ABSTRACT

All landscapes are, to a greater or lesser extent, cultural constructions; the bush, the veldt, the forest, field, and garden have all occupied the imaginative geography of the societies that engaged with them. Historians, geographers and art historians have internalized the idea that landscape is a message to be decoded,1 but this concept is only intermittently applied to Roman gardens, usually in the context of villa gardens which, as noted in the previous chapter, are part of the sum but not the whole of Roman garden space. From the emphasis placed on horti in literature it is clear that they were places eliciting a special set of responses. What is less evident is how these responses were produced. Our three main sources of evidence – textual, pictorial, and archaeological – do not entirely correspond with one another, confusing the perception of Roman garden space both as a literary and material construct. Consider, for example, the lush thicket in the garden painting of the House of the Golden Bracelet. is visualization of the garden bordered by a nearwilderness does not correlate to the physical evidence of the site with its controlled borders, straight paths, and neat beds, and neither version seems to bear a close relationship to the literary gardens of Martial or Vergil. Differences between the content, location, and use of these gardens are what constitute their cultural meaning, but the sheer variety of Roman gardens threatens to overwhelm any attempt to gauge the cultural meaning of individual sites.