Introduction

Archaeologists studying material culture often base their interpretations on a clearly defined and quantifiable dataset. Excavation reports typically include catalogues of finds, analyses of ceramic material, and numismatic reports. All of these are based on quantifiable information with an underlying dataset. As such, this approach should also be taken for the study of artworks found at a site. Instead of making generalisations based on years of research and selecting individual examples to support points, where possible, scholars should start with the number of artworks from a site or of a specific subject at a site, and separate them into categories. These categories need to be as objective as possible, as interpretations can then be made based on clear, quantifiable data.

Pompeii is a good case study for demonstrating previous approaches to Roman artworks, and how a quantitative approach can yield useful results, given the abundance of extant wall-paintings over a large area and the mass of previous scholarship on the site. There have been several attempts to catalogue Pompeii’s wall-paintings, most notably the catalogues of Helbig (1868) and Sogliano (1880), organised by subject, and of Schefold (1957), organised by location. Grouping wall-paintings by subject in a catalogue is problematic, as it relies on the user and the compiler having the same interpretation. For instance, if a user were to research representations of the ‘Love of Venus and Mars’ in a catalogue organised by subject, they would only find those images that the compiler identified as this scene type, omitting many other possibilities which the compiler identified as a different couple. The subjective nature of the identification of scenes makes it difficult to organise artworks in this manner. In order to study Roman art in such a way that scholars can build on the works of others, rather than start from scratch each time, a more systematic approach is needed.

Previous Approaches

Scholars of Roman and Pompeian artworks often make general statements regarding how often certain subjects occur in certain areas within houses without quantifying this. For instance, Ling and Ling (2005: 94) argue that the decoration of dining areas often features food, Bacchic scenes, or subjects relating to Greek culture. However, Ling (1995) himself disproved this previously when taking a quantitative approach to the decoration of triclinia (discussed further below). Clarke (2003: 227) states that sometimes the decoration of dining rooms is topical, for instance with banquets being depicted. He articulates the limits of his work, which is essential in order to avoid generalisations, by stating that these examples occur ‘sometimes’, and earlier emphasising that he intends to ‘single out individual instances’ rather than ‘analyzing all the decoration of a house’ (Clarke 2003: 222). However, a quantitative approach to the decoration of dining areas would offer clear evidence of how often this was actually the case. Without this transparency, it is difficult to place weight on statements such as that of Ling (1991: 135), when he claims that ‘myths of Venus and symbols of love are favoured in bedrooms’, or Ling and Ling (2005: 94), who state that ‘in cubicula the subjects are of a more intimate nature and often include love-stories or scenes relating to the goddess Venus’. These comments are made in passing, and no examples or statistics are given to support either of these statements. Although some examples in Pompeii would support these assertions, in order to ascertain their accuracy, the categories need to be further defined and quantifiable data needs to be compiled. ‘Symbols of love’ and ‘scenes relating to the goddess Venus’ are subjective and need to be defined before their location can be investigated. If this definition was restricted to ‘scenes featuring Venus’, the data could be more reliably determined. Furthermore, instead of giving a few examples where the content of scenes is influenced by the type of room they decorate, a sample is needed in order to establish the number of examples which support these points in relation to those which do not – for instance, comparing the number of scenes featuring Venus in cubicula with those in other areas. For these reasons, I have compiled all representations of Venus in the fixed artworks of Pompeii that have been excavated and are accessible, whether in situ, in a museum, or recorded in published materials, to ensure this study is comprehensive.

Roger Ling (1995) takes a quantitative approach to the decoration of Roman triclinia. In this study, he aims to ‘determine the extent to which Dionysiac themes appeared in the decorations of Roman dining-rooms’ through a ‘statistical check’. This is only a brief study, however, and Ling does not attempt to identify the function of rooms himself, but instead analyses triclinia identified by Schefold (1957) and Bragantini et al (1981–1992) in the Pitture e Pavimenti di Pompei (PPP). Ling focuses only on decorations in the centre of walls, ceilings, and pavements, and within this only on those containing figure-subjects (humans or animals), landscapes, or still-lifes. He examines 137 triclinia and identifies 7 with predominantly Dionysiac decoration, 39 with some Dionysiac elements, and 86 with non-Dionysiac decoration. Just one or two of the 137 triclinia feature scenes of non-mythological banquets and just seven of the 137 triclinia depict xenia (still-life paintings of food). Ling (1995) therefore discovers through statistical analysis that Dionysiac decoration and dining-related scenes are not particularly common in triclinia. He argues (1995: 245) that ‘decorators and their patrons were aware of the appropriateness of Dionysiac myths and motifs in the decoration of dining-rooms but did not feel the need to make them the exclusive, or even the dominant, theme.’ This quantitative approach would therefore suggest that the decoration of a room does not always reflect its function.

While this approach has clear value, it was rarely emulated until the work of Jürgen Hodske (2007) who used a quantitative approach to examine central mythological images in Pompeii. In this study, Hodske identifies which subjects appear more often in different room types and areas (i.e. the atrium zone and the peristyle zone), and changes in this between different stylistic periods. Hodske identifies 859 central mythological images across 202 Pompeian buildings. He discovers that these images appear more often in triclinia (27%) and cubicula (26%) compared to other room types (Hodske 2007: 69). The work of Hodske (2007) demonstrates the importance of a quantitative approach in forming conclusions, and he sometimes substantiates, but often disproves, the previously held assumption that decoration reflects the function of a room. For example, he discovers that Dionysiac subjects are not particularly dominant in triclinia and oeci (Hodske 2007: 76). It must be acknowledged, however, that Hodske only studies central mythological images and not all decoration, so this work is not an accurate reflection of all artworks in rooms across Pompeii, but can be used as a representative sample.

These quantitative approaches are in stark contrast with art-historical approaches. Art historians often over-interpret the artworks, especially with regard to iconography. For instance, Lorenz (2007: 667) argues that in the painting of Perseus and Andromeda in VII.16.10, Andromeda’s uncovered breast alludes to the Amazons and implies ‘that the location of Andromeda’s current situation is placed far away from civilization’. This is a possibility, but there is no empirical evidence to support Lorenz’s interpretation. My approach to artworks in Pompeii aims to break away from such art-historical approaches by separating each artwork into categories relating to its content and location. The categories are designed to exclude subjectivity as much as possible—for instance, Venus’ state of nudity, the number of figures in the scene, and the attributes of Venus present in the scene are generally not open to different readings. Naturally, some subjective interpretation is unavoidable when identifying Venus and the scene type, but this is kept to a minimum and clearly defined criteria are used to identify representations as Venus. Discussion can then be based on these clearly evidenced categories rather than on subjective interpretation of the artworks.

Overview of the Approach

This article is based on the results of a more extensive research project that aimed to investigate the significance of representations of Venus and their spatial contexts in Pompeii. More widely, this study sought to ascertain whether the location of an artwork influenced its composition. Pompeii was selected as a case study because it offers a unique opportunity to study the relationship between decoration and location in-depth across a town because of its exceptional preservation of fixed artworks. Venus was chosen over other deities because she was the patron deity of Pompeii (with the Roman colony of Pompeii named after her—Colonia Veneria Cornelia Pompeianorum), and could therefore be expected to feature prominently in the fixed artworks of the town.

The Roman names for deities and mythological figures are used throughout this study, such as Venus instead of Aphrodite, and Bacchus instead of Dionysus. This is both for consistency and because the iconography of syncretised deities often cannot be reliably distinguished. For example, this study argues that Venus Pompeiana probably represented the patron deity of the town – Venus rather than Aphrodite. However, in Greek mythological scenes, such as those depicting the ‘Judgement of Paris’, the deity represented is Aphrodite rather than Venus. Thus, all representations of Venus and Aphrodite are considered as ‘Venus’ in this study.

In order to investigate the significance of representations of Venus and their spatial contexts in Pompeii, I decided to adopt a quantitative approach to the artworks. Allison’s (2004a and 2004b) seminal work was influential in the creation of this quantitative approach. She created a database of material culture in 30 Pompeian houses using set criteria including artefact type and material, also recording the location of each artefact. I decided to take a similar approach to Pompeian artworks by separating my database into various categories relating to elements of composition and location. Allison’s room type criteria (2004a: Table 5.a) are also used in this study as she does not identify functions for rooms (except where a specific fixture suggests a function, such as a hearth) since this would be subjective. Instead, Allison identifies room types based on objective structural evidence, such as the openness and size of the room, and its location, which allows for the multifunctional and seasonal nature of rooms. Obviously, rooms in Roman houses could be made more open or closed using temporary partitions such as doors, curtains, screens, or windows with shutters (Pliny, Letters 2.17.5, 20–22; 5.6.19, 29, 38; George 1997: 317–318; Hodske 2007: 80; and Proudfoot 2013: esp. 93 and 100). Indeed, there is evidence of a wooden partition separating the atrium from the tablinum in the Casa del Tramezzo di Legno (III.4–12) in Herculaneum (De Kind 1998: 105). However, as evidence does not always survive of this, Allison (2004a) identifies rooms as open or closed based on tangible evidence, such as the presence of walls. Furthermore, although open rooms could be made more closed, it would be difficult to make structurally closed rooms more open.

Often, scholars of Roman art make conclusions based on years of research, but only use a few examples to support their points. Their works encouraged me to approach Roman art more systematically by creating a database recording various elements of the content and location of these artworks to provide conclusions based on numerical data. As a result, scholars can clearly see the evidence for such conclusions and can disagree with my interpretation of a scene but still use the contents of my database.

In order to identify representations of Venus in Pompeii, I first researched representations of Venus and Aphrodite outside Pompeii in the works of ancient authors who mention famous artworks and the Lexicon Iconographicum Mythologiae Classicae (LIMC). The LIMC is a multi-volume encyclopaedia of mythological subjects in classical art from across the world that separates individual subjects into typologies, but does not address them in a systematic way. Though it is detailed, it does not aim to list all known examples of a type, or all representations of a subject from a particular location, but it did enable me to create lists of Venus types, attributes of Venus, and scene types. The resulting lists were used to develop ‘Criteria for Identifying Venus’ so that other scholars can clearly understand why I identified certain representations as Venus without having to discuss the evidence for each one individually. The criteria were arranged into categories on a scale from (a) to (g) to identify how likely a representation is to depict Venus:

1. Representations of Venus which name or label the goddess within the artwork.

2. Representations of Venus which depict the goddess in recognisable narrative scenes, such as the Judgement of Paris.

3. Representations of Venus which adopt a pose of Venus (such as that of Venus Pudica) known from surviving artworks or ancient authors.

4. Representations which are very likely to be Venus (for instance, the figure appears with many of Venus’ attributes, or adopts a common pose of Venus and appears in the same scene as a deity).

5. Possible representations of Venus where some of her attributes are depicted or a possible narrative featuring Venus can be ascertained.

6. Representations which could be Venus, but there is insufficient evidence. For instance, if there is just one attribute of Venus, or a scholar or excavator suggests it represents Venus, but I disagree. Artworks from this category are excluded from the tables and discussion in this article but are included in the database so that other scholars can consider these for themselves.

7. Representations of Venus that were identified by earlier scholars or excavators, but for which no visual evidence survives today.

I compared my ‘Criteria for Identifying Venus’ against the entire Pompei: pitture e mosaici (PPM) – a detailed 11-volume catalogue of the artworks in the buildings of Pompeii — in order to identify representations of Venus. I identified 197 representations of Venus in the fixed artworks of Pompeii and compiled them into a database (Brain 2018). Fixed artworks — wall-paintings, reliefs, ceiling-paintings, and floor mosaics — were used because their location is permanent. Statuary and other material culture were not studied as their location is not fixed and they could be moved between rooms, making their usual location difficult to establish with certainty.

The information on artworks featuring Venus in the database is separated into two main groups: content and location. The information on content is separated into 11 categories: criteria for identifying Venus, type of scene, Venus type, Venus’ state of nudity, number of figures in the scene, number of cupids in the scene, deities depicted with Venus, other figures in the scene, type of headwear, type of jewellery, and type of attributes. The information on location is separated into eight categories: domestic or public context, region, building, room, room type, wall or façade, field of wall, and zone.

General Patterns: Location of Artworks

A quantitative approach to artworks featuring Venus in Pompeii reveals many insights into the content and location of Roman decoration, and also the relationship between decoration and room function. 162 representations of Venus are from domestic contexts, 12 are from more public areas (including shops, fulleries, cauponae, and the macellum), and 11 are on façades (Table 1). Six of the 162 Venus representations from domestic contexts are from unknown rooms (although for one of these we know the room type but not the actual room itself). Thus, the room type can be determined for 157 representations of Venus in domestic contexts. Venus representations occur in 105 different houses, and within these in 143 different rooms (excluding the six from unknown rooms), as some houses and rooms feature multiple representations of Venus.

I used Allison’s (2004b) database and room type criteria to identify the frequency and percentage of each room type in her sample of 30 Pompeian houses. I applied her room type criteria (slightly adapted to allow for non-atrium houses—Table 2) to rooms in Pompeian houses which featured a representation of Venus. I compared the percentage of Venus representations in each room type across all houses in Pompeii with the percentage of rooms of each type within Allison’s sample of 30 houses in Pompeii (Table 3). If the percentage of Venus representations in a room type was higher than the percentage of rooms of each type in Allison’s (2004b) sample, it would suggest that Venus was particularly popular in a room type. If the percentage of Venus representations in a room type was lower than the percentage of rooms of each type, it would suggest that Venus was less common in this room type. Obviously, this does not allow for some room types being more decorated than others — for instance, toilets were less likely to be decorated than other areas — but this serves as an indicator of whether Venus was represented more or less frequently than could be expected, given the frequency of each room type. This would be more accurate if compared with the number of decorated rooms of each type across Pompeii. However, to my knowledge, no study exists that quantifies the number of decorated rooms of each type across Pompeii. The nearest study to this would be that of Hodske (2007) who quantified the number of central mythological images in rooms of each type across Pompeii. However, it is unclear how Hodske identified room types and he did not use the same room type criteria as Allison. Therefore, his findings will be mentioned below but are not directly comparable with my quantification of representations of Venus in Pompeii.

This approach reveals that there are some room types in which Venus representations never occur. Her absence from these room types is easily explicable — she is absent in room types 1, 8, 17, and 19 (entranceways, corridors, and stairways), probably because these are transitional areas and thus often lack detailed scenes. Indeed, as Hodske (2007: 85) notes, house entrances are quite small and would be ill-suited as a location for wall-paintings. Hodske’s quantitative analysis revealed that only three central mythological images within two houses exist in house entrances. However, he notes that the Pompeians seemed to prefer the depiction of patron deities in this area, so perhaps the absence of Venus is noteworthy. Venus is also absent in type 2 rooms (rooms leading directly off the front entranceway), which is explicable as these are the least common room types in Allison’s sample and they often lack painted wall decoration. Venus’ absence from type 15 rooms (latrines) is unsurprising, given that there are few toilets decorated with figures in Pompeii (Jansen 1993: esp. 32; Moormann 2011: 61–63), although Fortuna is represented five times in Pompeian toilets (Jansen 2011: 167–170). Venus is also absent in type 18 rooms (secondary gardens or courtyards), probably because secondary gardens received less decorative attention than main garden areas (type 9 rooms), based on Allison’s (2004b) descriptions of the decorative state of individual rooms and spaces in her sample. I consider shops attached to houses (Allison’s room type 20) to be public spaces, so they are excluded from this discussion of domestic room types. No Venus representations are recorded in type 22 rooms (on upper floors), probably due to the collapse of many upper floors in the eruption or later.

When room types which do not feature Venus representations are removed from consideration, and the percentages of rooms of each type are redistributed, the percentage of room types in houses moves closer to the percentage of Venus representations in each room type. My dataset of Venus representations in individual room types appears to correspond to the decorative state of individual room types in houses across Pompeii, as Hodske’s (2007: 69, diagram 23) study of central mythological images in Pompeian houses yields similar results. Hodske discovered that most central mythological images are located in triclinia (27%), then cubicula (26%), tablina (8%), oeci (8%), peristyles (5%), alae (4%), exedrae (4%), atria (3%), other rooms (2%), bottegae (1%), and rooms of unknown function (12%). I do not use this terminology, but triclinia and oeci would roughly equate to large/medium rooms (types 6, 10, and 11 — where 43% of Venus representations are located); cubicula to small, closed rooms (types 4 and 12 — 23%); tablina to type 7 rooms (11%); peristyles to type 9 areas (6%); alae to type 5 rooms (4%); exedrae to type 13 rooms (3%); atria to type 3 rooms (4%); and ‘other rooms’ to the remaining rooms (types 14, 16 and 21 — 4%). Table 4 highlights types of room where Venus is particularly common or particularly uncommon, and types of room where Venus occurs proportionally as often as the room type. The percentages given in the below discussion of rooms of each type in Pompeian houses are based on the redistributed percentages of Allison’s (2004b) sample of 30 houses set out in Table 4. The percentages given in the below discussion of Venus representations in each room type are based on those from my sample of all Pompeian houses also set out in Table 4.

18% of rooms in Pompeian houses are large or medium rooms (room types 6, 10, and 11), and 43% of Venus representations occur in these room types. This indicates that Venus is particularly common in large and medium rooms; but larger rooms were probably used to receive guests in view of their capacity, and thus may have received more detailed decoration than other areas of the house.

Only 5% of rooms in Pompeian houses are type 7 rooms (open-sided rooms usually opposite the main entrance or leading to the garden, often referred to as tablina), yet 11% of Venus representations occur in this room type. Venus, therefore, appears more often in this room type than would be expected. Indeed, the study by Hodske (2007: 82–83, 90) of central mythological images in Pompeii reveals that Venus is particularly popular in this room type compared with other subjects, with the pairing of Venus and Mars alone accounting for 11% of the central mythological images in tablina. An examination of the other scene types depicted in type 7 rooms may offer some insight. Strocka (1997) notes that Venus and Mars scenes in tablina always occur amongst other love scenes. Indeed, by quantifying the other scenes represented in type 7 rooms alongside Venus, it is clear that Venus is always represented amongst love narratives or scenes which could relate to love but this is not the main focus of the scene, such as the ‘Punishment of Cupid’ scene in room f of VII.2.23, and Bacchus with Ariadne in the ‘Triumph of Bacchus’ scene in room 7 of V.4.a (Supplementary Table 1). This would suggest that the popularity of Venus in type 7 rooms might be due to the popularity of love scenes in this room type, but a more specific analysis of all the subjects in type 7 rooms across Pompeii would be needed to draw accurate conclusions.

Interestingly, despite scholars such as Ling and Ling (2005: 94) and Anguissola (2010: 435) assuming that scenes associated with Venus (as the goddess of love), and love scenes and themes generally, would be frequently represented in ‘bedrooms’ or cubicula, a quantitative approach to representations of Venus demonstrates that this is not the case in regard to representations of the goddess herself. In order to avoid connotations of room function, I do not use the term cubicula, and the nearest equivalent in my study would be small, closed rooms (types 4 and 12). Hodske’s (2007: 70) study revealed that cubicula were the second most common area of the house to feature central mythological images — with 26% of all central mythological images located in this room type, after triclinia with 27%. If representations of Venus were particularly common in cubicula, we would expect her to occur more often than other mythological images in this room type (thus more than 26%). However, just 23% of Venus representations occur in these types of room (though small, closed rooms are the second most popular location for representations of Venus, after large/medium ones with 43%). This low percentage is even more significant when compared with the proportion of small, closed rooms in Pompeian houses (32% of rooms in Pompeian houses are small, closed rooms based on the redistributed percentages of rooms of each type in Allison’s sample). Furthermore, just 21% of love scenes featuring representations of Venus in domestic contexts occur in small, closed rooms, whereas the larger proportion (46%) occur in large/medium rooms (types 6, 10, and 11). Love scenes featuring Venus, therefore, occur less often in so-called cubicula than in other areas of the house.

Type 14 rooms (cooking areas) account for 7% of rooms in Pompeian houses, and less than 1% of Venus representations occur in these rooms. Cooking areas would be unlikely to contain detailed paintings as the smoke from the hearth would have continually damaged them (lararium paintings are an exception).

Despite Venus’ association with gardens and bathing, the goddess is represented slightly less often in these areas than these areas occur in the house (7.3% of spaces in Pompeian houses are gardens, type 9 rooms, compared with 6.4% of Venus representations in these areas; and 3.1% of rooms in Pompeian houses are bath areas, type 21 rooms, compared with 1.9% of Venus representations in these areas). That said, this is perhaps unremarkable as these areas featured less mythological subjects in general — Hodske’s study (2007: 69, diagram 23) of central mythological images revealed that they were rarely located in peristyle areas (just 5% of all central mythological images were located in peristyles) or bath areas (less than 2%). Despite central mythological images and representations of Venus not being particularly common in peristyles compared with other areas of the house, Hodske (2007: 87) notes that Aphrodite is represented slightly more often than other subjects in central mythological images in peristyles. It is possible that representations of Venus were more popular in peristyles than we can identify now, as wall-paintings in gardens are more susceptible to weathering.

In addition to offering insights into the relationship between decoration and its location within the house, a quantitative approach can also provide evidence for the relationship between decoration and its location within rooms. For instance, Venus representations occur proportionally more on the north wall (29%) compared with the west wall (16%); 93% of Venus representations occur in the central zone, compared with 7% in the upper zone; and 80% occur in central fields, compared with 20% in side fields. In some cases, this information is not significant: Pompeii is not aligned to a true north-south axis, and some scholars may have confused the north and west walls (site north is used in this study). In other cases, these findings can be explained if they are broken down further. For example, there are 26 representations of Venus in side fields, and the majority of these depict non-narrative scenes (17 of 25, or 68% where the scene can be ascertained), which perhaps partially explains their location in these areas. Six of these 26 representations of Venus in side fields are busts (23%), which generally occur in side fields in order to complement the central field. Ten of these 26 representations (38%) could not be depicted in a central location as doors or windows occupied the central area, the wall consisted of just a left and right section, or for various reasons the wall should not be considered as having left/centre/right sections (for instance if Venus is depicted as part of a frieze where her location is determined by her place in the narrative and no central field exists within the frieze). A further two occur on aediculae or within lararium paintings, which would again explain their location in side fields rather than central fields.

A quantitative approach can also be used to examine the relationship between content and location outside of the home. For instance, there are 11 representations of Venus on façades. Aside from two of these representations occurring in scenes depicting the pantheon of twelve gods (and these scenes contain some variation, including which deities are represented), no scene type appears more than once, indicating that the commissioners of façade paintings wanted unique depictions of Venus. This is in contrast to domestic artworks, which tend to emulate well-known scenes. Of the 162 Venus representations in domestic contexts, the scene type is identifiable for 160 and, of these, 107 can be categorised into 15 scene types, highlighting the amount of repetition. The remaining 53 artworks from domestic contexts depict unique scene types grouped under ‘Other Narrative Scene’, ‘Other Non-Narrative Scene’, and ‘No Active Scene’ (Table 5). On façades, Venus appears alongside at least one deity in six out of the 11 (55%) artworks featuring Venus. Of these six, all except one (which depicts the ‘Love of Venus and Mars’ narrative) are non-narrative scenes, and four depict the deities without any kind of scene taking place (Venus and the other deities do not appear as part of a narrative scene and are not participating in any activity). These deities are thus represented for a reason other than the illustration of a narrative. All seven façade artworks featuring Venus that flank an entrance also feature at least one deity on the opposite side of the entrance (where this can be ascertained). This perhaps served as a protective axis at the entrance to the house, but may have also advertised the owner’s patron deities.

General Patterns: Content of Artworks

A quantitative approach can highlight the most popular Venus types and scene types in Pompeii, as well as identify trends within these, as will be shown in the two case studies of Venus types given below. The most common Venus types in Pompeii are ‘Venus and a Figure Touching/Embracing’ (45 of 184 where the Venus type can be ascertained, or 24%) and Venus Pescatrice (25 of 184, or 14%). The two most common scene types featuring Venus in Pompeii are the ‘Love of Venus and Mars’ (37 of 195 where the scene type can be ascertained, or 19%) and ‘Venus Fishing’ (25 of 195, or 13%). There is a substantial overlap between the two most common Venus types and the two most common scenes featuring Venus. All ‘Venus Fishing’ scenes feature Venus Pescatrice, and vice versa, and 33 of the 36 ‘Love of Venus and Mars’ scenes where the Venus type can be determined feature the ‘Venus and a Figure Touching/Embracing’ type (92%). This is significant because it demonstrates that some scenes require certain Venus types and vice versa.

To avoid subjective interpretations of Venus’ state of nudity, as these often depend on the individual scholar’s views and cultural-historical background, this study uses a quantitative approach to identify whether there are any correlations between Venus’ state of nudity and the Venus type, scene type, and location. This approach demonstrates that Venus’ state of nudity often depends on the Venus type. For instance, Venus Pudica (Venus holding her hand over her pubic area) is always naked, necessarily so if she is to cover or emphasise her pubic area. Venus Pescatrice (Venus as a fisherwoman), however, is never fully clothed but always semi-naked or with drapery falling below her pubic area, perhaps to reflect the association of eroticism with Venus and fishing (discussed below), but also because these scenes were probably all based on the same prototype. The Venus type used affects the state of nudity of Venus more than the location of the artwork. Despite the assertion by Jashemski (1979: 130) that Venus is always ‘scantily clad or nude’ when depicted in gardens, numerical data demonstrate that of the nine Venus representations in gardens (type 9 rooms) where the goddess’ state of nudity can be ascertained, she is clothed in four, naked in four, and semi-naked in one. Where the state of nudity and context are known, Venus is mostly naked or semi-naked in artworks from both domestic contexts (90 of 135, or 67%) and public contexts (8 of 11, or 73%), whereas she is more likely to be clothed on façades (7 of 11, or 64%). Venus’ state of nudity in the six representations of the goddess within lararium paintings or painted on aediculae again depends on the Venus type used: the four Venus Pompeiana representations are clothed, while the representations of Venus Anadyomene and Venus Pudica are naked.

To demonstrate how quantitative data on the content and location of artworks can be used to form new conclusions, two Venus types will be used as case studies: Venus Pompeiana and Venus Pescatrice.

Case Study 1: Venus Pompeiana

Venus Pompeiana appears in 12 wall-paintings in Pompeii. Representations are assigned to this type if Venus is standing in full regalia, heavily draped in a mantle, and is holding her right arm across her breast. Various other features are common to Venus Pompeiana, but do not appear in every representation of this type: for instance, her left arm may lean on a rudder, she may hold a sceptre and/or a branch of myrtle or olive, she may also wear a crown, and she may be accompanied by cupids.

Aside from one example in Boscoreale (Boyce 1937: 99 no. 496), no examples of the Venus Pompeiana type have been discovered outside Pompeii. Many scholars have argued that the Venus Pompeiana type represents Venus as a statue and was based on a cult statue of Venus from Pompeii which no longer survives (Balsdon 1951: 6; Moormann 1988: 74; Sampaolo in PPM VII: 374). Moormann (1988: 74) previously noted that Venus Pompeiana often appears as a statue, but the quantitative approach adopted in this paper allows us to state more positively that Venus Pompeiana appears exclusively as a rigid statue, rather than taking an active part in scenes as a live character. For instance, on the façade of IX.7.7, even when Venus is represented on a quadriga drawn by elephants, she does not hold the reins or attempt to drive it, but simply stands stationary atop the quadriga (Figure 1). By adopting a quantitative approach to identify the number of narrative and non-narrative scenes in which Venus Pompeiana occurs, we find that Venus Pompeiana only appears once within a narrative scene (the wedding of Hercules and Juventas scene from VII.9.47, Figure 2). In this scene, Venus stands on a podium in a shrine as the wedding party passes by, rather than interacting with any of the other figures, suggesting that she represents a statue rather than actively participating in the narrative. Venus Pompeiana was thus always intended by the artist to represent a statue, even when she occurs within narrative scenes. Perhaps the Venus Pompeiana type did indeed originate from a cult statue of Venus at Pompeii, retaining the religious and ritual aspect of the original statue and having especial religious significance.

Figure 1

Venus Pompeiana on an elephant-drawn quadriga on the façade between IX.7.6 and IX.7.7, Pompeii (Photo: Sophie Hay. Reproduced with permission of the photographer).

Figure 2

Venus in a shrine at the wedding of Hercules and Juventas from VII.9.47, Pompeii (Archivio dei Disegni della Soprintendenza Archeologica di Napoli 734, drawn and painted by F. Morelli. Reproduced with the permission of the Museo Archeologico Nazionale di Napoli).

This proposed religious significance can be assessed using quantitative data. With regard to content, Venus Pompeiana is mostly represented with other deities. The majority of paintings featuring Venus Pompeiana (58%) include other deities within the painting or in direct association with it: six paintings of this type feature other deities within the same painting (in VI.11.8–10; VII.4.20; VII.9.47; VIII.3.10–12; IX.1.12; and IX.7.7), while in house V.4.3 she is depicted within one of several paintings featuring deities around a lararium niche. Venus Pompeiana is sometimes represented in scenes which depict a religious or ritual setting. Four of the 12 paintings of Venus Pompeiana depict shrines or altars within the painting: one depicts Venus in a shrine overlooking her outdoor domain, where Venus and her domain are emphasised as the subjects of the sole painting on the one curved wall of the room (in I.10.4); two depict Venus with an altar in the painting as part of offering scenes (in VI.11.8–10 and VII.4.20); and one depicts Venus within a shrine (in VII.9.47).

With regard to location, the majority of artworks featuring Venus Pompeiana in Pompeii (83%) occur where they could be seen by many people: one in a public area (a garden behind a shop, VII.4.20), three on façades (VIII.3.10–12; IX.7.1; and IX.7.7), and six in spaces which mainly functioned as reception areas in houses (in I.9.3; I.10.4; V.4.3; VI.8.3/5; VI.9.6/7; and VII.9.47). The remaining two artworks occur in spaces which mainly functioned as non-reception areas of houses (a cooking area in IX.1.12 and a domestic bakery in VI.11.8–10). These are lararium paintings, however, which may explain their location in areas that mainly functioned as non-reception areas (for an explanation of how areas in houses can be grouped as ‘predominantly reception’ and ‘predominantly non-reception’ areas, rather than grouped as ‘public’ and ‘private’ areas, see Brain 2017).

Façade paintings featuring deities could be used to highlight the patron or favoured deities of the owner and to protect the building (especially at entrances) in a visible way. There are 11 representations of Venus on façades in Pompeii. Three of these (27%) represent Venus Pompeiana. This was the most popular Venus type to occur on façades, as just two other Venus types occur more than once on façades (each occurring twice: ‘Venus Leaning on a Pillar’ and ‘Venus with a Cupid on her Shoulder’). This suggests that Venus Pompeiana was considered by the Pompeians to be the most appropriate Venus type when representing the goddess on façades, possibly for protection.

Six representations of Venus occur on aediculae or in lararium paintings. Four of these six depict Venus Pompeiana (in V.4.3; VI.11.8–10; VII.4.20; and IX.1.12). As such, this is the most common Venus type to occur on aediculae or in lararium paintings. Additionally, Venus Pompeiana is the only Venus type to occur on aediculae or in lararium paintings within domestic contexts. This suggests that this type was particularly favoured in ritual contexts (especially in the domestic sphere) compared with other Venus types.

Four of the 13 representations of Venus in the upper zone in Pompeii (31%) represent Venus Pompeiana (in I.9.3; VII.4.20; VII.9.47; and VIII.3.10–12). Just three other Venus types occur more than once in the upper zone, all of which occur just twice (‘Venus Caricature’, both in the same frieze; ‘Venus and a Figure Touching/Embracing’; and the catch-all category ‘Venus Standing’). Venus Pompeiana is thus the most popular Venus type in the upper zone, which is even more significant given that this type is only represented 12 times in Pompeii and is not one of the most popular Venus types in the town. This type was clearly considered by the Pompeians to be the most appropriate Venus type for this location. Perhaps her popularity in the upper zone was intended to put her in a position to watch over inhabitants and passers-by from above, quite literally, and protect them. This is not to say that all upper zone representations of deities or Venus were intended for protection, however, it is likely for Venus Pompeiana given that this type represents a statue; only once appears in a narrative scene (where she overlooks a wedding as a statue in a shrine); often appears with other deities; is often found in paintings with a religious or ritual setting; and is the most popular Venus type to be represented on aediculae or in lararium paintings.

To summarise, previous scholars have argued that Venus Pompeiana represents Venus in her role as the patron deity of Pompeii, presumably because this type is almost exclusive to Pompeii (with one representation in nearby Boscoreale). Indeed, a quantitative approach that compares various categories makes it clear that Venus Pompeiana had particular religious and ritual significance to the Pompeians. With regard to content, this type is mostly represented with other deities, and is sometimes represented in scenes which depict a religious or ritual setting. With regard to location, this type is mostly located in areas viewed by many people, and is often located in ritual areas, or in areas which suggest a protective purpose. For instance, this is the most common Venus type on façades, in the upper zone, and on aediculae or in lararium paintings.

Case Study 2: Venus Pescatrice

Turning now to the second case study, there are 25 representations of Venus Pescatrice (Venus as a fisherwoman) in Pompeii. This type can be identified if the female figure is fishing, sitting on a rock, and semi-naked with her lower half covered in drapery. She typically uses one hand to support herself, holding a fishing rod in the other (Figure 3). She usually sits towards the right, often facing a cupid or accompanied by multiple cupids, who are themselves sometimes fishing. There is always landscape scenery in the background.

Figure 3

Venus Pescatrice from VI.14.28/33, Pompeii (Archivio dei Disegni della Soprintendenza Archeologica di Napoli 417, drawn by G. Discanno. Reproduced with the permission of the Museo Archeologico Nazionale di Napoli).

Venus Pescatrice scenes are so similar in composition that they may have all derived from the same original. However, the exact origin of this type is unknown. Scholars have suggested that it was either a sculpture or a Greek painting (see Allison and Sear 2002: 74), but there is no real evidence for either interpretation. The pose of Venus Pescatrice is known from representations of Aphrodite/Venus in other media, including two fifth-century BC coins from Eryx (Bernhart 1936: nos. 150–151) and a late third-century BC bowl or mirror from Mit Rahîne (Delivorrias 1984: no. 886; Pelizaeus-Museum 1921: 142 no. 1128; Figure 4); however, Venus is not fishing in any of these. The Venus Pescatrice type thus seems to have been derived from earlier representations of Aphrodite seated on rocks, with the fishing context adding a new element to the scene in Pompeii.

Figure 4

Aphrodite in a similar pose to Venus Pescatrice in a plaster cast of a late third-century BC bowl or mirror from Mit Rahîne (Pelizaeus-Museum 1921: 142, figure 55; in the public domain).

These paintings are so similar with regard to subject and pose that they clearly all depict the same woman fishing, though there is some variation. Previous scholars have identified the fisherwoman in these representations from Pompeii as Venus. By using quantitative data on the content of these artworks, we can identify whether or not this female figure actually represents Venus.

The presence of a cupid could identify the woman fishing as Venus, as cupids are amongst her attributes. However, fishing may have had more general erotic associations (Greek Anthology 5.67, 5.247, and 12.241; Apuleius, Apologia 27–41; Oppian, Halieutica esp. 4.1–438; Engemann 1969: 1009–1010; Davidson 1997: 8–11), so an ordinary fisherwoman could also appear alongside a cupid. Thus, the presence of a cupid cannot definitively identify the woman fishing as Venus. While the presence of a cupid may not be diagnostic, there is a clear preference for including at least one cupid in these scenes. There are 20 paintings of this female figure fishing where the presence or absence of cupids can be ascertained (as five paintings have deteriorated), and 90% of these 20 paintings include at least one cupid. Five paintings of this type also include a winged male in the background, probably Eros as an adult, given that he wears a floral crown and holds a branch, both attributes of Eros (Aristophanes, Acharnians 991–992; Plato, Symposium 196a–b). Depicting Eros alongside the woman fishing would be appropriate for Venus (he was one of her retinue), but this is not definitive due to the potential erotic symbolism of fishing.

The presence (or absence) of headwear on the fisherwoman can be ascertained in 16 paintings. In the majority (15 of 16), the fisherwoman wears some form of headwear, be it a crown, diadem, or hairband. The presence of crowns and diadems (the woman wears a crown in five representations and a diadem in four) suggest that the figure is a deity, queen, or other mythological figure, rather than just an ordinary fisherwoman. This is supported by the fact that the fisherwoman wears jewellery in 88% of paintings of this type, where this can be ascertained. Bracelets are the most common type of jewellery that she wears (depicted in 11 paintings), and notably in four paintings she wears the breast jewellery (or girdle) often associated with Venus/Aphrodite. This does not necessarily identify the fishing woman as Venus, as artists were inclined to ornament naked and semi-naked women with jewellery, perhaps to emphasise their erotic nature or their femininity. However, Venus is often described as wearing golden jewellery, so this may be an indicator of a Venus representation (Homeric Hymns 5 and 6; Ovid, Fasti 4.135–137).

Thus, it is difficult to determine whether these paintings featuring a woman fishing represent Venus or an ordinary fisherwoman, as the cupid and jewellery could indicate either, though the crown and Eros would perhaps be more indicative of Venus.

The vast majority of representations of Venus Pescatrice in Pompeii are located within domestic contexts (84%). Two Venus Pescatrice representations are from public contexts (a shop, VI.14.28/33, and the back room of a caupona, VI.10.1), one is from a domestic and public context (a caupona which later became part of a house, IX.8.3/7), and for another the domestic or public context cannot be determined as the building it originated from was not recorded.

Arnold de Vos (in PPM II: 1059) argues that all artworks featuring Venus Pescatrice occur in cubicula. This would suggest that artworks featuring this type exclusively occur in areas of the house that are slightly more private, as they lack the capacity for large gatherings and are closed rooms. This is demonstrably not the case, however, especially given the two paintings featuring Venus Pescatrice from public contexts in Pompeii. While the majority of representations of this type from domestic contexts are located in small, closed rooms (58% where the room type can be ascertained), the remainder (42%) are from more open or larger areas of the house, which were more likely to serve a predominantly reception purpose. While de Vos’s claim that artworks featuring Venus Pescatrice exclusively occur in cubicula is incorrect, it is clear that this was the preferred room type for these artworks. Interestingly, despite representing Venus as a fisherwoman, Venus Pescatrice is never found in atria or gardens, areas which may have included water features such as impluvia or fountains, appropriate settings for a fishing scene.

To summarise, a quantitative approach to paintings featuring Venus Pescatrice has highlighted three key points. First, the female figure can be identified as Venus when different categories are considered altogether, given the usual presence of at least one cupid with her (and sometimes Eros too) and the presence of headwear and jewellery. Second, previous scholars have argued that Venus Pescatrice occurs exclusively in cubicula, but we now know that this is not the case, with some representations of this type occurring in public areas and a substantial amount in other areas of houses. Third, despite representing Venus as a fisherwoman, this type does not occur in areas of the house with water features, such as atria with impluvia, and gardens with fountains.

Advantages of a Quantitative Approach

There are several advantages to adopting a quantitative approach. First, it enables artworks to be compared systematically. Rather than making a general point regarding a group of artworks and giving a few specific examples to illustrate it, an entire group of artworks can be easily compared. For instance, instead of stating that Venus Pompeiana is often depicted with other deities and giving a couple of examples to illustrate this, we can clearly state that six of the 12 paintings of Venus Pompeiana depict her with other deities within the same painting.

Second, it enables an evidence-led approach. By comparing these artworks more comprehensively using a quantitative approach, patterns may emerge that can then be interpreted, in contrast to the current approach where scholars research artworks, make conclusions based on incomplete data, and illustrate these with a few examples. For example, previous scholars have noticed that artworks featuring Venus often occur in small, closed rooms and from this have presumed that Venus, as goddess of love, would be particularly popular in these rooms (perhaps presuming these would be used as bedrooms), using only a few examples as evidence. However, they do not quantify how often she occurs in this room type or compare this with other room types. A quantitative approach demonstrates that just 23% of Venus representations are located in small, closed rooms within houses, and she actually appears far more often in larger rooms (43%).

Third, it enables arguments to be clearly supported by numerical data. Instead of drawing general conclusions from a group of artworks, the data produced by a quantitative approach enables conclusions to be clearly substantiated. For instance, previous scholars have argued that Venus Pompeiana represents Venus in her role as the patron deity of Pompeii, which can perhaps be partially supported by a quantitative approach, as this has highlighted that Venus Pompeiana had particular religious and ritual significance to the Pompeians. With regard to content, numerical data have revealed that this Venus type is mostly represented with other deities, and sometimes in scenes that depict a religious or ritual setting. With regard to location, this type is mostly located in areas viewed by many people, and often in ritual areas or areas which suggest a protective purpose (on façades, in the upper zone, and on aediculae or in lararium paintings).

Fourth, it allows other scholars to adapt and reuse data as the evidence to support conclusions is collected in a clear, comprehensive, and systematic way. As already noted, scholars of Roman art often make conclusions based on experience gained from years of research, and only give a few examples to illustrate their arguments. Personal knowledge is of course invaluable, especially for identifying and communicating general trends, and in the prose of a chapter it would be impractical for scholars to list every single example illustrating their point. However, such knowledge needs to be coupled with empirical evidence. Currently, scholars have to trust conclusions from those who have studied Roman art for decades when they note general patterns, as the full dataset is not accessible to them. Some of these generalisations gain authority with repetition but may not be accurate. Having a database means that other scholars can consult it; if they interpret a scene differently or want to use the information for a different analysis, they can adapt the database and make their own conclusions.

Challenges with Adopting a Quantitative Approach

There are also challenges with adopting a quantitative approach. Art is a creative and diverse medium, and given variations between artworks it can be difficult to set criteria to which all artworks conform. Although I created criteria for identifying Venus, further criteria were often needed in order to differentiate the goddess from other figures with similar iconographies, such as women at their toilet, nymphs, and hermaphrodites. Furthermore, criteria cannot cover every eventuality in art. For instance, in room n of VII.2.25, there is a wall-painting of a semi-naked female figure wearing a crown (which could be worn by Venus), surrounded by dolphins (attributes of Venus), and riding a sea-centaur (PPM VI: 715; PPM XI: 314–316). This could identify the figure as Venus, but the female figure appears to be holding a greave, with the sea-centaur holding a shield; rather than Venus, this is perhaps more likely to be Thetis (who acquired armour made by Vulcan for her son Achilles). Thus, some flexibility is needed. Furthermore, a quantitative approach is especially useful if one is studying all the artworks in a particular area, or depictions of a certain subject in a certain area. If one were studying Roman art more generally, or a range of locations with thousands of artworks altogether, it would be difficult to adopt such an approach.

Conclusion

This paper has demonstrated that a quantitative approach to Pompeian artworks can have important results as it enables a systematic approach, clarity of evidence for analyses through numerical data, reuse of the data by other scholars, and an evidence-led approach with fewer generalisations. Where it is not possible to adopt a quantitative approach, it is important that scholars articulate the limits of their work in order to avoid unsupported generalisations. Although the value of this approach has been demonstrated using artworks featuring one subject from one town, it could easily be applied to the study of other deities, other subjects, other areas, and artworks in other media. It is time we took a more ordered approach to artworks and considered them as we would other artefacts, rather than as a separate medium.

Abbreviations

Competing Interests

The author has no competing interests to declare.