele (1909-1911) is an exception in many ways, but peculiarly relevant for the issue of overall audience characteristics. It concerned a two-week festival, for which most of the tickets were sold by subscription. Even though we have no information on rank, the total amount of subscriptions (677 persons, to which have to be added some 160 persons from other cities, who have not been included in the database), related to the house capacity (ca. 1250), suggests that most of the spectators are identified (if every person bought tickets for two there would be some 1675 persons, which the house could not contain). Nine subscriptions from working-class persons indicate that these classes were hardly interested in this form of theatre. With respect to job categories and wealth, Paradise was and remained in the theatre stalls.

Money talks: this proverb is relevant for our subscribers. The over-representation of rich merchants, subscribing for first-rank tickets, indeed strongly influenced the other cultural variables. Here, too, the findings for melodrama and opera in the first half of the nineteenth century are reproduced over the whole period, 1773-1912.

The distribution of social and political functions is given in table 4a (appendix 3). The rate of persons holding functions diminished at the end of the century. The coupon booklets brought in more patrons outside the range of the political and social elite, thus we see a drop in percentages from the 1860s. The difference in functional weight between opera and drama, noticed above with respect to melodrama and opera, also proves structural. More opera-subscribers on average held more functions than drama-subscribers. The drama-cohort 1867-1868 stands out, but all its members, except a handful, also subscribed for the German Opera.

Functions and function-holding were biassed for wealth class. Table 4b gives the average number of functions per wealth class per nineteenth-century cohort. (26) All cohorts, then, contain a core of the rich and mighty, a core that proportionally became less 'powerful' in the course of the century, due also to broadening the basis of subscription.

Function-holding tended to cumulate, which shows the nature of merchant elite influence on all fields. Many of the holders of political functions had their hands in economic and financial organisations as well. Cumulation was also career-bound. A successful political career might well involve some four or five functions: elector, councillor, alderman, burgomaster, member of the provincial estates, member of one of the Chambers of Parliament that was the full cursus honorum. If one had succeeded in becoming a councillor, functions of the fixed city committees could follow, as could social and charitable, and military functions (these were often a prelude to a political career, too). Educational functions were to a less extent also tied to political or economic functions. In the course of the nineteenth century financial functions boomed, since a specific 'financial' elite focussed on banking and particularly in insurance. In order to spread the risks, the elite took part in a growing number of insurance societies, particularly for 'sea and fire risks', as they were called.

Most of the functions in the eighteenth-century cohorts were held after the revolution of 1795, thus, later than the early subscription lists. Many of the subscribers were dissenters, who were barred from holding public functions before 1795 (see also below). Their number of functions is suggestive of the potential prestige of the non-aristocratic merchant class. (27) Contrary to expectation, all cohorts had subscribers with high political status.

The late eighteenth and nineteenth century are known as the age of sociability. Table 5a (appendix 3) witnesses this. Measuring sociability needs archives, and though some important membership lists are missing for the early and middle nineteenth century, the proportion of members in those years is still the largest (76 to 88 per cent). Both the edges of the period do less well. For the eighteenth century we lack many data for the smaller and more informal clubs (private reading clubs), as well as those of the important literary society Verscheidenheid en Overeenstemming [Variety and Harmony]. From about 1880 we see the decline of enlightenment sociability and the rise of a more market-based organization of cultural life. Sociability, too, was biassed for wealth, as is shown in table 5b. Table 6a (appendix 3) presents data on education and political allegiance. As already noticed, data on education in the period before 1863 are hard to find. Particularly secondary education was opaque. Hence, we only tabulate here data for the cohorts after 1860. Although the outcome may look disappointing, in fact the proportion of subscribers that followed higher, gymnasium or burgher-school education was well above that of the city population. (28) Later twentieth-century surveys of participation in cultural events often take the MULO as the lowest level of extended education, leading normally the conclusion that theatre and concert audiences were generally educated above the average. (29) The data suggest that this is also true for the nineteenth century. Initially, most pupils did not finish the whole course of these new forms of education. In all cohorts this mounted to over 50 per cent of the pupils. Still, this does not mean that they were uneducated. Taken the number of high-educated subscribers in the first half of the century, and the figures for secondary and higher education after 1863, the assumption of an audience 'with little comprehension' cannot be sustained without new evidence. For now, we reject such a view.

Education also played its role in the discourse of the re-conquest of the stage after 1870. Rossing, the first historian of this movement, particularly stressed that the burgher-schools were the backbone of the re-education of the theatre audience. (30) Indeed, the burgher-schools focussed on modern rather than the classical languages (French, German, and English being obligatory), and often educated their pupils in relatively modern literature. Also, the practise of school drama returned. We agree, even on the basis of this slim evidence, that the subscribers in the cohorts around 1900 were well above-average educated, but with respect to the small impact of the burgher-schools both in the cohorts and in the population at large, we think it overdone to say that the burgher-schools were a decisive factor. To decide that, far more data are needed documenting the share of cultural education in boarding schools, the MULO, cultural societies, or at home.

We also hypothesised that, due to the census voting system, the wealthy theatre goers would show most active interest in politics. They did, but with a strong bias, first for democratic Patriotism, and later for liberalism (table 6a, appendix 3). The data for political allegiance were strongly biassed due to the fact that only liberal election clubs left membership lists. These cover the period 1869-1882. Officials and candidates of conservative and Christian-Democrat election clubs were listed in the address books, election-advertisements in journals, and newspaper reports on political meetings, but their numbers are a fraction of the liberal counts. Therefore, it is no surprise, hence, that liberal allegiance dominates the later nineteenth-century cohorts. Yet, this dominance is not fully due to archival bias. The conservatives, the former aristocratic regents, always formed a small fraction that had its Indian summer in neo-absolutism (1814-1848). About the same counts for conservative liberals, successors of the 'moderates' of the Batavian Republic. Both, moreover, often merged with what were later to become the religion-based parties. There, a strong resistance against theatre-going was traditional. This resistance tended to increase in the later decennia of the nineteenth century, due to the intensification of cultural differences known as the 'pillarisation-process'. Only a very few conservatives and Christian socialists who ran for elections can be found among the subscribers, even if they fitted the profile of theatre-goers in status, occupation and wealth.

With respect to political allegiance, we envision a continuity of what was already noticeable at the beginning of institutionalised theatre culture. The founders of the theatre were deeply involved in Patriotism (60 to 70 per cent of the two societies of shareholders were Patriots and only the second society of shareholders counted 8 per cent Orangists). (31) In both drama and opera cohorts, ancien regime councillors of both Patriot and Orangist allegiance, both before and after the coup-year 1787 were practically absent, even while the French Opera performed at the Stadtholder's court in Breda in the time of the Orange Restoration (1787-1795). There appears to have been a culture cleft between the theatre (on which the city government frowned from the beginning) and the chic concert series in the Bierstraat, in the former home of the Orangist councillor Cornets de Groot. There are subscription lists for these concerts for exactly the same years as the Rotterdam French Opera (1789-92). In the Bierstraat 24 Orangists gathered (42 percent of the cohort of 56 subscribers) against one Patriot (quite a few adolescent children of Orangist regents subscribed here, too). Without stating that democratic Patriotism automatically developed into liberalism (though often it did), there is a strong suggestion that the attitude towards culture of both political-ideological movements was the same, and was based on the ethics and aesthetics of the Enlightenment theatre. Disinterest in theatre-going (except, perhaps the Royal French Opera in The Hague), is manifest in the old regent caste, continuing in the anti-theatrical prejudice of the orthodox regents into the Restoration era and beyond, first as conservatism and later as Christian-socialism.

Our conclusion is that the hard core of the Rotterdam theatre audience from the ancien regime to the days of women's lib was wealthy and trade-based. The city (merchant) elite was as much 'compromised' in the culture of melodrama as it was later to become in naturalism and early symbolism. In fact, it often concerned successive members from the same families. Taste evolved, but the social composition of the audience remained stable with respect to occupation, wealth, power, sociability, education, and political 'liberalism'. Differences in data are rather due to lack of archives, than to real differences, or to the broadening of subscription opportunities, which tend to relatively diminish the dominance of the merchant class.

An Overview of the Implicit Social Variables of the Subscribers, 1773-1912

Besides the explicit variables relevant in the traditional theatre historiography, there are some implicit variables of theatre-going which were neglected in it. We consider three of them: religion, woman, and age.

Religion is relevant for pre-twentieth-century theatre-going because of the anti-theatrical prejudice in Calvinism, the official religion of the country. From the origin of commercial theatre in the early seventeenth century, orthodox Protestants agitated against theatre as being the very pomp of the devil. (32) From the first they were opposed by more liberal Protestants, particularly Arminians, who saw no evil in decent theatre, and in fact appreciated theatrical shows. As such, theatre tended to become an issue between Estate-minded regents (lenient), and Orangist regents (severe). Lenient Estate-minded regents, supported by Arminians, stood at the basis of the short-lived Rotterdam theatre (1631) and supported the performances given by Jean Baptiste van Fornenbergh and Jacob Sammers in Abraham Boschman's Tennis Court at the Kipstraat during the first stadtholderless period (1650-1672). Support for theatre from this angle was lasting. As table 3a (appendix 3) shows, both in opera but particularly in drama Arminians remained strongly over-represented in the theatre as compared to their position in the city. Their impact in the first shareholders cohorts (1773, 1780) was even greater: 37 per cent of the first shareholders and 44 per cent of the second were Arminians. Rich merchant families like Van Rijckevorsel, Van der Hoop, Havelaar, and Burger were among the staunch sponsors of a standing theatre.

The French Church ('Walen'), too, was well over-represented in the theatre, particularly in the opera cohorts. At the end of the seventeenth century these 'Huguenots', part of the official religion, were often stern Calvinists, but in the eighteenth century they became a major force in patronage of particularly music theatre. Mennonites followed them, as did a portion of the Lutherans. Restrictions on Jews were repealed after the 1795 Revolution, but it took about three generations before Jews participated regularly in the subscription lists. The later nineteenth century witnessed a boom of Jewish patronage. Best known are the bankers' family Ezechiel and the merchant family Jacobson. Louis Pincoffs, who caused the largest bankruptcy scandal in Rotterdam's history (1879) was surely the most notorious. The Van den Berghs, who stood at the basis of Unilever, moved to Rotterdam in the early 1890s.

Roman Catholics, who supported the early opera troupes quite well, withdrew in the course of the nineteenth century. The Restoration era showed a tendency to keep Catholics out of public offices. Their subsequent struggle for emancipation led to the re-institution of bishoprics in 1853, and from 1868 onwards to an ideological struggle with liberalism. In its wake, Catholic leaders discouraged theatre-going among their flock, which is indirect evidence for the extent to which a standing theatre was considered a liberal institution. Until about the First World War, the Catholic journal Maasbode refused to review theatrical shows, and opposed every attempt at municipal subsidy for the theatre. With respect to theatre, Catholic politicians for a long time sided with anti-revolutionary Protestants, who, of course, had a history of theatrical prejudice. Both the Roman Catholic and the Dutch Reformed denominations were underrepresented in the cohorts; Roman Catholics more so than Dutch Reformed. Table 3a shows that the city became slowly more Protestant. This was to a large extent the result of migration to the city from the Protestant countryside. The distribution of religions over cohorts is not due to chance. (33)

With respect to wealth, the two largest denominations (Roman Catholics and Dutch Reformed) were under-represented in the highest wealth class and often over-represented in the lower classes. For Arminians and French Church, this pattern was reversed. Lutherans and Jews formed more of a ragged pattern. In the course of the nineteenth century the tendency increased to change religious allegiance. Particularly the Dutch Reformed Church lost members to the smaller denominations with a higher status (French Church, Arminians). Noteworthy is the growing proportion of those who left the church. As table 3a shows, nondenominationality increased slowly, accelerating at the beginning of the twentieth century, but the rate was much higher for theatre-goers than for the city population in general. The hypothesis that religion was a relevant variable proved correct. The more orthodox denominations, particularly the Dutch Reformed, were under-represented, whereas the more liberal Protestant denominations were over-represented. These small denominations also were more wealthy. We had not expected that the Roman Catholics responded so negatively to the theatre.

The subscription lists, particularly those before 1860, suggest that the audience was predominantly male. However, from the beginning of the regular seasons (1773), some women independently subscribed for tickets. Their numbers are given in table 1c. Table 1c shows a slow increase of women independently subscribing, increasing from less then 2 per cent before 1850, to about 5 per cent in the 1860s to some 15 per cent in the 1910s. The relatively high figure for women at drama, 1773-1792 is due to an exceptional subscription in 1778-79. The theatre would be closed if it did not get 300 subscribers and hence, some shareholders had their wives and daughters subscribe independently.

Such small figures, however, are misleading, for male subscribers took regularly more than one ticket. Sometimes they had to state explicitly that these were for women for often women could enter cultural events at bargain prices when accompanied by a man. In other cases we are not sure that the extra tickets were for wives and daughters, but the number of extra tickets then gives at least an upper limit of the number of female companions. Table 1d enumerates the numbers of extra tickets taken by subscribers for the cohorts before 1860. We had to calculate numbers from the original lists and, hence, had to split the macro-cohorts. The lists of coupon subscriptions do not specify the number of persons for whom tickets were bought. Hence, there are no specified data after 1860.

The numbers give extra seats reserved by male subscribers and the male-female ratio gives the proportion of these extra tickets related to the number of male subscribers. Notice that in the years 1817-1820, the number of women in the boxes of the French Opera was probably larger than that of male subscribers and therefore is over 100 per cent.

It is clear enough that opera was far more a family outing than Dutch drama, and that the potential number of women taken to the theatre was lowest in the 'melodrama years'. This might confirm what critics remarked: a decent person could not take his wife to the theatre because of the piquancies in the plays, and, perhaps, the presence of prostitutes in the pit. (34)

In the 1860s a few patrician merchants' wives subscribed independently on guarantee lists of the German opera. Coupon buying women for the drama quite often ran a shop or a coffee house or worked as teachers. Widows without occupation or resident daughters also signed for coupons by the end of the century. At the end of the century wives quite often signed for the coupons, for which their husbands paid. There are more signs that women took an active interest in the household's cultural affairs. The 1894 yearly report of the Art Circle [Kunstkring], an elite society, complained of the fact that Rotterdam merchants preferred light entertainment to high art, so that the concerts suffered a 75 per cent women audience. Emancipation apparently was not cheered by all.

Age also is a relevant, but neglected, variable. The limited space here permits us only to say that the subscribers were relatively old, particularly after 1890. The melodrama cohort, however, was remarkably young, though settled.

Differentiation per Rank

The traditional narrative of audience composition in the long nineteenth century shows a tendency to equate social class and rank. Indeed, many nineteenth-century essayists have pictured the gallery as the domain of apprentices, servants, sailors, and working class persons, and juxtaposed it against the economic elite in the stalls and balcony. Descriptions of audience behaviour on the different ranks invited such a classification. Indeed, when kissing sounds were heard or soubrettes were yelled at from the galleries, its patrons swerved far from bourgeois mores. Is there a 'real' class division per rank?

For drama this relation is far more outspoken than for opera. For reasons of space we only compare German Opera, 1860-1882, with Dutch-spoken drama, 1902-1912 (table 1b, appendix 3). The first rank of the Opera (balcony and stalls) counted slightly more of the very rich 'trade' persons and 'academics' than the second rank (boxes and parquet) and the third (pit). In the pit, the third and fifth wealth classes were rather evenly spread for these occupations. 'Middle class' occupations and 'professionals', grouped on the third rank, and in wealth classes two and three. We surmise that this difference between the job categories 'trade' and 'academics' on one hand, and 'middle class occupations' and 'professionals' is due to the income structure in the city.

Van Eysden shows a clear change, per job category, per wealth class, per rank. Class-five 'trade' persons completely dominate the first rank, whereas their pit-fellows are mainly in the second and third wealth classes. The pattern of decreasing rank, relating to decreasing wealth is particularly clear with the 'professionals'. The large number of high-income first rank drama patrons without a profession to a large extent consisted of the unmarried daughters from the patrician class (the spinster-sisters Rueb, Van Reesema, Van Rossem, and so on). It must be recalled that the third rank was a bit exceptional, since--due to acoustics--it provided a bargain price for what in fact was a first rank (balcony).

In the German Opera, merchants are the largest group in all three ranks. With Van Eysden, 'middle-class' occupations are the largest group on the third rank (though 'professionals' almost equalled them) and 'professionals' dominate the fourth rank (pit), followed by 'middle-class' occupations at some distance. (35) We also inspected the French Opera (1817-1844) and Tivoli Theatre (1890-1894), which showed a more skewed though not radically different distribution. (36) Thus, there are differences in rank, wealth, and occupation in both opera and drama, but they are more pregnant for drama. Overall, these differences concern different bourgeois occupation categories, for working class persons hardly play a role (only at the drama), and are moreover, distributed over three ranks.

With respect to 'age' a calculation per rank did not make sense for all cohorts. The eighteenth-century and melodrama cohorts for instance consisted predominantly of subscribers for the pit. For such macro-cohorts as the French Opera 1817-1844 or the German Opera 1860-1882 the calculation of the mean age would, due to the spread of the cohort in time, lead to distortion rather than insight. At Van Eysden's most participants fell in the age-class 41 to 50 years on all ranks but the pit. In the pit the age-class 31-40 was the largest. Proportionally the first, third, and fifth ranks showed more age difference than the second and fourth ranks, as is shown in table 2b. The pit and upper boxes have a slightly younger audience than the other ranks, while the first rank tends more to a middle-age patronage. The differences are, however, small and the distribution is not statistically significant. (37)

Table 4c gives the distribution of function categories per rank per wealth class for the German Opera (1860-1882) and for Van Eysden (1902-1912). Function categories cumulate (one person can score in more function categories). The number of functions differs from that in table 4a, since we did not find wealth data for all subscribers. The distribution in both cohorts is not due to chance. (38) In the German Opera, first-rank patrons held an average of 3.2 functions; at the second rank this was 1.9, and on the third rank 1.2. At the drama of Van Eysden, first-rank patrons held 0.9 functions, second-rank patrons 0.2, third-rank patrons 0.2, and fourth-rank patrons 0.04. The fifth rank did not hold any functions. It is thus also true, particularly for the drama, that more functions were held when the price of the seat increased, and that involved occupation and wealth.

Table 5c gives the distribution of sociability categories per rank per wealth class for the German Opera, 1860-1882 and for Van Eysden, 1902-1912. Membership categories cumulate (one person can score in more categories of societies). The number of functions differs from that in table 5a, since we did not find wealth data for all subscribers. The distribution in both cohorts is not due to chance. (39) In the German Opera, first-rank patrons held an average of 4.2 memberships; at the second rank this was 2.9, and on the third rank 2.1. At the drama of Van Eysden, first-rank patrons held 2.3 memberships; second-rank patrons 0.7, third-rank patrons 0.5, fourth-rank patrons 0.4, and fifth rank patrons 0.4 memberships. It is thus also true, particularly for the drama, that more memberships were held when the price of the seat increased, and this implied an impact of occupation, wealth, and function holding. The drama cohort shows a dramatic skewness. Sociability in opera cohorts was more evenly divided over all ranks, in comparison with the later drama cohorts (the eighteenth-century and melodrama cohort predominantly subscribed for the pit). Yet, the decrease of average memberships is significant.

Table 6b gives the percentages of persons per cohort, for which we found data on education. Both per rank and per wealth class the number of 'educated' decreases when the rank is 'lower' or the income smaller for those cohorts which fell within the range of the new education system. The relation between wealth class and education is not due to chance. (40)

It turns out that occupation (trade), wealth, function-holding, sociability, and education are strongly linked with rank. It must be stressed that all such differentiations are in the bourgeois class, and have at best to do with the differentiation 'trade/academic' occupation categories versus 'middle class/professional' occupation categories. 'Objective' class differences per rank are very weak; the few working class persons we detected were not even on the same (low) rank.

A Stable Background

The traditional dominant narrative assumes on the one hand elite stability, on the other hand it assumes the opposite with respect to patrons of the melodrama. Such assumptions also concern the social background of subscribers. This background will be analysed from the point of view of social mobility. The implied hypotheses of the dominant narrative are that in the first half of the nineteenth century, opera patrons have a more stable ('elite') background than melodrama patrons, who perhaps show signs of upward mobility. As Dutch social historians generally assume that upward social mobility accelerated after 1870, it is also of interest to analyse the later cohorts (opera, 1860-1882; drama, 1890-1895 and 1902-1912) with respect to social mobility. We collected data on the social backgrounds of the subscribers, if these were born in the Rotterdam region. It turned out that about half of the subscribers generation was born in the region, and that per parental generation this half-life persisted. Because it regularly happened that fathers or grandfather themselves subscribed in one of the cohorts, we could also generate genealogical data for five generations. It must be stressed, however, that these fourth and fifth generation data are not of the same sort as the second and third generation data. We systematically searched for these latter data, whereas the fourth and fifth generations are just generated because a subscriber's father or grandfather had also subscribed. It did not make much sense to study the eighteenth-century cohorts in that way, because researching theatre-going in a city without a theatre is fruitless. We included, however, the father and grandfather generations of the early-nineteenth century cohorts, because of the assumption of social decline in the melodrama cohort.

Tables 7a-7d (appendix 3) sum up the results. (41) From Table 7a it appears that the social backgrounds of the fathers and grandfathers of the subscribers for opera and drama in the first half of the nineteenth century were not wildly different. There is a steady difference in the percentages of 'trade' and 'middle class' occupations. Noteworthy, however, is the absence of the expected massive lower middle class and working-class background of melodrama patrons. The outcomes for the first half of the nineteenth century tend to be reproduced over the whole period. Particularly in grandfather generation in the later nineteenth-century cohorts (Tivoli theatre!), there is an increase in working and middle class background, but it is small when compared to the category 'trade'. There is an interesting difference between the categories 'trade' and 'professionals' in the subscriber generation Tivoli (1890-1895) and Van Eysden (1902-1912). The Tivoli Theatre intended to become an intimate surrounding for an arty elite, and its seats in the tiers were relatively expensive. Apparently this helped to shut out subscribers of the occupation category 'professionals'. Curious, thus, that relatively more subscribers for this arty surrounding had a working-class grandfather.

The distribution of occupations among the first three generations is significant per cohort and therefore not due to chance. (42) Tables 7b and 7c indicate that also with respect to wealth, no sharp differentiations occur between the later nineteenth-century cohorts. That is to say, for the opera (1860-1882), the categories 'trade' and 'academics' stay relatively wealthy, while the other occupation categories have their basis in class 3. For drama (1902-1912), the same counts, except that 'middle class' occupations and 'professionals' in the father generation tend to be in the second wealth class instead of the third. Both for drama (1902-1912) and opera (1860-1882) the distributions of wealth class over the first three generations are not due to chance. (43)

The melodrama cohort (1824-1830, table 7d) which particularly interests us here, yielded less wealth data since the earliest list of rateable values stems from 1803. Data from the Marriage and Burial Tax (MTB) gave a better picture for the grandfather generation, of which 38 per cent was taxed in the highest class (a fortune of at least fl. 12,000.-). The tendency is that fathers in 'trade' were relatively more wealthy than in other occupations, but few fell in the lowest wealth class of the rateable value. The Marriage and Burial tax points roughly to the same. A tax sum of fl. 3.- or fl. 6.-suggests middle class wealth (fl. 1500.- to fl. 3000.-).

According to the traditional narrative, we should expect that the melodrama ancestors scored badly in social and cultural variables as well, but they did not. Of 178 identified fathers, 40 held political and social functions (22 per cent), and of the 43 identified grandfathers 10 did (23 per cent). Among the functions held by the father generation were a burgomaster, an alderman, and nine councillors. Moreover we found two consuls and a member of the Provincial Estates. The grandfather generation included a burgomaster, two members of the Provincial Estates and six councillors. For 31 of the 178 fathers we found data on sociability (17 per cent). These concerned the elite members of the cohort. They shared 59 memberships. It must be recalled that there are few data for eighteenth-century sociability. In the grandfather generation, therefore, the number was, as we expected, worse: ten persons (13 per cent) shared 17 memberships.

The score for the melodrama cohort is not extremely different from the data of the later cohorts. Comparing it to opera (1860-1882), and drama (1902-1912), which cohorts benefited from the generally better archive position of the nineteenth century, it turned out that the fifth wealth classes took by far the most functions. (44) Remarkably, the number of functions taken by drama-fathers was (slightly) higher than with the subscribers themselves. Political functions dominated in the background of these subscribers' parentage, so that those in power generated more theatre-minded offspring. (45) Of the opera-fathers, 24 were associated with liberal election clubs and four with conservative clubs; of the opera grandfathers, ten had been Orangists and 34 Patriots. Of the drama-fathers 67 had ties to liberal election clubs and five with Christian-Democrats (of 797 persons). The grandfather generation of the drama-subscribers counted eleven Patriots and two Orangists (of 472 persons), nine liberals, one conservative and a socialist. (46) These are small figures with respect to the cohorts, but they are of interest, since little political allegiance could be detected and they mesh with the findings for political allegiance in the subscribers' cohorts themselves.

The outcome of these analyses of the social backgrounds of the subscribers does in no way support the implied assumptions of the traditional narrative with respect to the social background of melodrama subscribers. The analyses demonstrate an overall stability in the social composition of subscribers in time. Many of the late eighteenth-century merchant families, who did not belong to the regent caste, still subscribed for theatre tickets at least until the Great War. The drama and opera cohorts show some differences which allow the conclusion that opera patrons had a socially 'better' background, but the differences are relatively small.

4. Theatre as a prison of longue duree?

The survey of the social variables relevant for testing the traditional historiography on the composition of the audience in the long nineteenth century points to the untenableness of the traditional historiography. Subscribers formed a considerable part of the audience and were generally wealthy, powerful, sociable, well educated, and politically engaged. They were over-represented in the religious denominations that were tolerant of profane culture and rather old. Other cultural variables also show continuity rather than change over the period. Rank differences in social background do not automatically imply a non-bourgeois social background of the rank as a whole. As regards the genealogical background of subscribers, theatre-going had its basis in wealthy traders under the level of the aristocrat regent families, and spread into middle class professions both in service ('professionals') and the economic activity ('middle class'). The theatre, therefore, was essentially and throughout the whole period of time we studied a form of amusement for a solid citizenship. Theatre-going ran in the family. The effects of upward social mobility per generation in at least the city-born subscribers was relatively small, but tended to upward mobility. This is more spectacularly noticeable on the individual career level. This often meant that the rich got richer.

Our analyses point to tenacious elements in the social background of theatre-goers. They were often wealthy, occupied in trade, mighty, and took an interest in other cultural organizations. The beginnings of stimulating profane theatre culture in Rotterdam were located in the Estate-minded regents, who opposed the orthodoxy then related to Orangists. After 1672 some of them became dissenters (Arminians), others became the enlightened regents, who had awkward relations with the democratic Patriot burghers in the later eighteenth-century, often also dissenters and the founding fathers of the first standing theatre. As liberals--and increasingly unchurched--they enter the twentieth century. It looks likely that the theatre, too, is a prison of longue duree. (47)

Other aspects of longue duree were found in the time series analyses of the ticket sales, where the impact of determining season and audience loyalty diminished over time, but remained of effect and--for Rotterdam--was never replaced by supply variables before 1916 (the end of the data series).(48) Innovation and resistance against it must not be interpreted in terms of social class differences, but as battles over changing tastes or the acquisition of cultural capital related to getting actively involved in such battles. This, however, concerns a rather small group of people who set trends, since the social composition of subscribers does not allow the identification of large differentiations per theatre, or forms of theatre (before 1916). Quite often persons made subscriptions for all venues and forms of theatre (e.g. German Opera and drama both at the Tivoli and the Grand Theatre). The time series analyses also denied the significance of differentiations shaped by so-called innovations, either positively or negatively. A third aspect of longue duree is the structural underfinancing of the cultural circuit, due to a resistance to pay for tickets that covered the real costs, and a chronic over-estimation of the number of potential patrons. (49)

The question is, whether the results reached here can be reproduced in other cities, and whether things radically changed in the twentieth century, under the impact of the welfare state and the 'theatre revolution' of 1970. The crisis in Dutch theatre around 1970 has traditionally been linked with a change in audience composition, but research into it is still in its infancy. (50)

The results of this present analysis opens up possibilities for theatre historiography along the lines of a Braudel or Vovelle, since it is clear that the level of the event (e.g. innovating individuals), the medium level (e.g. theatre companies) and the long-term processes apparently interact in a more intricate way than the old hero and dragon stories of theatre history supposed. The perspectives of the event are multifaceted, according to the level from which the event is viewed. This relates also strongly to the question of what the object of theatre history actually is: the performance, as most theatre studies scholars claim, or the social system. In the end we need both in order to give relevant explanations.

Appendix 1. The subscription lists

Eighteenth Century Lists

Dutch Drama, seasons 1773-74 to 1777-78, 44 subscribers season tickets and prepaid tickets

Dutch Drama, season 1778-79, 113 subscribers for season tickets to rescue the theatre

Dutch Drama, season 1792-93, 25 subscribers for season tickets

French Opera, seasons 1789-90 to 1791-92, 120 subscribers for season tickets

Concert Series, 1789-90 to 1791-92, 56 subscribers for season tickets.

Nineteenth Century Lists

Dutch Drama, season 1824-25 to 1829-30, season tickets, 253 subscribers

Dutch Drama, season 1867-68, coupons, 115 subscribers

Dutch Drama, season 1890-91 to 1894-95, coupons, 582 subscribers (Tivoli Theatre)

Dutch Drama, seasons 1879-80 to 1898-99 (incomplete!), coupons 363 subscribers

French Opera, seasons 1817-18 to 1819-20 (boxes only), season tickets, 37 subscribers

French Opera, seasons 1836-37 to 1838-39, season tickets, 208 subscribers

French Opera, season 1843-44, season tickets, 81 subscribers

German Opera, season 1860-61 to 1861-62, 147 shareholders with box diploma

German Opera, season 1865-66 to 1866-67, 560 discount subscribers to guarantee fund

German Opera, season 1867-68, coupons, 442 subscribers

German Opera, season 1879-80, coupons, 469 subscribers

German Opera, season 1881-82, season tickets, 130 subscribers

Twentieth Century Lists

Dutch Drama, seasons 1902-03 to 1911-12, coupons, 1476 subscribers

German Opera, Maifestspiele 1909 and 1911, 677 subscribers from Rotterdam

Italian Opera, season 1933-34, 248 subscribers in Rotterdam

In view of the overlap of subscribers, and also in order to prevent too many analyses on small cohorts of subscribers, we also formed combinations of cohorts, to wit:

Dutch-spoken Drama, 1773-1792: 157 persons

French Opera, 1817-1844: 276 persons

German Opera, 1860-1882: 1150 persons

Appendix 2. An assessment of the data

Since the aggregated cohorts contain more than one season, the highest rank occupied by a theatre-goer has been taken for our analyses. In some cases persons switched rank. This could be related to upward social mobility, but also to ranks no longer being available. The very rich sometimes bought coupons for a more ranks. To include these in all these ranks would, however, distort the facts of theatre-going, since in many of these cases tickets were bought to give to employees or domestic servants (a chauffeur, the lady-companion to the mistress, or maids to go out on Sunday). Patrons, however, were overwhelmingly loyal to one rank or even to a seat. In many cases persons bought more than one season ticket. Most likely they were accompanied by their wives. These have not been separately included, except when the ticket is explicitly put on their names.

The cohorts vary considerably in size; subscription lists for season tickets in the eighteenth and early nineteenth centuries are much smaller than coupon subscription after 1860. The many cohorts made it unhelpful to draw representative samples, since the rules of that game demanded that we include most of the subscribers anyway. So, we gathered information about all subscribers. All cohorts had a small amount of subscribers who could not be identified.

In relative numbers, the subscribers before 1792 form only a small part of the audience. The melodrama season ticket holders (1824-30) constitute about 50 per cent of the average pit occupation rate per season. Since boxes and amphitheatre were hardly occupied when Dutch drama was staged, this is not a bad score. It is, of course, likely that buyers of single tickets may have had less income than the season ticket holders, but in the nineteenth-century (Dutch) context it is highly unlikely that the rich and mighty would mix with the lower classes. Hence, the reasonable expectation on the basis of the social composition of the season-ticket holders is, that the pit had a solid citizen profile even in the melodrama years. The 1867 drama list is less likely to be representative.

The figures for the Tivoli theatre, 1890-1895 are more complex. This theatre, opened in 1890, saw its patronage decline from the first season onwards. Coupon sales, however, remained relatively stable at 10 per cent average for the first and second rank and about 5 per cent for the pit, so that its relative weight became larger. The use of coupons also depended on rank and weekday. Using average ticket sales per weekday over all five seasons, coupons-tickets on the first ranks formed about 25 per cent of the audience on Sunday, and on Saturday 33 per cent. Monday and Thursday are in between these figures (29 to 31 per cent). For the second rank these percentages are 20, 19, 19, and 20 respectively, and for the third rank (the pit) it was about 7 per cent on all occasions.

In the Grand Theatre, Van Eysden gave Sunday plays and performed on Tuesday and Friday. The first rank on Sunday had an average season ticket and coupon participation of some 25 per cent, which increased for Tuesdays to some 73 per cent, and for Friday to some 65 per cent. For the second rank these figures are 28, 50, and 45 per cent respectively, whereas for the pit (fourth rank) they dropped to about 16, 26, and 22 per cent respectively. The figures for the third and particularly fifth rank (side balcony and side boxes) were even lower. The third rank, moreover, was not always available at its reduced price. It was normally part of the first rank. These figures are also based on the average ticket sales for the normal performance days over the whole period of Van Eysden's management (1900-1916). For the first and second ranks in both theatres, the scores are particularly good, for the pit they are less solid, but still not bad. For the lower ranks the score is too low to estimate the social composition. The side boxes, amphitheatres, and galleries were normally very poorly occupied, except for Sunday plays with reduced prices and cheap Monday plays. The fact that spectators for these ranks had to enter through the back door and were not allowed in the main foyers caused bad blood.

In the later eighteenth and the first half of the nineteenth century boxes were en vogue for the opera. We have most of the box audience for the Rotterdam French Opera 1789-92 and the Royal French Opera of The Hague, 1817-20. For these latter years no subscriptions for the pit have been preserved, but for the pit of the 1789-92 opera seasons the number of subscribers was considerable (some 35 per cent of the average occupation rate). The best figures come from the French Royal Opera seasons 1836-38, where we have about all the pit, amphitheatre and box audience. The Amsterdam French Opera, 1843-44 had 200 subscribers for the pit (over 50 per cent of the seats). Here no subscriptions were made for the boxes, although the ticket sales accounts prove that these were sold, too, but not on a season-ticket basis. The 1150 subscribers for the German opera must have formed, on the basis of their number and the occupation rates, a considerable part of the audience of the stalls, balcony, boxes, and pit. On the whole, we think these are sufficient data to work with.

Data on address, occupation, religion, and age were found in the address books and the population registers. Occupations led to an immensely fragmented list. We use here six macro-categories: trade, middle class, academics, professionals, working class, and none (see for the specification of the job categories the legend of table 1a, appendix 3). Wealth classes were based on several city taxes. These wealth classes all have the following characteristics: class 1 is the lowest category, based on an acknowledged small income; class 2 forms a transition segment to middle class incomes; class 3, hence, covers solid middle class incomes. Class 4 represents a transition segment to the very rich, who are in class 5. From 1803 to about 1850 the rateable value was the basis for most local taxes and for the national wealth tax; after 1856 a direct poll tax, roughly based on expenditure, was calculated with diverse, complicated formulas, until in 1893 a formal income tax was instituted. This lasted till 1922, when local direct taxes were abolished. The changes in the tax system make it impossible to exactly compare the outcomes of wealth for the several cohorts in time, although they also are not widely apart. (51) The outcomes for rateable value and the diverse sorts of income tax differ, due to the limit on the former (hence, there are relatively fewer class 5 rateable values than class 5 incomes). (52) In the analyses we have taken account of inflation.

Sources for functions were the official lists published by the city, the address books, published genealogies in the series Nederlands Patriciaat [Dutch Patriciate], and the first and fifth volumes of series Bronnen voor de geschiedenis van Rotterdam [Sources of Rotterdam History], compiled at the Municipal Archive: De Regeering van Rotterdam, 1328-1892, ed. J.H.W. Unger (Rotterdam, 1892), and De Vroedschap van Rotterdam, 1572-1795, ed. E.A. Engelbrecht (Rotterdam, 1973).

Sources for sociability were membership-lists of societies and clubs. These were divided into the categories as given in table 5a (appendix 3). Data were also derived from address books, which, however, only give the committee members of a club. Membership lists were not always available for the whole period. For instance, the lists for the important Society for Public Welfare (abbreviated as the Nut, founded in 1784) are not available for the period ca. 1810-1880, although some of its active members were listed in the address books; those for the Society for the Advancement of Music [Toonkunst] are only available after about 1900. For Harmony [Harmonie], one of the most important cultural clubs in the nineteenth century, the membership lists for the early years are missing. (53) There is also a bias for the clubs frequented by the well-to-do, since they existed longer and their members were more careful archivists. (54) The 1940 bombing destroyed many archival materials which were kept at home.

For eighteenth-century political allegiance, we benefited much from Eric Palmen, De Kooplieden van Rotterdam [Rotterdam Merchants], internal report MAR (1994). Nineteenth-century data came from the New Rotterdam Journal and the archive of the election club Burgerpligt. School archives (secondary training) are only available from the 1860s onwards. We consulted the school archives of the Burgher-schools (five and three year course), and the Latin school (later gymnasium). Data on education before 1860 were restricted to higher education. Medical doctors, lawyers, and judges had to finish the university; whereas surgeons had to finish the Illustrious School of surgery. Merchant class children often took one of four courses (or mixed them): they got practical training in the business companies of their class, they went to private French schools, or to boarding schools. Private teaching at home was the most expensive option. Neither of these road maps to knowledge has left sufficient archival documents. There are also data for technical education, but these are complicated and incomplete, while in fact the occupation of a subscriber already signals what sort of training had likely been followed. It is a pity that no archive materials are left of the comprehensive primary teaching schools (ULO and MULO), in which many of the middle classes were (supposed to have been) trained. In 1889 some 1500 boys and 1450 girls attended those schools.

Appendix 3. Basic Tables

Table 1A Opera, Wealth Class by Job Category

Wealth class
Jobcat         1      2     3      4      5      sum   %

               French Opera. 1789-1792
               Rateable Values

trade          --     --     11     15     15     41   80.4
middle class   --     --     --     --     --     --   --
academics      --     --      1      1      1      3    5.9
professionals  --     --      3      2     --      5    9.8
none           --     --      1      1     --      2    3.9
sum            --     --     16     19     16     51  100
percentage     --     --     31.4   37.3   31.4  100

               French Opera. 1817-1844
               Rateable Values

trade          --     --     46     57     49    152   64.4
middle class   --     --      7      4      2     13    5.5
academics      --     --      7     15     13     35   14.8
professionals  --     --     16      9      3     28   11.9
none           --     --      2      2      4      8    3.9
sum            --     --     78     87     71    236  100
percentage     --     --     33.1   36.9   30.1  100

               French Opera. 1852-59, approx.
               Poll Tax

trade          --     11      3     16     48     78   70.3
middle class   --      2      5      2     --      9    8.1
academics      --     --      1      2      5      8    7.2
professionals  --     --      2      3      1      6    5.4
none           --      1      1      4      4     10    9.0
sum            --     14     12     27     58    111  100
percentage     --     12.6   10.8   24.3   52.3  100

               German Opera, 1860-1882
               Poll Tax

trade           1     44     97    130    237    509   56.7
middle class    1     28     70     41     15    155   17.3
academics       1     10     15     24     36     86    9.6
professionals   1     37     39     19      8    104   11.6
working class  --     --     --     --     --     --   --
none            1      5      7     11     20     44    4.9
sum             5    124    228    225    316    898  100
percentage      0.6   13.8   25.4   25.1   35.2  100

               Maifestspiele. 1909-1911
               Income Tax

trade           2      9     37     29    106    183   33.9
middle class   14     16     25      9     11     75   13.9
academics      --      3      7     10     14     34    6.3
professionals  56     73     48     11     13    201   37.2
working class   7     --      2     --     --      9    1.7
none            4      6     11      8      9     38    7.0
sum            83    107    130     67    153    540  100
percentage     15.3   19.8   24.1   12.4   28.3  100

Wealth class
Jobcat         1      2      3      4      5      sum   %

               Drama, 1773-1792
               Rateable Values

trade          --     --     15     13      9      37   75.5
middle class   --     --      2     --     --       2    4.1
academics      --     --     --      1      1       2    4.1
professionals  --     --      3      2     --       5   10.2
none           --     --      1      1      1       3    6.1
sum            --     --     21     17     11      49
percentage            42.9   34.9   22.4  100

               Drama, 1824-1830
               Rateable Values

trade          --      1     60     43     23     127   62.0
middle class   --     --     12     13      7      32   15.6
academics      --     --      3      6      2      11    5.4
professionals   1     --     26      1     --      28   13.7
none           --      1      2      4     --       7    3.4
sum             1      2    103     67     32     205  100
percentage      0.5    1.0   50.2   32.7   15.6   100

               Drama, 1867-1868
               Poll Tax

trade          --      2     11     12     30      55   55.6
middle class   --      4     13      2      3      22   22.2
academics      --     --      2      1      6       9    9.1
professionals  --      3      3      2      4      12   12.1
none           --     --     --      1     --       1    1.0
sum            --      9     29     18     43      99  100
percentage     --      9.1   29.3   18.2   43.4   100

               Tivoli, 1890-1895
               Income Tax

trade           6     24     55     39    103     227   54.3
middle class    5     16     18      8     14      61   14.6
academics       1     10     13      8     14      46   11.0
professionals   6     13     19      7      7      52   12.4
working class   2     --     --     --     --       2    0.5
none            4      6     12      4      4      30    7.1
sum            24     69    117     66    142     418  100%
percentage      5.7   16.5   28.0   15.8   34.0   100

               Van Eysden. 1902-1912
               Income Tax

trade           9     44    130     87    282     552   44.2
middle class   18     48     82     37     36     221   17.7
academics       2      5     14     19     33      73    5.9
professionals  39     91     92     33     20     275   22.1
working class   4     --      4     --     --       8    0.6
none           13     26     29     22     28     118    9.5
sum            85    214    351    198    399    1247  100
percentage      0.8   17.2   28.1   15.9   32.0   100

Opera 1789-72: N = 120. Valid Total 83 (69.2%); Opera 1817-44: N = 276.
Valid Total 236 (85.5%); approximation of French Opera, 1852-59: N =
134. Valid Total: 111 (82,8%). Opera 1860-82: N = 1150. Valid Total: 898
(78.1)%; Opera. 1909-11: N = 677. Valid Total: 540 (79.8%). Drama
1773-92: N = 157. Valid Total 49 (31.2%); Drama 1824-30: N = 253. Valid
Total 205 (81.0%); Drama 1867-68: N = 115. Valid Total 99 (86,1%); Drama
1890-95: N = 582. Valid Total 418 (71.8%). Drama 1902-1912: N = 1476.
Valid Total 1247 (84,5%).
Main job-categories contain the following sub-categories: Trade:
bankers, stock traders, insurers, merchants, transit-traders, major
industrials, trade agents, and the in-between categories of trade-and-
finance, trade-and-industry, trade-and-transport. Middle Class:
wholesalers, retailers, artisans, building-trade, book-trade, hotel and
catering industry. Academics: medical professions, notaries, attorneys
and advocates. Professionals: employees, civil servants, clergymen,
technicians, officers in army and navy, education, art and journalism.
Working Class: skilled and unskilled industrial and artisan labourers,
domestic servants, seamstresses. None: unknown, not relevant
(institutions), private persons.

Table 1B Rank, job-category, and wealth

Opera, 1860-1882, Poll Tax        Drama, 1902-1912, Income Tax

Wealth class
Jobcat  1   2   3   4   5    sum   1  2   3   4   5    sum

trade
Rank 1  --   1   3  23  103  130   2  14  67  70  268  421
Rank 2  --   7  21  26   44   98   2  13  39  13   13   80
Rank 3  --  23  46  40   48  157  --   3   7   3    1   14
Rank 4                             5  14  15   1   --   35
Rank 5                            --  --   2  --   --    2

middle class
Rank 1  --   1   1   1    1    4   2   3  14  16   28   63
Rank 2  --   1   7   4    4   16   2  20  37  10    7   76
Rank 3  --  19  38  15    6   78   3   6  14   8    1   32
Rank 4                             9  16  16   3   --   44
Rank 5                             2   3   1  --   --    6

academics
Rank 1  --  --   1   5   16   22   1   2  10  17   33   63
Rank 2  --  --   4   6    6   16   1   2   4   1   --    8
Rank 3   1   5   8   7    5   26  --  --  --   1   --    1
Rank 4                            --   1  --  --   --    1
Rank 5                            --  --  --  --   --   --

professionals
Rank 1  --  --  --   2    1    3   4  12  31  25   19   91
Rank 2  --  --   1   2    4    7   7  19  45   7   --   78
Rank 3   1  17  24   7   --   49   5  16   8   1   --   30
Rank 4                            21  40   7  --    1   69
Rank 5                             2   4   1  --   --    7

working class
Rank 1                            --  --  --  --   --   --
Rank 2                             1  --   3  --   --    4
Rank 3                            --  --  --  --   --   --
Rank 4                             2  --   1  --   --    3
Rank 5                             1  --  --  --   --    1

none
Rank 1  --  --  --   2   10   12   4   8  17  22   28   79
Rank 2  --   2  --   3    1    6   2   6   8  --   --   16
Rank 3  --   2   2   3    1    8   2   5   2  --   --    9
Rank 4                             4   6  --  --   --   10
Rank 5                             1   1   2  --   --    4

Table 3A Religious denominations [source for city data: Municipal
Reports, and A.Th.C. Kersbergen, Mentink & Van der Woude (ca. 1820)]

cohort & perc.   drama      opera      city, %   drama       opera
inhabitants      1773-92    1789-92    ca. 1820  1824-30     1817-44
religions        N = 157    N = 120              N = 253     N = 276

Dutch Reformed   48 (30.6)  26 (21.7)  61.0%     126 (49.8)  107 (38.8)
English Church    7 (4.5)   10 (8.3)               4 (1.7)     5 (1.8)
Scottish Church   1 (0.6)    2 (1.7)              --          --
Reformed,        --         --                     1 (0.4)    --
dissenting
French Church    15 (9.6)   20 (16.7)             11 (4.3)    32 (11.6)
Lutherans        11 (7.0)    4 (3.3)    6.5%      24 (9.5)    20 (7.2)
Arminians        41 (26.1)  16 (13.3)   1.5%      28 (11.1)   36 (13.0)
Mennonites        4 (2.5)    2 (1.7)    0.2%       2 (0.9)     5 (1.8)
Old/Roman        16 (10.2)  23 (19.2)  28.0%      41 (16.2)   49 (17.8)
Catholics
Jewish           --         --          3.5%       5 (2.0)    14 (5.1)
None             --         --         --         --           1 (0.4)
Not relevant     --         --         --         --          --
Unknown          14 (8.9)   17 (14.2)  --         11 (4.3)     7 (2.5)

cohort & perc.   opera      city %  drama      opera       city, %
inhabitants      1850-59    1865    1867-68    1860-82     1880
religions        N = 134            N = 115    N = 1150

Dutch Reformed   52 (38.8)  57.4%   58 (50.4)  529 (46.0)  57.2%
English Church    3 (2,2)    0.2%    2 (1.7)    10 (0.9)    0.1%
Scottish Church  --          0.1%   --           2 (0.2)    0.1%
Reformed,        --          0.5%   --           2 (0.2)    1.3%
dissenting
French Church    14 (10.4)   0.9%    8 (7.0)    77 (6.7)    0.7%
Lutherans         9 (6.7)    4.4%    7 (6.1)    82 (7.1)    3.8%
Arminians        24 (17.9)   1.1%   13 (11.3)   98 (8.5)    1.1%
Mennonites        1 (0.7)    0.4%    1 (0.9)    23 (2.0)    0.4%
Old/Roman        21 (15.7)  30.8%   16 (13.9)  201 (17.5)  30.5%
Catholics
Jewish            9 (6.7)    4.1%    9 (7.8)    84 (7.3)    4.4%
None             --          0.03%  --           6 (0.5)    0.4%
Not relevant     --         --       1 (0.9)    10 (0.9)   --
Unknown           1 (0.7)   --      --          26 (2.3)   --

cohort & perc.   City %  drama       city %  drama;      opera
inhabitants      1893    1890-95     1908    1902-12     1909-11
religions                N = 582             N = 1476    N = 677

Dutch Reformed   57.8%   257 (44.2)  58.3%   623 (42.2)  302 (44.6)
English Church    0.1%     1 (0.2)    0.04%   --           1 (0.1)
Scottish Church   0.04%   --          0.01%    1 (0.1)    --
Reformed,         2.3%     2 (0.3)    6.8%     5 (0.3)     9 (1.3)
dissenting
French Church     0.4%    35 (6.0)    0.2%    68 (4.6)    20 (3.0)
Lutherans         3.1%    39 (6.7)    2.3%    95 (6.4)    59 (8.7)
Arminians         1.1%    61 (10.5)   0.9%   168 (11.4)   70 (10.3)
Mennonites        0.4%     7 (1.2)    0.3%    22 (1.5)    16 (2.4)
Old/Roman        29.9%    91 (15.6)  26.2%   222 (15.0)   95 (14.0)
Catholics
Jewish            3.5%    74 (12.7)   2.6%   207 (14.0)   63 (9.3)
None              1.3%     8 (1.4)    2.2%    22 (1.5)    21 (3.1)
Not relevant     --        1 (0.2)   --        4 (0.3)    --
Unknown          --        6 (1.0)   --       39 (2.6)    21 (3.1)

Table 4A Political and Social Functions

                   1789-92      1817-42      1850-59      1860-82
                   Opera        Opera        Opera        Opera
                   N = 120      N = 276      N = 134      N = 1150

Administration      23           22            9            17
Civil                7           25           19            73
Cultural                          4            3            15
Economic            43           52           37           135
Education            1           58           42           134
Financial           16          118           91           373
Juridical            7            2            1             3
Church              16           88           65           251
Social                           18           21           120
Military            57           34           22           101
Political           63          145           70           157
Charitative          6          165           95           300

Sum functions      239          731          475          1679
Sum persons:        71 (59.2%)  169 (61.2%)   96 (71.6%)   621 (54.0%)
Functions/persons    3.4          4.3          4.9           2.7
Functions/cohort     1.9          2.6          3.5           1.5

Function Level (sum total of functions)
International        3 (1.3%)    11 (1.5%)    21 (4.4%)     82 (4.9%)
National            19 (8.1%)    45 (6.2%)    34 (7.2%)    138 (8.2%)
Province             5 (2.1%)    23 (3.2%)     9 (1.9%)     29 (1.7%)
Regional            17 (7.2%)    12 (1.7%)     4 (0.8%)     15 (0.9%)
City               191 (81.3%)  635 (87.5%)  407 (85.7%)  1415 (84.3%)

                   1909-11      1773-92      1824-30      1867-68
                   Opera        Drama        Drama        Drama
                   N = 677      N = 157      N = 253      N = 115

Administration       2           15            4            2
Civil                5           13           24           20
Cultural             2                         1            2
Economic            29           31           22           23
Education           38                        17           12
Financial           49           11           46           53
Juridical           --            4            1            1
Church              62           13           33           23
Social              25                         4           24
Military            22           45           24           18
Political           29           32           47           22
Charitative         42            3           72           52

Sum functions      305          169          295          252
Sum persons:       137 (20.2%)   68 (43.3%)   99 (39.1%)   71 (61.7%)
Functions/persons    2.2          2.5          3.0          3.5
Functions/cohort     0.5          1.1          1.2          2.2

Function Level (sum total of functions)
International       22 (7.2%)     3 (1.8%)     6 (2.0%)    12 (4.8%)
National            22 (7.2%)     5 (3.0%)    12 (4.4%)    19 (7.5%)
Province             3 (1.0%)     3 (1.8%)     4 (1.4%)     4 (1.6%)
Regional             2 (0.7%)    15 (8.9%)     3 (1.0%)     3 (1.2%)
City               256 (83.9%)  143 (84.6%)  269 (91.2%)  214 (84,9%)

                   1890-94      1902-12
                   Drama        Drama
                   N = 582      N = 1476

Administration       2            3
Civil               17           40
Cultural             8            7
Economic            48           62
Education           62           94
Financial          120          152
Juridical                         2
Church              58          115
Social              53           72
Military            55           59
Political           38           73
Charitative         95          113

Sum functions      556          792
Sum persons:       230 (39.5%)  353 (23.9%)
Functions/persons    2.4          2.2
Functions/cohort     1.0          0.5

Function Level (sum total of functions)
International       21 (3.8%)    49 (6.2%)
National            40 (7.2%)    54 (6.8%)
Province             5 (0.9%)    18 (2.3%)
Regional             1 (0.2%)     7 (0.9%)
City               489 (87.9%)  664 (83.8%)

Table 5A Sociability

                    1789-92      1817-42      1850-59      1860-82
                    Opera        Opera        Opera        Opera
                    N = 120      N = 276      N = 134      N = 1150

Cultural             17          256          173          1239
Cultural/political   13            3            7            61
Sociability          31          185           85           344
Literary             31           21           14            54
Social                5           74           75           786
Musical              10           12            6           164
Paining/drawing      --            9            2             1
Sports               --            4            2            22
Theatre                            2           --             3
Scientific            2           11           11            74
Sum Membership:     109          577          375          2748
Sum persons:         56 (46.7%)  196 (71.0%)  112 (83.6%)   875 (76.1%)
Membership/persons    1.9          2.9          3.3           3.1
Membership/cohort     0.9          2.1          2.8           2.4

                    1909-11      1773-92      1824-30      1867-68
                    Opera        Drama        Drama        Drama
                    N = 677      N = 157      N = 253      N = 115

Cultural            336           4           136          151
Cultural/political   18           4             7           --
Sociability          68          20            61           46
Literary              4          29             7            8
Social              135           1            24           90
Musical             186          --             4           11
Paining/drawing      --           1             4           --
Sports                2          --             1            6
Theatre               1          --             1            2
Scientific           32           3             3           11
Sum Membership:     782          40           248          325
Sum persons:        297 (43.9%)  62 (39.5%)   104 (41.1%)  101 (87.8%)
Membership/persons    2.6         1.6           2.4          3.2
Membership/cohort     1.2         0.3           1.0          2.8

                    1890-95       1902-12
                    Drama         Drama
                    N = 582       N = 1476

Cultural             657           960
Cultural/political    52            57
Sociability          123           172
Literary              21            24
Social               335           390
Musical              157           384
Paining/drawing       --            --
Sports                 3             9
Theatre               --            10
Scientific            60            78
Sum Membership:     1408          2084
Sum persons:         443 (76.1%)   760 (51.5%)
Membership/persons     3.2           2.7
Membership/cohort      2.4           1.4

Table 6A Education and Political Allegiance after 1860

                             1860-82      1909-11      1867-68
                             Opera        Opera        Drama
                             N = 1150     N = 677      N = 115

Sum Higher Education         106 (9.2%)    59 (8.7%)     6 (5.2%)
Sum Gymnasium related         57 (5.0%)    13 (1.9%)     7 (6.1%)
Sum 5-Burgherschool related   15 (1.3%)    31 (4.6%)    --
Sum 3-Burgherschool related   --           15 (2.2%)    --
Teacher Training               3 (0.3%)    35 (5.2%)     1 (0.9%)
Boarding School                6 (0.5%)     4 (0.6%)    --
Different/Not relevant        10 (0.9%)    --            1 (0.9%)
Sum Total                    197 (17.1%)  157 (23.2%)   15 (13.0%)
Unknown                      953 (82.9%)  520 (76.8%)  100 (87.0%)

Political Allegiance
Conservative Election Club     9 (0.8%)    --            2 (1.7%)
Liberal Election Club        175 (15.2%)   20 (3.0%)    27 (23.5%)
Christian Democrats            4 (0.6%)    --           --
Not Relevant                   9 (0.8%)    --            1 (0.9%)
Unknown                      953 (82.9%)  653 (96.5%)   85 (73.9%)

                             1890-94      1902-12
                             Drama        Drama
                             N = 582      N = 1476

Sum Higher Education          57 (9.8%)     99 (6.7%)
Sum Gymnasium related         26 (4.5%)     38 (2.6%)
Sum 5-Burgherschool related   32 (5.5%)     74 (5.0%)
Sum 3-Burgherschool related   15 (2.6%)     62 (4.2%)
Teacher Training               3 (0.5%)     21 (1.4%)
Boarding School                1 (0.2%)      3 (0.2%)
Different/Not relevant         1 (0.2%)      4 (0.3%)
Sum Total                    135 (23.2%)   304 (20.6%)
Unknown                      477 (76.8%)  1172 (79.4%)

Political Allegiance
Conservative Election Club    --    --
Liberal Election Club         64 (11.0%)    56 (3.8%)
Christian Democrats           --             6 (0.4%)
Not Relevant                   1 (0.2%)      4 (0.3%)
Unknown                      517 (88.8%)  1410 (95.5%)

Table 7A Occupation per Generation

                           Middle     Aca-       Profes-
                           Trade      class      demics     sionals
                         N    %      N    %     N    %     N     %

Opera 1817-1844
subscriber               169  61.9   13   4.7   43  15.6   35  12.7
father                   105  59.0   15   8.4   14   7.9   24  13.5
grandfather               72  72.7    4   4.0    4   4.0    6   6.1
great-grandfather         23  67.6    2   5.9   --  --      5  14.7

Drama 1824-1830
subscriber               141  55.7   40  15.8   16   6.3   39  15.4
father                    85  47.8   36  20.2    8   4.5   17   9.6
grandfather               39  50.0    8  10.3    7   9.0    3   3.8

Opera, 1860-1882
subscriber               628  54.6  176  15.3  100   8.7  142  12.3
father                   318  57.3  106  19.1   30   5.4   78  14.0
grandfather              158  52.8   54  18.1   12   4.0   25   8.4
great-grandfather         74  71.8    3   2.9    3   2.9    8   7.8
great-great-grandfather   27  67.5    2   5.0    5  12.5   --  --

Tivoli, 1890-1895
subscriber               304  52.2   83  14.3   56   9.6   89  15.3
father                   160  49.1   80  24.5   22   6.7   48  14.7
grandfather               84  43.1   57  29.2    5   2.5   21  10.1
great-grandfather         43  64.2    9  13.4    1   1.5    2   3.0
great-great-grandfather   17  56.7    3  10.0    2   6.7    3  10.0

Drama, 1902-1912
subscriber               594  40.2  242  16.4   83   5.6  331  22.4
father                   357  44.8  233  29.2   26   3.3  129  16.2
grandfather              184  39.0  154  32.6   11   2.3   68  14.4
great-grandfather         84  60.0   20  14.3    7   5.0    5   3.6
great-great-grandfather   40  71.4    2   3.5    4   7.0    2   3.5

                         Workers    None      Sum
                         N    %     N    %     N

Opera 1817-1844
subscriber               --  --      16   5.8   276
father                   --  --      20  11.2   178
grandfather              --  --      13  13.1    99
great-grandfather         1   2.9     3   8.8    34

Drama 1824-1830
subscriber               --  --      17   6.7   253
father                    4   2.2    28  15.7   178
grandfather              --  --      21  26.9    78

Opera, 1860-1882
subscriber                1   0.1   103   9.0  1150
father                    9   1.6    14   2.5   555
grandfather              18   6.0    32  10.7   299
great-grandfather         1   1.0    14  13.6   103
great-great-grandfather   1   2.5     5  12.5    40

Tivoli, 1890-1895
subscriber                4   0.7    46   7.9   582
father                   10   3.1     6   1.8   326
grandfather              21  10.8     7   3.6   195
great-grandfather         5   7.5     7  10.4    67
great-great-grandfather   1   3.3     4  13.3    30

Drama, 1902-1912
subscriber               16   1.1   210  14.2  1476
father                   37   4.6    15   1.9   797
grandfather              37   7.8    18   3.8   472
great-grandfather        11   7.9    13   9.3   140
great-great-grandfather   1   1.8     7  12.3    56

Table 7B Wealth Class and Occupation per Generation. Opera, 1860-82

wealth class
occupation & generation                     percentages:
trade          1    2    3    4    5   sum  1      2       3     4

gen. 1          1   44   98  133  238  514   0.1   4.9    10.9  14.8
gen. 2         --    1   93   79   73  246  --     0.3    23.9  20.3
gen. 3         --    1   43   49   22  115  --     0.6    24.9  28.3
gen. 4         --   --   10   12    8   30  --     --     27.8  33.3
gen. 5         --   --   --    1    2    3  --     --     --    25

middle class    1    2    3    4    5  sum  percentages

gen. 1          1   28   69   42   16  156   0.1   3.1     7.7   4.7
gen. 2          1    6   57    5    1   70   0.3   1.5    14.7   1.3
gen. 3         --    5   18    2   --   25  --     2.9    10.4   1.6
No data for generations 4 and 5

academics       1    2    3    4    5  sum  percentages

gen. 1          1   10   15   24   37   87   0.1   1.1     1.7   2.7
gen. 2         --   --    5   10   12   27  --    --       1.3   2.6
gen. 3         --   --    1    3    2    6  --    --       0.6   1.7
gen. 4         --   --   --   --   --   --  --    --      --    --
gen. 5         --   --   --    1   --    1  --    --      --    25.0

professionals   1    2    3    4    5  sum  percentages

gen. 1          1   37   39   18    8  103   0.1   4.1     4.3   2
gen. 2         --    3   30    7    1   41  --     0.8     7.7   1.8
gen. 3         --    5    7    1   --   13  --     2.9     4.1   0.6
gen. 4         --    1    1   --    1    3  --     2.8     2.8  --
No data for generation 5

work. class     1    2    3    4    5  sum  percentages

gen. 1         --    1   --   --   --    1  --     0.1    --    --
gen. 2         --    1   --   --   --    1  --     0.3    --    --
No data for generations 3, 4, and 5

none            1    2    3    4    5  sum  percentages

gen. 1          1    5    7   11   20   44   0.1   0.6     0.8   1.2
gen. 2         --   --    2    1    1    4  --    --       0.5   0.3
gen. 3         --    1    5    4    3   11  --     0.6     2.9   1.2
gen. 4         --   --    1   --    2    3  --    --       2.8  --
No data for generation 5

Sum             1    2    3   4    5   sum  percentages

gen. 1          5  125  228  228  319  905   0.6  13.8    25.2  25.0
gen. 2          1   11  187  102   88  389   0.3   2.8    48.1  26.2
gen. 3         --   12   74   57   27  170  --     7.1    43.5  33.3
gen. 4         --    1   12   12   11   36  --     2.8    33.3  33.3
gen. 5         --   --   --    2    2    4  --    --      --    50

               percentages:
trade          5      sum

gen. 1         26.4   56.8
gen. 2         18.8   63.2
gen. 3         12.7   66.5
gen. 4         22.2   83.3
gen. 5         50     75

middle class   percentages

gen. 1          1.8   17.3
gen. 2          0.3   18
gen. 3         --     14.5
No data for generations 4 and 5

academics      percentages

gen. 1          4.1    9.7
gen. 2          3.1    6.9
gen. 3          1.2    3.5
gen. 4         --     --
gen. 5         --     25.0

professionals  percentages

gen. 1          0.9   11.4
gen. 2          0.3   10.5
gen. 3         --      7.5
gen. 4          2.8    8.3
No data for generation 5

work. class    percentages

gen. 1         --      0.1
gen. 2         --      0.3
No data for generations 3, 4, and 5

none           percentages

gen. 1          2.2    4.9
gen. 2          0.3    0.8
gen. 3          1.7    6.3
gen. 4          5.6    8.3
No data for generation 5

Sum            percentages

gen. 1         35.2  100
gen. 2         22.6  100
gen. 3         15.9  100
gen. 4         30.6  100
gen. 5         50    100

Table 7C Wealth Class and Occupation per Generation. Drama, 1902-12

wealth class
occupation & generation                      percentages
occupation      1    2    3    4    5  sum   1      2       3     4

gen. 1          9   44  130   87  282   553   0.7   3.5    10.4   7.0
gen. 2          2   43   39   58  100   242   0.4   9.2     8.4  12.5
gen. 3          1    2   33   42   33   122   0.4   0.8    13.4  17.1
gen. 4         --    1   16   20   15    52  --     1.4    22.2  27.8
gen. 5         --   --    3    5    4    12  --    --      20.0  33.3

middle class    1    2    3    4    5   sum  percentages

gen. 1         18   48   82   37   36   221   1.4   3.8     6.6   3.0
gen. 2          7   70   41   18    2   138   1.5  15.1     8.8   3.9
gen. 3          2   13   68    1   --    84   0.8   5.3    27.6   0.4
gen. 4         --    2    7   --   --    10  --     4.2     9.7  --
No data for generation 5

academics       1    2    3    4    5   sum  percentages

gen. 1          2    5   14   19   33    73   0.2   0.4     1.1   1.5
gen. 2         --    5    5    3    5    18  --     1.1     1.1   0.6
gen. 3         --   --    3    2    2     7  --    --       1.2   0.8
gen. 4         --   --    3    2   --     5  --    --       4.2   2.8
No data for generation 5

professionals   1    2    3    4    5   sum  percentages

gen. 1         39   91   92   33   20   275   3.1   7.3     7.4   2.6
gen. 2          2   43    8    6    2    61   0.4   9.2     1.7   1.3
gen. 3          1    7   13    2   --    23   0.4   2.8     5.3   0.8
gen. 4         --    1   --   --   --     1  --     1.4    --    --
gen. 5         --   --   --    1   --     1  --    --      --     6.7

working class   1    2    3    4    5   sum  percentages

gen. 1          4   --    4   --   --     8   0.3  --       0.3  --
gen. 2         --    3   --   --   --     3  --     0.6    --    --
gen. 3          2    3   --   --   --     3   0.8   1.2    --    --
No data for generations 4 and 5

none            1    2    3    4    5   sum  percentages

gen. 1         13   26   29   22   28   118   1.0   2.1     2.3   1.8
gen. 2         --   --   --    1    2     3  --    --      --     0.2
gen. 3         --   --    3    1    1     5  --    --       1.2   0.4
gen. 4         --    1    1    1    1     4  --     1.4     1.4   1.4
gen. 5         --   --    1   --    1     2  --    --       6.7  --

Sum             1    2    3    4    5   sum  percentages

gen. 1         85  214  351  198  399  1247   6.8  17.2    28.1  15.9
gen. 2         11  164   93   86  111   465   2.4  35.3    20.0  18.5
gen. 3          6   25  131   48   36   246   2.4  10.2    53.2  19.5
gen. 4         --    6   27   23   16    72  --     8.3    37.5  31.9
gen. 5         --   --    4    6    5    15  --    --      26.7  40.0

               percentages
occupation     5      sum

gen. 1         22.6   44.2
gen. 2         21.5   52.0
gen. 3         13.4   49.6
gen. 4         20.8   72.2
gen. 5         26.7   80.0

middle class   percentages

gen. 1          2.9   17.7
gen. 2          0.4   29.7
gen. 3         --     34.1
gen. 4         --     13.9
No data for generation 5

academics      percentages

gen. 1          2.6    5.9
gen. 2          1.1    3.9
gen. 3          0.8    2.8
gen. 4         --      6.9
No data for generation 5

professionals  percentages

gen. 1          1.6   22.1
gen. 2          0.4   13.1
gen. 3         --      9.3
gen. 4         --      1.4
gen. 5         --      6.7

working class  percentages

gen. 1         --      0.6
gen. 2         --      0.6
gen. 3         --      2.0
No data for generations 4 and 5

none           percentages

gen. 1          2.2    9.5
gen. 2          0.4    0.6
gen. 3          0.4    2.0
gen. 4          1.4    5.6
gen. 5          6.7   13.3

Sum            percentages

gen. 1         32.0  100
gen. 2         23.9  100
gen. 3         14.6  100
gen. 4         22.2  100
gen. 5         33.3  100

Table 7D Wealth Class and Occupation per Generation. Melodrama cohort,
1824-1830

wealth class (rateable value)
occupation & generation
trade                     1      2      3     4     5    sum

gen. 1                    --     1     53    44    20    118
gen. 2                    --    --     15    19     8     42
gen. 3                    --    --     --     3     2      5

middle class              1      2      3     4     5    sum

gen. 1                    --     1     18    12     2     33
gen. 2                    --     1     12     3    --     16
gen. 3                    --    --      1    --    --      1

academics                 1      2      3     4     5    sum

gen. 1                    --    --      3     5     2     10
gen. 2                    --    --      2     4    --      5
gen. 3                    --    --      1    --    --      1

professionals             1      2      3     4     5    sum

gen. 1                     1     1     24     1    --     27
gen. 2                     1    --      3    --     1      5
No data for generation 3

work. class               1      2      3     4     5    sum

no data

none                      1      2      3     4     5    sum

gen. 1                    --    --      2     3     1      6
gen. 2                    --    --      2    --    --      2
No data for generation 3

Sum                       1      2      3     4     5    sum

gen. 1                     1     3    100    63    25    192
gen. 2                     1     1     34    26     9     71
gen. 3                    --    --      2     3     2      7

MTB                       0      3.-    6.-  15.-  30.-  sum

gen 2                      1    10     10     5    14     40
gen 3                     --    10      7     1    11     29

                          percentages:
trade                     percentages

gen. 1                    --     0.5  27.6  22.9  10.4   61.4
gen. 2                    --    --    21.1  26.8  11.3   59.2
gen. 3                    too few data to be relevant

middle class              percentages

gen. 1                    --     0.5   9.4   6.3   1.0   17.2
gen. 2                    --     1.4  16.9   4.2  --     22.5
gen. 3                    too few data to be relevant

academics                 percentages

gen. 1                    --    --     1.6   2.6   1.0    5.2
gen. 2                    --    --     2.8   5.6  --      7.0
gen. 3                    too few data to be relevant

professionals             percentages

gen. 1                     0.5   0.5  12.5   0.5  --     14.1
gen. 2                     1.4  --     4.2  --     1.4    7.0
No data for generation 3       no data

work. class               percentages

no data

none                      percentages

gen. 1                    --    --     1.0   1.6   0.5    3.1
gen. 2                    --    --     2.8  --    --      2.8
No data for generation 3             No data

Sum                       percentages

gen. 1                     0.5   1.6  52.1  32.8  13.2  100
gen. 2                     1.4   1.4  47.9  36.6  12.7  100
gen. 3                    too few data to be relevant

MTB                       percentages

gen 2                      2.5  25.0  25.0  12.5  35.0  100
gen 3                     --    34.5  24.1   3.4  37.9  100

Table 4B Average number of functions per cohort per wealth class (19th.
c. only)

wealth class     1      2     3    4    5     mean
cohort

Opera 1817-44    --     --    1.4  2.9  4.7   2.9
'Opera' 1850-59  --     1.6   2.3  2.2  4.8   3.5
Opera 1869-82    0.2    0.6   0.6  1.5  3.1   1.7
Opera 1909-11    0.04   0.08  0.2  0.3  1.5   0.5
Drama 1824-30    --     --    0.6  1.7  4.2   1.4
Drama 1867-68    --     0.1   0.9  1.6  4.2   2.3
Drama 1890-94    0.2    0.5   0.8  1.1  1.6   1.0
Drama 1902-12    0.04   0.1   0.2  0.5  1.4   0.6

Table 5B Average number of memberships per cohort per wealth class
(19th. c. only)

wealth class     1    2    3    4     5    mean
cohort

Opera 1817-44    --   --   1.2  2.1  3.5   2.2
'Opera' 1850-59  --   1.6  1.9  2.8  3.6   3.0
Opera 1869-82    1.2  1.2  1.6  2.8  3.9   2.7
Opera 1909-11    0.2  0.3  0.8  1.3  3.2   1.4
Drama 1824-30    --   --   0.8  1.3  2.7   1.2
Drama 1867-68    --   --   2.2  2.7  4.3   3.0
Drama 1890-94    0.6  1.4  2.1  2.7  3.9   2.6
Drama 1902-12    0.1  0.4  0.8  1.6  3.3   1.6

Table 1C Number of women subscribers and percentage in the cohorts

Opera    number  percentage  Drama    number  percentage

1789-92     2     1.7%       1773-92     8      5.1%
1817-44    --    --          1824-30     3      1.2%
1860-82    48     4.1%       1867-68    --     --
1909-11   106    15.7%       1878-00    20      5.5
                             1890-95    43      7.4
                             1902-12   228     15.4

Table 1D Number and potential ratio of accompanying women to male
subscribers

Opera    number  male-female ratio  Drama    number  male-female ratio

1789-92    30        .25            1791-92    6         .24
1817-20    40       1.08            1824-30   16         .6
1836-38   195        .94

Table 2B Proportional distribution of age-classes to rank, Van Eysden,
1902-1912

rank         1     2     3     4     5
age-class

11-20       1.5   4.7   1.9   7.2   8.3
21-30      13.9  21.1  13.3  18.3  11.0
31-40      19.6  21.6  25.4  25.6  29.7
41-50      27.8  25.4  36.0  19.8  31.0
51-60      24.1  21.8  15.7  19.1  14.5
61-70      10.9   4.3   7.0   7.9   5.5
71-80       2.3   1.0   0.7   2.1  --

Table 4C Rank, Wealth Class and Function Categories

         German Opera, 1860-1882      Drama, Van Eysden, 1902-1912
class    1   2   3    4    5    sum   1   2   3   4   5    sum

rank 1   --   3    3   72  488   566   1   3  50  78  548  680
rank 2   --   6   20   65  186   277  --   9  16  12    7   44
rank 3   --  36   88   96  161   381   2   4   4   3    1   14
rank 4   not available                --   5   3  --   --    8
rank 5   not available                --  --  --  --   --   --
missing   1  24   25   97  141   288
Sum       1  69  136  330  976  1512   3  21  73  93  556  746

Table 5C Rank, Wealth Class and Sociability

         German Opera, 1860-1882       Drama, Van Eysden, 1902-1912
class    1    2    3    4    5    sum   1   2   3    4    5     sum

rank 1   --    3   16  125  597   741   4  13  133  257  1270  1677
rank 2   --   14   71  112  213   410   4  27   82   41    36   190
rank 3    4   71  193  168  203   639   1   8   14   21    --    44
rank 4   not available                  2  25   23    6     1    57
rank 5   not available                  1   5    2   --    --     8
missing   2   68   80  227  221   598
Sum       6  156  360  632  696  2388  12  78  254  325  1307  1976

Table 6B Percentages of higher education data per rank and per wealth
class

cohort   Tivoli  Eysden  Maifest           Tivoli  Eysden  Maifest
rank                              class

rank 1   30.2    26.2             class 1    4.2     4.7    15.7
rank 2   10.2    13.2             class 2   21.7    14.4    19.6
rank 3    4.3    12.8             class 3   17.1    17.9    22.3
rank 4            0.0             class 4   28.8    27.8    26.9
unknown  12.5             25.0    class 5   30.3    28.3    35.3

Drama                               Opera
tax sum        3.-  6.-  15.-  30.-  3.-  6.-  15.-  30.-

trade           5    2    6    74     1   --    6    71
middle class   --   --    1    --    --   --   --    --
academics      --   --   --     5    --   --   --     4
professionals  --    1    3     6    --   --   --     5
none/unknown   --    1   --     7    --   --   --     3
Sum             5    4   10    92     1   --    6    83

ENDNOTES

The research was subsidized by the Dutch Science Foundation (NWO). The organization of the work was as follows. Henk Gras did the historical research; Harry van Vliet developed the software applications for data gathering and analysis and surveyed the statistics. We especially thank Bennie Pratasik for his assistance. He did not live to see this result. We dedicate this essay to his memory.

1. Henk Gras and Philip Hans Franses, "Theater Going in Rotterdam, 1802-1853. A Statistical Analysis of Ticket Sales," Theater Survey, 39, (1998), 73-98; Henk Gras, Philip Hans Franses, Marius Ooms, "Did Men of taste and Civilization Save the Stage? Theater-Going in Rotterdam, 1860-1916. A Statistical Analysis of Ticket Sales," in: Journal of Social History, 36:3 (2003), 615-655.

2. Gras, Franses, and Ooms, "Did Men of Taste and Civilization Save the Stage" (2003), 624.

3. Michel Vovelle, "Histoire des Mentalites, histoire des resistances, ou les prisons de longue duree," in: Le monde Alpin et rhodanien (1980). We consulted the Dutch translation in: Michel Vovelle, Mentaliteitsgeschiedenis (Nijmegen, 1985), pp. 142-169.

4. See for the traditional view particularly B. Hunningher, Een eeuw Nederlands toneel (Amsterdam, 1949) [A Century of Dutch Theatre], pp. 12-13; the same, Het dramatisch werk van Schimmel in verband met het Amsterdamsche tooneelleven in de negentiende eeuw (Amsterdam, 1931) [The Dramatic Work of Schimmel, as related to Amsterdam Theatre Culture in the Nineteenth Century], p. 3. Also B. Albach, Helden, draken en komedianten (Amsterdam, 1956), [Heroes, Blood and Thunder Plays, and Comedians], p. 141; H.H.J. de Leeuwe, "De geschiedenis van het Amsterdamse toneelleven in de negentiende eeuw (1795-1925)," in: A,E. d'Ailly, ed., Zeven eeuwen Amsterdam, 6 Vols. (Amsterdam, n.d. [1948]) ["The History of Amsterdam Theatre Life, 1795-1925," in: Seven Centuries Amsterdam], V, pp. 6 and 12, but see also the same, "Silepochen des hollandisches Theaters im 19. Jahrhundert," in: Maske und Kothurn, 3:4 (1957), 340-62; and the same, "Antoine Jean le Gras, een Nederlands regisseur der negentiende eeuw in Rotterdam," in: Rotterdams Jaarboekje ["Antoine Jean le Gras, a Dutch director in Nineteenth Century Rotterdam"] (1975), 209-256, see p. 211. S. Koster, Comedie in Gelderland (Zutphen, 1997).

5. De Leeuwe, "Antoine Jean le Gras" (1975), 221.

6. Jim Davis and Victor Emeljanow, Reflecting the Audience. London Theatregoing, 1840-1880 (Iowa City, 2001), pp. 97-98. The works in question are Allardyce Nicoll's A History of English Drama, 1660-1900, Vol. 4, Early Nineteenth-Century Drama 1800-1850 (Cambridge, 1970) and Vol. 5, Late Nineteenth-Century Drama 1850-1900 (Cambridge, 1946). George Rowell, The Victorian Theatre 1790-1914 (Cambridge, 1978).

7. Davis and Emeljanow, Reflecting the Audience (2001), 98.

8. Davis and Emeljanow, Reflecting the Audience (2001), 98.

9. More likely, he gave up when the duke of Weimar's mistress, the actress Caroline Jagermann was given control over the company. The Weimar theatre in its heyday staged three plays per week, of which one was a musical play, and another a Von Kotzebue of Iffland play. See Marvin Carlson, Goethe and the Weimar Theatre (Ithaca, 1978), John Prudhoe, The Theatre of Goethe and Schiller (Oxford, 1973).

10. Bruce A. McConachie, Melodramatic Formations. American Theatre & Society, 1820-1870 (Iowa City, 1992).

11. McConachie, Melodramatic Formations. (1992), xii.

12. Tony Bennett, "Texts in History: the Determinations of Readings and Their Texts," in Derek Attridge, et. al., eds. Poststructuralism and the Question of History (Cambridge, 1987). Particularly David Morley's work on audience-supply relations, written in the Birmingham Centre of Cultural Studies tradition, has done much to reject such schematic relations. See for an overview, David Morley, Television, Audiences, & Cultural Studies (London and New York, 1992). MacConachie plays with vague concepts such as 'Paternal', 'patriarch(al)', (unspecified) 'elite', 'patrician', 'gentry', 'urban yeomen', most of which terms do not define 'class' in a comparative way. Conclusions are derived from three plays. In view of the huge repertoire this cannot be considered a representative analysis. The one piece of visual evidence offered with respect to melodrama audiences, is not helpful, for, just after the reader has been told that the early nineteenth-century theatres stored the 'elite' in boxes, above the 'pitties', McConachie explains the plate on p. 10, to show "many of the New York male elite (...) seated in the pit" [our italics].

13. We are not, of course, the first to criticise the traditional view. Hennie Ruitenbeek "Byna te arm tot schouwburgtydverdryf': Sociale geledingen en levensstandaard van het publiek in de Amsterdamse Stadsschouwburg in de eerste helft van de negentiende eeuw," in: Economisch-Sociaalhistorisch Jaarboek, 56 (1993), 94-149 ["Almost too poor for theatre-going. Social Layers and Standard of Life of the Audience of the Amsterdam City Theatre"], concluded that it was unlikely that working classes could afford a theatre ticket. Davis and Emeljanow, Reflecting the Audience (2002) criticised the narrative of decline and resurrection of quality in the London theatre.

14. Election clubs were formed from 1848 onwards. It took till about 1870 before non-voters were allowed in election clubs in Rotterdam.

15. Hunningher, Century of Dutch Stage (1949), 26-27; De Leeuwe, "Antoine Jean le Gras" (1975), 211.

16. Pierre Bourdieu, transl. Richard Nice, Distinction. A Social Critique of the Judgement of Taste (London, 1989), originally published in 1979.

17. These shareholders never got free entrance, but had the right of pre-subscribing for seats. As the minutes make clear, most of them did so for the balcony, boxes, or the pit. We refer to this cohort as pseudo-French Opera, 1852-1859. The number of season ticket holders at the French Opera was considerably higher than the 134 shareholders (more like 300 to 400).

18. An analysis of poll-tax in the years 1860 of about 350 members of the literary society Amicable Literary Exercise, which society consisted half of white collar workers and half of shopkeepers and artisans, showed that more than half of them could spend some fl. 500.- to fl. 1,250.- per year (lowest tax scales), while a considerable number of them were below this threshold (GAR, VLO, membership lists and minutes). Herman Heijermans, journalist for the liberal New Rotterdam Journal regarded a weekly income of fl. 10.- per week (too) high for a working class labourer (Sunday Journal, 5-12-1868).

19. We lack, for instance, membership lists of the Society for Public Welfare, Harmony, and the earliest membership lists of the Society for the Advancement of Music.

20. Higher education as defined in the tables include the university, the medical schools, the technical high-schools, and the high-school for trade. All of these forms of education now have a university status.

21. This is curious in a way. Orangism after 1787 was, of course, dominated by the aristocrat regents, but the Orangist 'party' itself, its street fighters so to say, consisted for a large part of the classes, which the traditional theatre historians defined as melodrama audience.

22. We remind readers that the cohorts before ca. 1850 appear to be slightly out of line because of the different measure (rateable value instead of spending or income tax).

23. We tested the relation between the distribution of wealth classes for opera and drama linking the melodrama cohort (1824-1830) to the French Opera cohort (1817-1844). This relation was significant (chi-square = 23.6; df = 2; p < 0.00--we had to ignore the first and second class of the melodrama cohort). The distribution of wealth was also significant for the German Opera (1860-1882), Tivoli (1890-1895), and Van Eysden (1902-1912). (Chi-square = 80.0; df = 8; p < 0.00).

24. The Marriage and Burial Tax (MTB) was based on wealth at the time of death. In the highest class one paid fl. 30.-; in the second class fl. 15.-; in the third fl. 6.-; in the fourth fl. 3.-. There was also a pro deo class for the poor. The highest class, however, was related to a fortune of fl. 12,000.- (it halved per tax class) and it turned out that for the theatre-goers this measure hardly differentiated wealth.

25. Opera: French Opera, 1789-92: no valid Chi-square test; French Opera, 1817-44: Chi-square = 27.6, df = 4, p < 0.00; German Opera, 1860-82: Chi-square = 537.7, df = 12, p < 0.00; Maifestspiele: no ranks known. Drama: Dutch troupes, 1773-92: no valid Chi-square test; Hoedt & Bingley, 1824-30: almost all data on one rank; Albregt & Van Ollefen: no valid Chi-square test; Tivoli: Chi-square = 59.2; df = 8; p < 0.00. Van Eysden: Chi-square = 629.5; df = 12; p < 0.00.

26. The number of functions identified is different from that in table 3a, for we did not find wealth data for all our subscribers.

27. The rapid changes in the constitution helped to increase the number of function-holders in the early nineteenth century, whereas councillors were over and again reelected in the late nineteenth century.

28. N.L. Dodde, Tot der kinderen selfs profit. Een geschiedenis van het onderwijs te Rotterdam (Den Haag, 1991), states that the proportion of pupils in the new extended education system between 1865 and 1900 was negligibly small (p. 217), and estimated that only 8 per cent of the city population could even consider sending their children to such schools (p. 231). He calculated that 2 per cent of the boys of 12-17 years of age visited the gymnasium (1825-1880); 3 per cent of that age group visited the HBS-5 at the end of the century, and 2 per cent of that age group visited the HBS-3 about 1880 (pp. 229, 234, 241).

29. See for an overview of the twentieth-century surveys: H. Ganzeboom, Cultuurdeelname in Nederland (Assen, 1989); W. Knulst, Van Vaudeville tot Video (Rijswijk, 1989); H. van Maanen, Het Nederlands toneelbestel (Amsterdam, 1997).

30. The idea was already current about 1916, when Rossing used it in his De Koninklijke Vereeniging Het Nederlandsch Tooneel (Amsterdam, 1916). Robert H. Leek, Shakespeare in Nederland (Zutphen, 1988), returns to it (p. 95).

31. The opposition 'Patriot' against 'Orangist' is in some ways misleading. In fact three parties were relevant: The Orangists, the democratic Patriots and the 'Estates-minded' regents. The best overview in English is Simon Schama, Patriots and Liberators (London, 1977).

32. For the growth and formation of the field of theatrical amusements in Rotterdam, before the construction of the first theatre (1773-74) see Henk Gras, "Wat er aan de schouwburg vooraf ging. Vorming en structuur van het veld van theatrale vermakelijkheden in Rotterdam, ca. 1630-1773," in: Rotterdams Jaarboekje (1997), pp. 152-212.

33. We had to leave out the eighteenth-century drama and opera cohorts, because they contained no Jews. Also we had to leave out the Mennonites, whose share before ca. 1870 was too small, and the category 'none' and 'unknown', since these contained empty cells. English, Scots and Reformed dissenters, were counted with the Dutch Reformed, as 'Protestants'. Chi-square = 117.7, df = 30, p < 0.00.

34. See the NRC, 1844, nr. 128 (October 5), and nr. 145 (October 22).

35. The coupon buyers of 1879 (no rank known), which subscribed for this season only, were not considered here, since rank is pertinent in this discussion.

36. Most 1817-1844 opera subscribers sat in the pit or in the boxes. Merchants were the largest group on all three ranks. Of those in the boxes (47 persons), 42 were in the fourth or fifth wealth classes, whereas 67 of 86 merchants in the pit were in the third or fourth wealth classes. The bulk of the academics (28 out of 35) sat in the pit, where 23 were in the fourth or fifth fifth wealth classes. The weight of the middle class occupations and professionals were in the third wealth class (18 out of 37). There were no subscribers in wealth classes one and two. Most melodrama subscribers sat in the pit (198 out of 205). In the Tivoli Theatre so many subscribers took first-rank tickets, that the distribution was very uneven. Merchants were the largest group on all ranks. The category 'trade' showed more of a ragged pattern, but the other job categories clearly showed that the cheaper the rank, the lower the wealth class.

37. A chi-square test gave the following results: chi-square = 21.3; df = 16; p < 0.17. For this test we had to merge the first two and last two age classes.

38. Data from table 4a. Opera, 1860-1882: Chi-square = 305.9; df = 9; p < 0.00. Drama, 1902-1912: Chi-square = 298.1; df = 6; p < 0.00. We had to add up the two lowest wealth classes, and rank 3 to 5, due to empty cells, a fact that hinders such tests for other cohorts. It was also not very useful to give all results in tables, since in the older cohorts one rank often dominated (the pit). For the French Opera 1817-1844 the results were also significant, with a good distribution of data in wealth classes three to five: chi-square = 107.6; df = 4; p < 0.00.

39. Opera, 1860-1882: Chi-square = 526.8; df = 9; p < 0.00. We had to add up wealth the two lowest wealth classes. 3.2; df = 8; p < 0.00. We had to add up rank 3 to 5, due to empty cells, a fact that hinders such tests for other cohorts. It was also not very useful to give all results in table, since in the older cohorts one rank often dominated (the pit).

40. We tested the relation with a Kruskal-Wallis test: H = 12.17; df = 4; p < 0.00. We did not test the relation between rank and education, since there were no data for the Maifestspiele and an empty cell (Van Eysden, rank 5).

41. We had, of course, to use different sorts of taxes. For the opera, 1860-82 the wealth class of generation 1 was based on the poll tax (901 of 1150, valid percentage 78.3); generation 2 on rateable value (389 of 555, valid percentage 70.1); generation 3 on rateable value (173 of 299, valid percentage 57.9). For drama, 1902-12 generation 1 was based on income tax (1247 of 1476, valid percentage 84.5); generation 2 on the poll tax (465 of 797, valid percentage 84.5); generation 3 on rateable value (246 of 472, valid percentage 52.1).

42. The data are derived from Table 7a (number of occupations per job category). Opera 1860-82: chi-square = 45.5; df = 8; p < 0.00; drama, 1902-12: chi-square = 255.1; df = 10; p < 0.00. Tivoli 1890-95: chi-square = 98.5; df = 10; p < 0.00. For opera we had to leave out the category 'working class'.

43. The data are derived from tables 7b and 7c (sums of wealth classes). Drama, 1902-12: Chi-square = 176.8; df = 8; p < 0.00. Opera, 1860-82: Chi-square = 114.8; df = 8; p < 0.00. Due to the small sums in class 1, the results for the opera cohort are just valid (20% of the expected values were < 5). For this cohort we had to leave out the working class (empty cells). We did not produce a table for drama, 1890-95 for sake of space.

44. Opera-fathers: 49%; drama-fathers: 62%, drama-grandfathers: 52%, only with opera-grandfathers the fourth class dominates with 61% over the fifth class.

45. Opera-subscribers: 9%, opera-fathers: 13%, opera-grandfathers: 25%; drama-subscribers: 10%, drama-fathers: 8%; drama-grandfathers: 18%.

46. Only in the great-grandfather generation we reach the political struggles of the old regime: 25 great-grandfathers were Patriots, against 3 Orangists (140 persons).

47. Vovelle, "Histoire des Mentalites, histoire des resistances, ou les prisons de longue duree," (1980).

48. Gras and Franses, "Theatre-going in Rotterdam, 1802-1853" (1998), and Gras, Franses, and Ooms, "Did Men of Taste and Civilization Save the Stage" (2003).

49. Present-day cultural economics from T. Moore, "The Demand for Broadway Theater Tickets" [1966], reprinted in M. Blaug, The Economics of the Arts (London, 1976), pp. 227-242, reckon with price-inflexibility, which is to say that patrons are prepared to go to shows regardless of the price. This should not be taken for granted in early modern times, where particularly subscribers could violently react to a rise of entrance prices.

50. A start is Hans van Maanen, Het Nederlands toneelbestel (1997).

51. See for the attempts to relate the diverse forms of direct city taxes to income, Boudien de Vries, Electoraat en elite. Sociale structuurmobiliteit in Amsterdam, 1850-1895 (Amsterdam, 1986), and Henk Visscher's critique "'Een goede waterweg en elastieke financien'. Een speurtocht naar de optimale belastingreeks voor het welvaartsonderzoek in de tweede helft van de negentiende eeuw," in: Rotterdams Jaarboekje (Rotterdam, 1999), pp. 312-336.

52. Dutch-spoken drama, 1773-92: rateable values (tax registers 1803, 1826, 1833), or on the on the Marriage and Burial tax (abolished in 1810); Dutch-spoken drama, 1824-1830: rateable values (tax-registers 1803, 1822, 1826, 1833, and registers of real property, 1832, 1848); Dutch-spoken drama, 1867: city poll tax (assessment lists 1856, 1859, 1865, 1869, 1880); Tivoli, 1890-95 and Van Eysden, 1902-1912: city income tax (assessment lists, 1893, 1902, 1905, 1908, 1912). French Opera 1789-92: rateable values (tax registers 1803, 1826, 1833), and the Marriage and Burial tax; French Opera, 1817-43: rateable values (tax registers 1803, 1826, 1833 and registers of real property, 1832, 1848); German Opera, 1860-1882: city poll tax (assessment lists 1856, 1859, 1865, 1869, 1880); Maifestspiele 1909, 1911: city income tax (assessment lists 1893, 1902, 1905, 1908, 1912).

53. There is, though, a brochure from 1840, signed by some fifty members, while also all committee members are known.

54. These clubs were also more prone to donate their archives to the Municipal Archive, which survived the 1940 bombing.

By Henk Gras and Harry van Vliet

Research Institute of Culture,

University of Utrecht

Kromme Nieuwe Gracht 29

NL-3512 HD Utrecht

Netherlands