Ancient History & Civilisation



1.1 Roman house and status display

The social flux in the Roman world around the time of the eruption of Vesuvius in 79 CE increased both the need and the means to demonstrate status. Any rise in legal status was slow, particularly for groups such as former slaves, but luxury and public consumption offered methods to display success in other areas of life.1 One possibility was to construct luxurious dwellings, and the Roman house is known to have been a significant means to demonstrate its owner’s identity, including wealth and status.2 However, this was not reserved only for the upper class and rich, and the archaeological material reveals how the sub-elites also utilized their houses for their identity building.

This book examines how Pompeian peristyle gardens were used to represent the house owner’s socioeconomic status. In the Roman world, a peristyle was a colonnaded courtyard, often featuring a garden.3 This is the first examination and comparative analysis of all 252 Pompeian peristyle gardens excavated in Pompeii. The comprehensive approach permits an understanding of the different levels of wealth and social status that were transmitted by these colonnaded spaces throughout the entire city.

Pompeii has always been considered one of the main sources of information about so-called daily life in the Roman world. The city has been interpreted to reflect something ordinary compared to, for instance, the political history of Rome. Research carried out on Pompeii concentrates on the private dwellings, which are, indeed, the resource through which the city contributes most to our knowledge of antiquity. However, such scholarship has primarily focused on the largest and most decorated houses. Additionally, other buildings such as tabernae, workshops, and even brothels have been studied to illuminate the world of the poorer strata of society,4 but the hundreds of middle-sized and small houses have been studied only occasionally and sporadically. This study clarifies the life and social interactions of the so-called Roman middle class. This group has been overshadowed in the scholarship by the highest socio-political elite. The gap between the rich and the poor was enormous, and the economic group that belonged to this middle ground was the largest in the ancient world. Our understanding of antiquity will always be partial if this mass of people is not studied in detail.

Over the last 20 years or so, scholarship has questioned the functions of the traditional room types in the Roman house, yet the peristyle curiously remains one of the spaces which is still seen to be used mainly for display purposes.5 There are several Pompeian peristyles where this is the case, but the broader picture – examining all the peristyles of the city – reveals a different situation: a vast number of peristyles were not planned or used for display purposes. I will construct a novel view of why the peristyles were built and how they were utilized.

All architecture reflects something about the socioeconomic status of the owner, as has been hypothesized by several theorists from different fields. For instance, Amos Rapoport underlines the character of architecture as a means of communicating status, power, and roles. Rapoport notes that architecture provides information about human behavior, and on the other hand also influences human behavior. He maintains that the architecture of a space was planned with a view towards its proper function, and therefore the aim is to design the space to be as well suited to the intended activity as possible.6

Pierre Bourdieu instead sees that cultural practices and preferences are related to a person’s social origin and education. This leads to the conclusion that the limits of necessity select for the most economical alternative – which can also mean the most practical alternative – whereas a taste for liberty or luxury favors conventions and tends to ignore practicality. In this view, practical solutions in domestic architecture are favored by the lower classes, particularly by people who work with their hands, as Bourdieu’s study demonstrates.7

Rapoport’s and Bourdieu’s views are the basis of my theoretical framework. Even if the function of the space is altered, it must be functionally suited to its new purpose – otherwise it would not have been selected for it – and, therefore, the qualities of a space reveal something about its use in the past. Those qualities reveal the needs of the people who used the space, and on this basis we can interpret the economical level of the inhabitants, as different levels of society had different possibilities and needs.

My ultimate aim is to examine how peristyles reflect the socioeconomic status of their owners. Several other questions must be answered before reaching this goal. First, in Chapter 3, I investigate the role of the peristyle inside the house: what was its purpose and function, and what activities took place there? Then, in Chapter 4, I move on to examine: what tools could be utilized for socioeconomic display? After defining these tools, I answer a set of questions: in what types of peristyle were these different means adopted, and how did they reflect their owner’s wealth, and how did they influence each other? Chapters 5 and 6 are built around these questions. Chapter 7 explores the connection between wealth and social status in Pompeii.

Peristyles are a part of several studies of ancient Pompeii and/or the Roman house, but oftentimes the research focus limits their examination to a few selected houses and peristyles – perhaps even choosing those that are best suited to their argument, while those which do not easily fit are ignored.8 The Roman house, which is mainly modeled in contemporary research on the basis of the writings of Vitruvius and two excavated Campanian cities, Herculaneum and Pompeii, had several important functions in Roman social life. One of these was to display acquired wealth and social status. The house provided an opportunity to do this when other areas of life were more socially constricted. For example, the clothing of the upper class male was controlled and regulated, which left little scope to display nuanced socioeconomic status.9 The house, on the other hand, opened up many possibilities to do this – although it was not free from the criticism of Roman moralists. Yet, even Cicero, who often championed himself as the supporter of traditional values, was known for the Greek ornaments that decorated his villa – something that could perhaps be considered a dangerous type of Hellenistic privata luxuria by some Roman standards.10

Scholars have examined many of the largest Pompeian houses, trying to connect them to the texts of ancient authors such as Cicero, Pliny the Younger, and – of course – most importantly Vitruvius’s De Architecura. Conclusions about the functions of the rooms were made on the basis of the architect’s descriptions, and archaeological material played a secondary role in the process. These interpreted functions provided a simplified model of the Roman house created by 19th and early 20th century archaeologists and scholars,11 and overlooked most of the domestic material in Pompeii. During the past 20 or 30 years, researchers have questioned these functions, and the ensuing deconstruction of room functions has shaped debate for decades.12 It has been noted frequently that the municipal city of Pompeii was very different from Rome, the huge capital city of the known world, making it somewhat problematic to use Roman literary sources in the Pompeian context.13 The Roman house is now often viewed as a multifunctional space; it seems that the rooms, courtyards, and gardens seldom had one clear function, although models easily give this impression. This means that socioeconomic representation was not separated from other possible functions, as is demonstrated several times in this book. Of course, having the capacity to allocate a space mainly for display purposes was a sign of wealth, but the multifunctional nature of the rooms meant that the display would be seen by several types of audiences.

Penelope Allison’s many contributions to this field of study – for example Pompeian Households: An Analysis of the Material Culture (2004) – played a significant role in the reinterpretation of the room functions of the Pompeian house. Emphasizing the analysis of the archaeological finds, she questioned the literature-based analysis. This deconstruction of the interpretation of the Pompeian house has considerably changed our view of the Roman house and their daily life, and it has left space for new reconstructions and interpretations. The more than fifteen years that has passed between the publication of the work have resulted in several excellent contributions to the study of Roman urbanism, and in many cases Pompeii has been a key source. Nonetheless, studies covering the entire city and combing through all of the material related to a single type of space have been a rarity, and for instance several important space types, such as the peristyle, have thus far remained unstudied.

Andrew Wallace-Hadrill’s Houses and Society in Pompeii and Herculaneum (1994) has been deeply influential in the study of socioeconomic representation in urban and domestic space in the Roman world. In his work, the peristyle was defined as one of the most important display spaces in the Roman house.14 Although the heyday of this explanation of the socioeconomic function of the Roman house is often considered to be the 1990s or early 2000s, it was not a new idea; similar interpretations were made already in the 19th century.15 However, the conclusion relies heavily on the writings of Vitruvius, and leaves space for an analysis of the archaeological material.

Although the archaeological material of Pompeii mostly corresponds to the year 79 CE and the time immediately associated with the eruption of Vesuvius, it can also be interpreted as reflecting the overall situation of the early Principate – or partly even the ideas and fashions of the late Republican period. The birth of the Imperial government has often been seen as representing a change in the ruling class of Rome: the Emperors nominated the equites, liberti, or even slaves to important positions in the Empire.16 These groups, historically situated below the highest senatorial elite, thus gained more status and power, which they then needed to display to society.17

The Pompeian peristyles primarily reflect something that could be called a Roman middle class, even though in this case it is better to talk about the middle classes – in plural. Although this term can be seen as anachronistic, there was a distinct group of people situated between the richest and poorest inhabitants,18 and the vast majority of the Pompeian peristyle owners belonged to this group. The Roman middle class is often neglected by scholars due to the fact that the written sources concentrate on the elite, having been mainly written by the elites themselves. Of course, the middle stratum is not entirely neglected by scholars and there are, for example, studies that focus on individual parts of it, such as freedmen.19 However, the picture is still very incomplete and more work needs to be done, and Pompeii offers a location and body of material well suited to this purpose that should not be disregarded.

The traditional understanding of the movement of ideas in the Roman world has been built according to the top-down model, where the upper levels of society produced new trends and the lower social groups passively adopted these ideas.20 This book will question that approach. Organizing the houses and their peristyles according to their architectural remains – which on a general level reflects the house owner’s wealth – demonstrates that not all types of decorations and designs can be found in the houses of the wealthiest Pompeians; rather, some means of display seem to have been developed by the middle or lower echelons of society.

Having now introduced the larger historical context and the interpretations I follow in my study, the controversial concepts of the top-down model and the middle class still require further examination and consideration before moving on.

1.2 Top-down model?

The top-down model suggests that influence in a society moves from the upper social levels to the lower. Justin Walsh compares the social elite to the fashionistas who set the trends because they have the wealth, knowledge, and leisure lifestyle to do so.21 Literary sources indicate that the Roman elite thought of itself as a role model – including in architecture – for the lower social groups, which led to competition where the elite tried to stay ahead of the lower classes and also competed against other members of the elite.22

Paul Zanker has adopted the top-down model to interpret Roman private dwellings. He believes that the upper classes dictated the trends, which were then copied by the lower echelons. In particular, Zanker interprets the nearby villae of Pompeii as major influencers on the architecture of other houses. Although Zanker focuses mainly on the non-urban villae, he also interprets large houses as “town villae,” and therefore within his context some Pompeian houses can be considered as models for imitation. Among the forms of imitation are garden architecture, fountains, sculpture, and paintings – all of which can be connected to the peristyle gardens. Zanker’s idea has been incredibly popular among scholars.23 Nonetheless, it has several problems when it is further examined.

Use of the top-down model is not limited to the classical world. It has been developed by several social theorists, such as Thorstein Veblen, whose work has often been inspirational for the study of Pompeian domestic space.24 Veblen suggests that humans display their social rank through their consumption, and he introduces the term conspicuous consumption as a means to express a high social position. According to Veblen’s theory, the leisure class – the highest level of society – influences the lower classes, who attempt to achieve the standards of the upper classes.25

As with any theory, Veblen’s ideas have also received some criticism; after all, they are more than a hundred years old. On the general level, Colin Campbell notes Veblen’s shaky evidence, the question of consciousness of conspicuous consumption, and the extent to which the action is driven by intention, instinct, or other motives. Furthermore, he notes the vague definition of the conspicuous consumption, as it has been adopted into everyday language.26 Walsh has connected conspicuous consumption with similar terms, such as costly signaling, wasteful advertising, and wasteful display.27 Signaling, advertising, and consumption are intentional activities, something not fully present in mere display, as having something on display may be unintentional. Consequently, as the level of intention is often a difficult attribute to confirm from archaeological sources, I will mostly use the term socioeconomic display, or just economic display (as the sources primarily suggest wealth or economic rank). Nevertheless, other terms such as conspicuous consumption are also utilized if they are fitting.

The top-down model has also been criticized in the Roman context. Shelley Hales notes that the efforts of the lower classes to make an impression on other members of society are frequently underestimated as imitative in the scholarship of the Roman house, and she calls for more research on the lower social strata.28 Wallace-Hadrill does not believe that the model proposed by Veblen is a likely explanation for behavior in Roman society, and posits that the motivation behind adopting new fashions was likely an urge to create distance from those who were inferior in the social hierarchy, more than merely mimicking their superiors.29 Emanuel Mayer, instead, considers that Roman art was so standardized that this in itself explains the similarity of the paintings and statues of the elite and the lower classes. In addition, he concludes that the middle class house owners may have wanted decorations similar to those of their patrons, but they were also not afraid to alter them according to their own taste.30

Katherine von Stackelberg criticizes Zanker’s top-down model in the context of gardens, and proposes that the peristyle was possibly evolved from the hortus – and, therefore, the domus architecture did not necessarily need a villa as a model for the peristyle garden.31 Equally, Lauren Petersen has focused on some gardens and garden paintings in Pompeii, noting that they are not necessarily only villa imitations. According to her, they were even more ideal representations of nature than the large villa gardens, which she has proposed to be their examples.32 In general, the architecture and decoration of villae and Pompeian houses share similar features, but the perspective that the lower classes just duplicated the upper strata of society is too simple. It does not take into account the ground up or lateral movement of ideas, and it sees the lower classes as passive imitators without the ability to make their own innovations.

In light of the above, this book carries out a large-scale examination to determine which elements are present in the peristyles of different economic classes, from the wealthiest to the poorest peristyle owners. This comparison will clarify which were copied from the upper classes and which were instead typical of the lower classes but were not commonly found in the peristyles of the wealthy, and were therefore innovations of the lower strata of society meant to demonstrate their economic success.

1.3 Middle class, middling group, middle group, sub-elites?

The subject of inequality in Pompeii has been increasingly studied. Miko Flohr has recently examined the distribution of household wealth in the city, and he comes to the conclusion that the group between the rich and the poor – Flohr calls them middling groups – was a significant economic factor in the city.33 Many of Pompeian peristyles belong to this group, and perhaps in the wider perspective they represent something that could be defined as the ancient middle classes.

There is debate over whether a concept such as the middle class can be used in classical studies, or if the modern connotations are too strong and direct, leading us to imagine things that did not exist in ancient societies.34 This point of view is demonstrated by the criticism of Mayer’s study The Ancient Middle Classes: Urban Life and Aesthetics in the Roman Empire, 100 BCE–250 CE (2012). In particular, his definition of the middle class and his assessment of the possibility of recognizing and differentiating the middle class from the elite in the archaeological sources have been challenged.35

In any society – including the Roman – the borders between social groups are rarely clear.36 For instance, an individual can belong to several groups, or persons can appear to be somewhere between two groups, and it is almost impossible to define to which group they belong. Furthermore, the borders between the groups are constantly shifting. To avoid as much as possible setting arbitrary boundaries between groups, I have classified the peristyles according to their common archaeological features, and therefore the groups are determined by the archaeological remains, instead of a need to identify certain social groups in different houses of Pompeii. Of course, my grouping only reflects the image given by the peristyles, and it may be that other areas of the house might have expressed a different socioeconomic standing, which can be controlled for – at least partly – by comparing the peristyle to the other elements of the house architecture.

As is generally the case when examining antiquity, the source material favors the highest strata of society. It is relatively easy to locate the peristyles of the highest municipal elite – or at least the peristyles with the wealthiest owners. These are large and richly decorated. However, descending in the social ranking makes the interpretation more complex, and the differences between the lower classes are not necessarily so obvious, or the groups are very vague. For instance, some interpretations of the Pompeian houses propose that very different types of dwellings belong to the middle class: from the famous Casa dei Vettii (VI,15,1) to the less well-known house I,11,14.37 Indeed, both houses fit in the middle class, if it is defined broadly, but nobody would easily compare them to each other or suggest that they reflect similar owners.38 One possible solution is adding more subgroups to the division; for example, the upper middle class and lower middle class.

Even if drawing a line between the elite and middle class is difficult,39 the fact remains that most of the Pompeian peristyle owners belonged somewhere between the top elite and the lowest stratum of the Roman world. The social status of the wealthiest peristyle owners was perhaps not equal to the highest senatorial or Imperial elite, as persons of this rank are not known inhabitants of Pompeii,40 meaning that the top political class of Rome was absent from its social stratigraphy. However, the Pompeian upper class might have been a part of the Roman elite in some other aspects, such as cultural taste or wealth.41 The wealthiest peristyle owners in Pompeii might have perhaps competed with the wealthiest persons in the era; at least the architecture of some houses has been noted to be equal to – if not more lavish than – the palaces of some Hellenistic royalty.42 On the other side of the social spectrum, owning a peristyle required a certain wealth and social status that permitted property ownership, which excludes the lowest levels of society.43

In Pompeii, the peristyles were a feature of the city’s economic middle class, but also its upper class. The houses studied in this book cover almost the entire economic elite of Pompeii, excluding only a few of the largest houses, possibly some citizens living outside the city walls, and perhaps a few elite houses that are not yet excavated. The major part of the houses examined, however, can be assigned to a group that could be called the Pompeian middle class – or likely even upper middle class.44 Their architecture and decoration indicate that the owners were neither the richest nor the poorest persons in Pompeii.

The term middle class in this book is mainly used to describe wealth, making it a so-called “objective class,” and it resembles a Weberian definition of class, although the house – let alone peristyle – does not always coincide well with the likelihood of the owner’s economic success.45 If we try to move beyond the economic definition towards a so-called “subjective” class and determine whether this middle group formed a culturally distinct group, or perhaps had a common identity, as Mayer has tried to do,46 we quickly find a lack of sources from which to draw this type of conclusion. Any connections with social status, ethnicity, or other possible identities are frequently difficult to deduce from the Pompeian archaeological material, as is demonstrated in Chapter 7.47 Several different social groups, such as freedmen, plebs, or plebs media, and perhaps even some equites, can be included in this economic group, but besides a very few singular exceptions, we can only speculate on the social standing of the Pompeian house owners.48 Consequently, a comprehensive comparison of the house owners can only be carried out using economic standards.

Despite the problems, Mayer’s idea of exploring the use of art and architecture that is distinctively different between the elite and the classes below it is good. The evidence of material culture – particularly when it was visible in the city or landscape – tells us about a certain pride of belonging to a group that may not have been at the top of society, but nevertheless had its own achievements. However, the problem that remains with the archaeological evidence is that we cannot know whether this pride of social status meant pride in belonging to the economic middle class, or perhaps belonging to some other group inside this massive classification of people. For instance, there could be a pride in belonging to a specific collegia rather than the middle class as a whole. The economic middle class as portrayed in various archaeological sources is often diverse and complex, and likewise in Pompeii it was not a single unit. My purpose is to demonstrate the different nuances and smaller groups of peristyle owners – avoiding the lure of dividing Roman society into elites and others, or rich and poor, or of simply defining a single group as a middle class.

The concept of middle class can be replaced, for example with the word sub-elite(s), which has been used in the scholarship. However, the peristyle owners do not include the poorest or the lowest social levels, such as slaves, making sub-elites too broad of a term to use, as it also includes these groups. Middle class is more precise than sub-elite in this case, and therefore I have decided to use it. My application of the middle class is a working tool used to clarify the context for the reader, not to suggest that ancient society was absolutely divided in a similar way to the modern.

The belief in the usefulness of the concept of middle class seems to follow language barriers. The opposition comes mainly from the Anglophone countries, while the rest of the world does not seem to share this strong concern.49 It is possible that there is an English meaning that prevents the use of the middle class in its classical context, and then we who decide to use the language just have to accept it and stop using it. However, this reason is somewhat unclear. The concept is anachronistic, but so are many other terms that scholars keep using, such as, for example, elite, or public and private.50 Should one be extremely strict and logical with this argument, one could even question using English as the language of ancient studies at all. Yet, writing in Latin, ancient Greek, or another ancient language would not solve this problem, because our use of the words would still have different meanings than in antiquity. If we were to strictly use the words only known from ancient sources, it would lead to a situation where we were basically just copying the ancient sources, and this would hardly produce new information, which is the purpose of research.

My example is extreme, but it highlights that using modern concepts as research tools is one acceptable method. They can help us see our sources differently and produce new information about ancient life. Modern phenomena such as feminism or racism are not found as such in the ancient material, yet these terms can still be used as research tools to interpret the ancient world and to explore similar phenomena. Equally, later dwellings, for example from 19th century England and France, have been utilized to interpret the Roman domus.51 This is, of course, anachronistic, but modern examples – as well as terms – can still provide new points of view.

There is always a need in scholarship for a discussion about the uses of modern concepts such as the middle class – their implications and pitfalls. I am using middle class as a term to define a group between the rich and the poor, and therefore it cannot be thought of in strictly Marxist, Weberian, etc. terms. I have identified several smaller groups inside this middle group. These are named according to the decorative and architectural features of the peristyles – opulent, large full, ornamental, large painting, imitation, minor decoration, and architectural peristyles – but additionally they correspond more-or-less to the different economic groups of the peristyle owners: upper class, upper and lower middle class.52 It is impossible to say whether the Pompeians themselves noticed that there was a large economic middle group in their city, but they likely perceived some type of similarity inside these smaller groups that I have defined.


1. Wallace-Hadrill 2008, 439–440.

2. Mazzoleni 1993, 7, 290, 293. Wallace-Hadrill 1994, 3–4, 147. Zanker 1998, 10–11. Hales 2003, 2–3.

3. For a more detailed discussion of the definition of the peristyle garden, see Chapter 2.2. In this book, both peristyle and peristyle garden are generally used for describing open spaces that had at least one colonnade on one side. The words are used synonymously, which is a scholarly convention. In addition, I sometimes utilize the term garden with one portico to make a distinction between them, peristyles, and pseudo-peristyles. Nevertheless, the gardens with one portico are also referred to as peristyles, particularly if they belong to a group that includes gardens that have more than one colonnade. In general, I use the space and room names and numbers that are presented in Pompei: pitture e mosaici, except all the vestibula are called fauces. Although Latin nomenclature is used, the names do not signify room functions (for the problems with Latin names, see Allison 2006, 405), but rather refer to the scholarly tradition.

4. E.g. Mustilli 1950, Laurence 1994, McGinn 2002, Peña & McCallum 2009, Monteix 2010; 2017, Flohr 2011; 2013; 2013b; 2021, Ellis 2004, 2011; 2011b; 2018.

5. On the display function of the peristyle, see Leach 1997, 52, Sampaolo 1997, 428; 1998 974, Zanker 1998, 12–13, Dickmann 1998, 452, Bragantini 1999, 142, Farrar 1998, 19, Wallace-Hadrill 2008, 193, Ciarallo and Giordano 2012, 389, n. 41, 537 n. 280, 556 n. 306, Trentin 2014, 1. For new interpretations of the Roman house, see Allison 1993; 2004, Berry 1997, Wallace-Hadrill 1997, Nevett 2010, 89–118, Flohr 2011, Tuori 2015, Green 2015, Nissin 2015.

6. Rapoport 1990, 11.

7. Bourdieu 1979, 1–2, 6, 248 fig. 10. See also Wallace-Hadrill 1994, 148; 2008, 326–327, Walsh 2014, 80–81.

8. E.g. Zanker (1998, 247) has 35 houses in his index, and Hales (2003, 7–8, 290–291) has 32 houses in her index. Some scholars have utilized sample areas, such as Wallace-Hadrill (1994, 65-72) using Insulae I,6–12 and VI,9–16. Other studies applying sample areas are: Grahame 2000, 38–39, Allison 2004, 6–7, 29–30, Lohmann 2015, 71–71. There are some studies covering several houses around the city, such as Dickmann 1999 (see index pp. 379–381) and Anguissola 2010 (see pp. 1, 513–573). On the peristyle in general, see Richardson 1988, Zaccaria Ruggiu 1995, Jashemski 1993, 15–19, Jones & Bon 1997, 3–5, Meyer 1999, Hodske 2007, 17–22, Poehler, Flohr & Cole 2011, 1–8, Zarmakoupi 2014, 8–13, Tuori 2015, 9–10, Nissin 2016, 14–16. Kawamoto (2015, 111–195) has a very thorough and good overview of the history and scholarship of garden archaeology, particularly for the Bay of Naples.

9. On the house, see Mazzoleni 1993, 7, 290, 293, Wallace-Hadrill 1994, 3–4, 147; 2008, 193, Leach 1997, 52, Zanker 1998, 10–13, Farrar 1998, 19, Hales 2003, 2–3. On the clothing, see Edmondson 2008, 23–26, 32–37, Wallace-Hadrill 2008, 41–57, Larsson Lovén 2014, 266–270.

10. Cic. Att. 1.6.2. See Wallace-Hadrill 2008, 171–173, 315, 329–338, La Bua 2019, 319.

11. See Mau 1908, 250–289. See also Becker 1838, 70–102, Clarke 1991, 2–19, George 1997.

12. See Wallace-Hadrill 1994, 7, Nevett 1997, 282–288, 294–298; 2010, 93–97, Berry 1997, 183–185, 193–195; 2016, 125, 141, Allison 2004, 161.

13. Ciarallo & Mariotti Lippi 1993, 116. Pesando 1997, 6, 9. Allison 2001, 53; 2004, xv, 14. Petersen 2006, 128, Viitanen & Ynnilä 2014, 142. Speksnijder 2015, 88.

14. Wallace-Hadrill 1994, 83.

15. See e.g. Niccolini & Niccolini 1854, Casa detta del Fauno, 1.

16. See e.g. Tuori 2016, 104, 157, 187, 195, 222, 277–278, 289.

17. Although the status of these groups are diverse and, for example, equites could be defined as an elite group, as Davenport (2019, 5) has done.

18. On the middle class in antiquity, see Mayer 2012. Interpreting ancient social and economic structures through modern terms is not a new method, Rostovtzeff (1957, 10, 21, 195) for example uses such terms as bourgeoisie and capitalistic. For critical views, in particular about the term middle class, see Wallace-Hadrill 2008, 354, Archibald & Davies 2011, 1–2, Newby 2016, 20–24, Ellis 2018, 12.

19. See Petersen 2006, 11.

20. See Zanker 1998, 16, 19, 20, 192–193, 199–202. For the adaptation of the model, see Bragantini 1991, 34, Sampaolo 1993, 613, Fröhlich 1993, 641; 1996, 116, Inserra 2008, 23, Loccardi 2009, 69. On the top-down model in general and its criticism, see Hales 2003, 8, 250 n. 27, Von Stackelberg 2009, 21–22, Mayer 2012, 166–167.

21. Walsh 2014, 83.

22. Wallace-Hadrill 1994, 4–5, 143, 146–147.

23. Zanker 1998, 12–14, 16, 19–20, 142, 145, 160, 168, 192–193, 199–202. Zanker has partly published the same examination of and conclusions on Pompeian houses already in the article: Die Villa als Vorbild des späten pompejanischen Wohngeschmacks (1979). I use the English translation (Pompeii: public and private life, 1998) of Zanker’s study, as it is the most recent version of the book, and although the main text is mostly similar to the previous editions, Zanker (1998, viii) states that the notes are updated. Zanker’s view has been extensively quoted and followed, see, e.g., Wallace-Hadrill 1994, 14, 169–174; 2015 186, Jones & Robinson 2005, 696, Dickmann 1999, 159–255, 299–300, 355, Hales 2003, 7–8, 137-138 (also includes critique of Zanker’s ideas), Hodske 2007, 22, Zarmakoupi 2014, 9, Tuori 2015, 10, Morvillez 2017, 36. See also Petersen 2006, 128–129 for the critique. The interpretations of villa imitation in the Pompeian house: Bragantini 1991, 34, Sampaolo 1993, 613, Fröhlich 1993, 641; 1996, 116, Inserra 2008, 23, Loccardi 2009, 69, Kuivalainen 2019, 68. Zanker (1998, 135) mentions that the idea of villa features in the domus is not original, and that scholars before him mention the connection (see, e.g., Bechi 1835, 10).

24. See, e.g., Wallace-Hadrill 1994, 5–6, Zanker 1998, 12, Dickmann 1999, 308, 374, Jones & Robinson 2005, 700, Von Stackelberg 2009, 22 (also criticism).

25. Veblen 1957, 68–101, 103–105, 126–128.

26. Campbell 1995, 37–40, 45–46.

27. Walsh 2014, 84, 86.

28. Hales 2003, 8, 250 n. 27. For criticism, see Mayer 2012, 166–167.

29. Wallace-Hadrill 2008, 436.

30. Mayer 2012, 166–212, 214–216, 218.

31. Von Stackelberg 2009, 21–22.

32. Petersen 2006, 123–162.

33. Flohr 2017, 54–55, 75–81. There is also an article by Kohler & al. (2017) about the distribution of household wealth in premodern cities around the globe. Pompeii is among the studied cities. Unfortunately, the data set provided by the study was not transparent enough, and it was impossible to control for which houses were selected to represent Pompeii in this study. Additionally, counting the Gini coefficients on the basis of household size excludes the large portion of people who did not own property, and therefore comparison with income-based Gini calculations – used for modern society – is not possible.

34. See Wallace-Hadrill 1994, 172–173; 2008, 354, Newby 2016, 20–24, Ellis 2018, 12.

35. See, e.g., Newby 2016, 20–24. There have been several critical reviews of Mayer’s book, as Newby mentions in note 97. Mayer (2012, 7–8) himself also admits that the middle classes and the elite might have overlapped in some areas of life.

36. Mayer 2012, 8–14.

37. See also Section 7.1. I use Italian house names, as this is currently the custom in Pompeian studies. It is possible to translate names using Google Translate or Deep-L. In addition, the English names, as well as alternative Italian names, for several houses can be found in the Pompeii in Pictures website. The house names used in this study are adopted from Pompei: pitture e mosaici, excepting I,2,24, II,8,2/3, VII,10,5, VII,11,6–8, VII,11,11/14 and IX,1,12, which are simply referred to as houses (not caupona, lavanderia, or albergo). Several houses have two names or additions (e i suoi annessi) in Pompei: pitture e mosaici, but here only the main part of the name is used: house VI,14,20 is simply Casa di Vesonius Primus, VI,14,43 Casa degli scienziati, VII,4,59 Casa della Parete nera, VII,6,3 Casa di M. Spurius Saturninus, VII,14,5 Casa del Banchiere and VIII,4,4/49 Casa dei Postumii. In Pompei: pitture e mosaici house I,6,9 is separated from house I,6,11 which is called as Casa dei Quadretti teatrali. As the houses are linked through a door between peristyles (nn. 20, 21, n. or nn. refers to number(s) of the peristyle in the Appendix) and there is no reason to expect that the houses were separate units, I deal with them as one house called Casa dei Quadretti teatrali. A similar situation is found with houses VII,7,2 and VII,7,5; therefore they are also considered as the same house, called Casa di Trittolemo.

38. See nn. 42, 134, 135.

39. See e.g. Petersen 2006, 83.

40. See Camodeca 2008, 25, noting that there are no known senatorial class members in Pompeii. There might have been persons who had contacts with the Imperial family, such as T. Suedius Clemens, but whether they were house owners in Pompeii is unknown. See Chapter 7 about the social status of the peristyle owners.

41. E.g. Wallace-Hadrill (2008, 438–439) in the context of absorbing Hellenistic fashions, regards the Pompeian elite to be equal to the Roman political elite.

42. See Castrén 1975, 40, Dickmann 1997, 123, Wallace-Hadrill 2008, 438–439.

43. On the economic status of Pompeii compared to the entire Roman Empire, see Ray 2017, 106 and Verboven 2017, 369–370, although the low number of data sets only enables tentative conclusions (Jongman 2017, 421).

44. Flohr (2017, 80) situates the houses with peristyles or porticoes in the upper 25–30 percent of Pompeian households, which is a general assessment, but there are some houses with a peristyle or portico that do not belong to this group. All of Flohr’s peristyles and porticoes are included in the houses of the top 40 percent. However, the data does not include the inhabitants that did not own property, meaning that on a larger scale all of the peristyle owners very likely belonged to the upper middle class in wealth. On the missing data of Pompeian houses, see Flohr 2017, 60–62.

45. Weber 1978, 302–307; 1978b, 926–939. On the “objective” class, see Hall and Stead 2020, 13.

46. Mayer 2012 18–21. On the “subjective” class, see Hall and Stead 2020, 13.

47. See also Castrén 1975, 31–33, Mouritsen 1988, 13–27, Wallace-Hadrill 1994, 108, Allison 2001, 57, Painter 2001, 35, Mayer 2012, 33, 53, Viitanen, Nissinen & Korhonen 2012, 67.

48. Jongman (2017, 425) suggests that a major part of 500–600 atrium houses was inhabited by freedmen.

49. For non-English use, see e.g. Miniero 1990, 598, Peters & Moormann 1993b, 409, Strocka 1994, 648, Seiler 1994, 714, Sampaolo 1998, 1091, Zanier 2009, 229, Mayer 2012, 8.

50. On the public and private, see e.g. Russell 2016, 12–16.

51. Wallace-Hadrill 1994, 9–10, 12–14.

52. See Section 6.1.



If you find an error or have any questions, please email us at Thank you!