Rivers as boundaries in the "Empire without end": An ancient perspective on a modern debate

(1)

Rivers as boundaries in the “Empire without end”

An ancient perspective on a modern debate

Fig. 1: Caesar’s Rhine Bridge, John Soane (1814).

Pelle Wijker

Student number: 10411453

Telephone number: ---

E-mail address: ---

Master’s thesis

Research Master’s, History

Universiteit van Amsterdam

Supervisor: prof. dr. E.A. Hemelrijk

Second reader: dr. J.A. van Rossum

14 July 2014

(2)

(3)

General introduction

Introduction

“Do rivers make good frontiers?” asked the British historian Boris Rankov, provocatively, in a lecture given at a congress of 'Roman frontier studies' in Pecs, Hungary in 2003.1 To his audience, the question will have been a familiar one; indeed, it has been asked and answered, sometimes explicitly and often implicitly, in scholarly works on the frontiers of the Roman Empire for several decades now. To understand why this question was foremost in their minds, one has only to glance at a map to see that the borders of the great Empire, as traditionally described, were for the greater part of its history located largely along three major rivers: the Rhine and Danube in the north, and, to a somewhat lesser extent, the Euphrates in the east. Only a number of small areas in the north and east (Britain across the sea and the so-called Agri Decumates between the Rhine and Danube), as well as the long African border on the Sahara desert, appear to have been 'dry frontiers'.

Too much stock should not perhaps be placed in these maps we find in schoolbooks and on Wikipedia. Although their function, meaning, and apparent veracity are obvious to the modern observer, it must not be forgotten that the Romans themselves did not have access to maps of this sort.2 As both a reason for and a consequence of this fact, their understanding of the world, their Empire, and its borders, may well have been very different from our own, and rather more complex than these lines on a map might suggest. Indeed, this is what many historians have argued over the past decades.3 Nevertheless, a great many passages in Roman texts do in fact refer, quite simply and explicitly, to the great rivers as demarcating the limits of the Empire.4 Perhaps most famous among these is the assertion by Tacitus that, thanks to the conquests made by Augustus, the Empire was “enclosed by the Ocean or distant rivers”.5 The complex debate on the river frontiers of the Roman Empire, in which this thesis is engaged, can perhaps be boiled down to the essential question of how seriously and how literally we ought to take statements like this.

Rankov's answer to his own titular question was a resounding “yes”: rivers do make good frontiers, and the Rhine and Danube were indeed the linear borders of the Empire. In other words: Tacitus was spot on. This may have come as somewhat of a surprise to his colleagues, however, because over the past few decades a fairly broad consensus has emerged in the field which holds that rivers are poor lines of defence, and that the great northern rivers functioned not as boundaries or military frontiers, but as lines of communication and transport in a broad and vaguely defined

1 Rankov (2005). See also Austin & Rankov (1995), p 173-180. 2 Nicolet (1991), Brodersen (2001), Salway (2001).

3 See below.

4 See especially Chapter II, “Political boundaries and the borders of the Empire”, below.

(6)

'frontier zone'.6 According to this theory, passages by Roman authors that seem to contradict this interpretation by referring to the Rhine or Danube as a meaningful boundary, like Tacitus' cited above, reflected little more than literary tropes, and any reference to rivers as boundaries were made out of convention and bureaucratic convenience – after all, the visible line of a river is simply an easy line to point to and describe – rather than having any stronger meaning.7

This consensus, which can still in some ways be recognized in Brian Campbell’s recent benchmark work on the subject of rivers in the Roman Empire, Rivers and the power of ancient

Rome,8 is the product of a debate that can be traced to the highly influential yet controversial work of American strategist Edward Luttwak in the 1970's. In his book The Grand Strategy of the Roman

Empire he attempts to discern a consistent frontier strategy in three centuries of Roman imperial

policy. The great rivers were a key part of this strategy; according to Luttwak, the emperors understood rivers to be rational military boundaries, and aiming to establish a system of 'scientific' boundaries, they expanded the Empire to the banks of Rhine, Danube, and Euphrates, but no further.9 The fortifications they constructed there were intended to accommodate a flexible and pre-emptive – as opposed to static – yet essentially defensive strategy. In Luttwak's system, these great rivers acted as definitive boundaries, and rationally so.10

Luttwak's work was criticized by historians on nearly all fronts, however. The very idea of a 'grand strategy' behind the wars and policies of the various emperors quickly fell out of favor, especially after an innovative study by Claude Nicolet in 1991, which demonstrated that the Romans did not possess the accurate geographic knowledge to implement any rational geopolitical strategy in the modern sense of the term.11 More importantly for the purposes of this thesis, though, his assertion that rivers were rational boundaries was firmly rejected. In 1987, V.A. Maxfield stated categorically that rivers are in fact “militarily weak”. Instead of operating as a natural line of defence or separation, they act as links between peoples on both sides: “they are highways which unite, not barriers which divide.”12 This maxim has taken on an almost canonical status in the historiography on the subject.

Not the least among those who followed it was C.R. Whittaker, whose work in the 1990's essentially established the current consensus.13 In his model, the great northern rivers were not truly

6 Formulated most definitively by Whittaker (1994, 2004), see below.

7 For a recent example of this analysis of the sources, see Campbell (2012), p 53, 63, 98-99.

8 See note 7, above. At other times Campbell is almost entirely in agreement with my argument, although he fails to draw clear conclusions in response to the debate. See Campbell (2012), p 186-199.

9 Luttwak (1976), p 60. 10 Luttwak (1976), p 78.

11 Nicolet (1991). See for example Whittaker (1994), p 66-67. 12 Maxfield (1987), cited in Rankov (2005), p 175.

(7)

boundaries at all; instead they were only a part of a broader ‘frontier zone’ that stretched well beyond the rivers themselves. Whittaker agrees with Maxfield, then, that rivers do not separate peoples, but bring them together. Moreover, in the military sphere they disappoint and cannot be held a defensive line – although, as Rankov points out, 14 he offers little in the way of evidence for this statement – and the legionary camps and other fortifications along their banks were intended not to defend the Empire’s borders from invasion, but to protect the river itself as a logistical asset. In Whittaker’s model, then, rivers were by no means useless in the grand scheme of things, but they were not boundaries; any indication to that purport in the sources reflects little more than

convenience, and should not be taken too seriously.

It is against this broadly shared understanding of the river frontiers of the Roman Empire, then, that Rankov rebelled. He explains that the difficulty of crossing a river is often underestimated – in fact, it is no easy task to, especially with a large and cumbersome army. Swimming is certainly not an option while carrying heavy weapons and armour, and to bring the whole army across on boats takes a long time, rendering it vulnerable to attack. The only viable option is to build a bridge across the river. This was a task at which the Romans were adept, but their enemies in the north less so. If the bridges were tightly controlled, then – and Rankov cites a number of sources

demonstrating Roman caution in this regard – rivers could function quite effectively as military boundaries.15 To Rankov, Roman authors like Tacitus, asserting that the Empire was “enclosed by rivers” should be taken quite seriously after all.

Although this all sounds very reasonable to me, and I am generally in agreement with Rankov, I am, like most historians, not a strategist or a military expert. Rankov and I draw the same conclusions, but in different ways – I have no real basis to support one appraisal of the defensive value of rivers over another. More importantly though, Rankov’s argument is still framed by a rationalist, modernist perspective, which presupposes that modern strategic theory is singular and universal, applicable and intelligible at all times throughout history. This is an essential

anachronism: whatever our understanding of strategy, however 'rational' it might be, the Romans themselves may well have seen things differently. Rankov himself, to his credit, clearly

acknowledges this,16 although I feel he does not entirely escape the issue. Perhaps this is why, a decade later, the consensus has yet to shift in his favour.

To my mind, the key to a real breakthrough in this debate can be found in a very short article

version, even before Luttwak – see Mann (1974). 14 Whittaker (1994), p 61. Rankov (2005), p 176.

15 Rankov cites Tacitus, Annals 1.69, Ammianus Marcellinus 24.3.9 (English transl. H.G. Bohn 1862), and Cassius Dio 68.13.5-6 (English transl. E. Cary 1914-1927).

(8)

by David Braund, dating back to 1996.17 Though he too believes rivers to be “militarily weak”,18 Braund, too, offers a dissenting voice. He invokes Roman ‘environmental psychology’, arguing that the cultural meaning of rivers, including a strong religious component, made them into boundaries in the eyes of the Romans. He concludes that “from a Roman perspective, rivers were indeed natural boundaries in a sense that includes their religiosity, their natural power and their tendency to divide and to bound. When modern strategists leap to the criticism that rivers are inadequate (even non-functional) boundaries and dividers, they miss much of the point which lies embedded in the environmental psychology of the Roman world.”19 In other words, only through a broader

understanding of the Roman perception of the world, their culture and the role of rivers therein, can the issue of river frontiers be properly tackled. Sources like Tacitus, which are direct expressions of this perception, should not be so casually dismissed.

Though Braund’s call to action, too, has so far gone unheeded, I aim to rectify this situation in this thesis, by elaborating on his theory and working out its argument in greater detail. In what follows, I shall attempt to demonstrate that he is right in emphasizing the specifically Roman cultural perspective in answering the question of rivers as boundaries to the Roman Empire. With this perspective in mind, it shall become apparent that the Romans did in fact have a strong tendency to associate rivers, great and small, with boundaries of various sorts, and that these boundaries were infused with both a religious and a military value. As a consequence, it was perfectly rational for them – regardless of whether it seems rational to us – to build the limes, their borders, along the natural boundaries of Rhine, Danube, and Euphrates. In addition, I will

complement Braund’s vision with an analysis of what such a boundary meant and represented in the expansive and imperialistic world of the self-styled “Empire without end”.20

What this thesis will not be is a systematic exposition of the historical development of the frontiers of the Roman Empire and the campaigns that led to their establishment; I am not myself concerned with the issue of ‘grand strategy’ or the reasons the Empire failed to expand further than it did. Nor will it be a detailed analysis of the physical configuration of the frontier, its forts, camps, armies, and fleets. I am interested instead in how the Romans, or at least those among them who wrote the sources available to us, thought about these rivers and their status as boundary, and the cultural background of these patterns of thought. Their historical actions and physical constructions can tell us something of these cultural notions and beliefs as well, of course, and they will not be entirely neglected, but my emphasis will be on the literary sources, like Tacitus, which have clear

17 Braund (1996).

18 Braund (1996), p 43, citing Maxfield. 19 Braund (1996), p 47.

(9)

and revealing views on the subject that are, to my mind, far too casually dismissed within the current consensus. Above all, then, my aim for this thesis is to promote a new perspective on the issue of river boundaries – or more to the point, an ancient one, that restores these sources to their proper place and context.

The structure of this thesis is as follows. In the remaining sections of this introduction, I shall explain my approach to the sources, provide a number of key definitions, and elaborate briefly but crucially on the nature of 'landscape' and how our perception of it is shaped by our cultural background. It is with this understanding that I subsequently venture into the main body of the thesis, which can be divided into two parts. The first and larger section comprises two chapters dealing generally with the Roman perception of rivers as boundaries in different yet interconnected ways. Chapter I is concerned with the cultural background of religion and the divinity of rivers in general, while Chapter II deals with the military uses of rivers and their role as official political boundaries, with an obviously somewhat stronger focus on the three frontier rivers. Together, these chapters intend to demonstrate and explain the fact the Romans believed rivers were natural

boundaries. Following this exposition of my general theory, I will, in the second part (Chapter III), test its merit by application to a specific case study: the river Rhine. In so doing, I will illustrate my argument further and simultaneously position it more pointedly within the broader debate, arguing that the Rhine was in fact understood to be a boundary to the Roman Empire.

The problem of sources

It is difficult, if not impossible, in the field of ancient history to add anything new in the way of sources; the relatively small number of our sources, and their inherent limitations, are an inescapable reality of our discipline. Certainly on a subject as well-worn as the borders of the Roman Empire, it will be very hard to introduce any relevant, untapped textual sources (for these are my primary concern), and I will not do so here. This is an essentially argumentative, rather than investigative, thesis, and it does not introduce any entirely new facts or pieces of evidence, though it does aim to widen the scope of the debate by emphasizing the broader cultural perception of rivers in the Roman world, in concordance with Braund’s article. Most importantly however, this thesis calls for a reappraisal and a restoration of the evidence already at our disposal.

The evidence under review here is of a highly scattered nature. Though there are a few outlying sources from the late Empire and – especially on religious matters – the early Republic, it is centred mostly around the very end of Republican times and the first two centuries of the

Imperial (and Common) era. This is due to the simple facts that the borders of the Empire that lie at the heart of this debate were formed during this time, and that the majority of Roman literary

(10)

sources in general date from this period. Of course, there are no Roman texts available to us that are concerned solely with the subject of rivers and boundaries, giving a full record of the limits of the Empire, how these frontiers functioned, what the function of rivers was in this system, and whether or not these rivers were seen as natural boundaries. In the absence of such a treatise, we are reliant on brief, scattered passages from all sorts of texts that happen to mention rivers in this context. In most cases, I was led to these passages through the citations in previous historical work on the subject; Campbell’s recent book, dealing with virtually all aspects of the river’s place in the Roman world, has been especially useful in this regard.21 I have attempted in all cases to convey the context of these passages as well as possible without being overly tangential.

Lastly, it must be acknowledged that 'the Roman' did not of course exist. Any work of cultural history or analysis is to some extent impressionistic and reductionist; culture is too large and complicated a thing to ever fully describe or even grasp, and no culture is monolithic. There is always room for conflicting visions, voices and attitudes. Moreover, for Roman history in particular the sources we have are few and highly skewed towards a small part of Roman society: virtually all of them are products of the urban elite. That should not be taken to mean, however, that nothing can be said about culture. It is certainly unavoidable that within the space afforded by this thesis I can offer only fragments of the full story of the Roman view of the world, its rivers, and its boundaries. Still, I believe these fragments may, taken together and understood in their proper context,

demonstrate certain tendencies in Roman thought on landscape that help to explain why their Empire took on the shape it did.

Moreover, the image these various sources and passages convey of Roman beliefs about rivers seems to me to be a highly consistent one. It is very telling that those historians who argue that rivers are not, and were not understood to be, natural boundaries, and that the Rhine, Danube, and Euphrates were not actually the borders of the Empire at all, usually fail to cite any Roman authors to support this view.22 To the contrary, as I will show, there is a large body of Roman sources that clearly and explicitly declare the opposite: the Empire was bounded, in a number of ways, by rivers, and this appeared entirely rational to their mind. Indeed, although the authors of our sources were, due to their socio-economic background, by no means representative for the whole of Roman society, it can be inferred that their views and arguments at least seemed credible and reasonable to their audience.

21 Campbell (2012). 22 See note 14, above.

(11)

Boundaries and their definition

Let us start with a basic definition: a boundary is an imaginary line in the landscape that demarcates the limits of a certain area. This line may have a very well-defined practical purpose and be marked by constructions like walls, border posts, or boundary stones, or it may be nothing more than a wholly abstract line on a map, imperceptible to the eye. In some cases it might coincide with a pre-existing natural feature, such as a river or a mountain range, but even then, it is crucial to

understand that the boundary line itself is in all cases drawn by human hands and the result of human choice. As such, the notion of a natural boundary – a line shaped by nature itself to excel as a boundary – becomes a complicated one. After all, in the strictest sense, natural boundaries do not exist;23 it is not universally self-apparent that rivers should be boundaries, and a boundary line in itself, even one based on natural features, exists and has meaning only in people’s minds.

Yet it is precisely this mental reality that I am concerned with in this thesis. I do not mean to categorically prove that rivers are universally natural boundaries – an impossible task – but rather to show that they were considered to be natural boundaries by the Romans, or, in other words, that the authors of our sources had a general tendency to associate rivers with boundaries, and that it seemed 'natural' to them to do so. It is in this context that my use of the term natural boundary should be understood. The same applies to the related concept of the rational boundary.24 I am not interested, as Luttwak and Rankov are, in the 'pure' logic of the strategic pros and cons of choosing a river for a boundary, and whether this is as such a 'rational' choice, but only in what Roman sources reveal about these choices and the reasons for making them. When I myself make use of the term rational boundary, then, it is not a supposed universal rationalism I am referring to, but the particularly Roman rationalism or way of thinking revealed to us by these sources.

Lastly, I make reference throughout this thesis to three specific types of boundary: the

religious boundary, the military boundary, and the political boundary, though these terms may

appear self-explanatory. The first can be defined quite simply as a boundary infused with meaning of a religious nature. Perhaps it defines the limits of a sacred space, or the line itself carries religious significance; supernatural phenomena may be connected to it, and rituals may be involved in its crossing. A military boundary, then, is a boundary line with military significance and a possibly defensive function. I use the term frontier as essentially synonymous with military boundary.25 Lastly, a political boundary, which might in many cases also be referred to simply as a border, is a

23 A fact acknowledged by scholars on both sides of the debate. For example, see Mann (1974), p 513, and Rankov (2005), p 176.

24 Luttwak (1976) uses the term scientific frontier in essentially the same way, but I do not use this term unless referring to Luttwak’s theories.

(12)

boundary which separates different political or administrative territories or entities from each other. In the main body of this thesis I will argue that to Roman eyes certain important rivers fulfilled several of these functions at once, and the three categories were in fact intertwined.

Landscape and culture

Such a view of boundaries can only be sustained by understanding landscape to be a construct of human imagination and narrative. And it is; although the rivers, mountains and valleys, and even the creatures that dwell there are of course a physical part of reality, the meaning they provide to an observer is culturally determined. Two people of different cultural backgrounds might look at Mount Everest, for example, and see very different things: where the Westerner might consider it a dangerous challenge – something to be overcome by the brave and adventurous – the native

Himalayan perhaps sees little more than a massive rock, unfit for farming or habitation. Naturally, their subsequent actions towards this feature of the landscape will differ as well. The Westerner may choose to climb and conquer it, and tell all who will listen about his heroics – or at least marvel at the accomplishments of those more brave than he. The Himalayan, meanwhile, might join such an expedition for the pay, but is more likely simply to laugh at this absurd pastime and stay safely at home.

The realization that the landscape and the way we, as a species, make use of it is not a purely physical reality seen through a purely rational lens has many repercussions, and the principle has found applications in many fields in recent decades. History is among these fields, of course; Simon Schama’s Landscape and memory is a classic study of Western thought on landscape.26

Anthropology,27 geography, 28 and archaeology have benefited from this insight as well.29 Still, to my mind this crucial understanding has not been applied with the necessary rigor in the debate over the river frontiers of the Roman Empire. Certainly, it has been convincingly argued that the

Romans' lack of precise geographical knowledge – which made them perceive the landscape in ways different from our own – prevented them from adopting a unified 'grand strategy' as Luttwak had seen it.30 But this and other arguments on both sides of the debate are still framed by a modern Western perspective, and the deeper issue has rarely been addressed: the question of whether rivers made sense as boundaries for the Roman Empire cannot be given a meaningful answer solely by modern rationalist geography, which can tell us the measure of a river but not its meaning. It is only

26 Schama (1995). For Roman history, the essential work was done by Nicolet (1991). See also Beagon (1992) and Spencer (2011).

27 Kommers (1994) has been my prime inspiration, see below. 28 Ingold (1993) and Cosgrove (1984).

29 Derks (1998). 30 Nicolet (1991).

(13)

within the context of Roman culture that the matter can be truly understood.

The same point is excellently made in a wholly different context by the Dutch anthropologist Jean Kommers, in an article on the geography of the Australian Aborigines.31 Kommers proposes the term 'mythical geography' for their culturally determined, narrative-based understanding of the landscape, utterly different from our own, and laments the fact that though anthropologists often acknowledge these differences between Western and native conceptions of landscape and geography, their analysis of these differences is made entirely from the perspective of 'objective' Western geography. They apply an essential dichotomy between 'fact' and 'fiction'; only that which corresponds with Western notions is regarded as true geography, while the rest is relegated to other cultural domains – religion, for instance – and so severed from their original context. As a

consequence, the essential connection and continuity between these notions is lost.

It seems to me that the very same mistake is made by those who argue that the Roman Empire was not in fact bounded by the great rivers of Rhine, Danube, and Euphrates because, based on our modern understanding of geography and strategy, rivers are not rational boundaries. They take into account only those matters which correspond to this modern point of view – the 'rational' and 'objective' strategic factors – and disregard the other, decidedly un-modern meanings the

Romans attached to rivers. In this way, they project their own rationalism onto the past, pushing out the less relatable aspects of the Roman worldview and understanding of the landscape – most prominently their religious beliefs concerning the natural world. It is precisely these 'other matters' that I will elaborate on in the first chapter of this thesis, as it is my contention that the role and function of rivers and boundaries within the Roman view of the world cannot be understood without them.

(14)

(15)

I. Rivers in Roman religion

Introduction

All things considered, rivers are remarkable things, or in the words of Pliny the Elder, “a truly wondrous provision of nature”:32 they spring from the depths of the earth, often in remote mountainous regions, and spew forth endless quantities of water without ever emptying, as if by magic. The violent noise of their stream, rushing towards the sea, might remind the listener of a roaring bull, and stands in stark contrast to their great visual beauty and serenity, often remarked upon by poets. Yet they possess power beyond mere poetry: rivers bring life, fertility, and sustenance to the communities living on their banks. On the other hand, in some circumstances rivers can be a destructive force too, ravaging the neighbouring lands, destroying their crops and drowning their inhabitants. It is not be wondered at, then, that these mysterious and unpredictable waters have often come to be associated with the supernatural and divine. Indeed, even today some of the world’s rivers, like the Ganges in India and Bangladesh, are revered as sacred in ways that may well have been remarkably familiar to a Roman.

If we are to know if rivers were natural boundaries in Roman eyes, we must first understand the cultural meanings they attached to rivers in general. There are many possible avenues of

investigation here, but I have chosen to focus on one particular aspect: their religious views on rivers. Roman thought on rivers, in whatever context, cannot in my view be properly understood without recourse to this religious aspect. More to the point, to ask the question if rivers were considered to be natural boundaries is to ask what the nature of a river was seen to be, and to a Roman, a highly significant factor in answering this question would certainly have been the fact that rivers were divine. In this chapter, I will explore this phenomenon, the supernatural characteristics associated with rivers, and their worship. As such, this chapter is detached somewhat from the rest of this thesis, which is more pointedly concerned with river boundaries. Additionally, there is a lot to cover in this chapter, and a certain measure of generalization and lack of nuance is sadly

unavoidable. Still, I think the cultural background provided here allows us to better understand the issues discussed in what follows.

This chapter is subdivided into three paragraphs. In the first two, I shall discuss the Roman religious conception and worship of rivers along general lines. The first paragraph concerns the role of rivers or their waters in ritual and cult practices of a broad nature, the healing power of river water, and the more specific sacred character of springs, while the second paragraph covers river gods, the deities associated with the river itself, both in their appearance and function in art and

(16)

mythology, and, more importantly, their actual worship in official state religion as well as local cults. The third paragraph deals with the religious significance of crossing a river, and the rituals associated with this act, which illustrate the crucial connection between the divinity of rivers and their function as boundaries.

Sacred water

In order to gain some understanding of the way rivers functioned as boundaries in the Roman world, it is vital to take into account their religious function. Before turning to the worship of the rivers as gods in their own right,33 let us first examine the role of rivers in religion in a broader sense. Rivers and their waters had a role to play in the cult of other gods, too, and their sacred character was perhaps not wholly bound to their own divinity. In addition, we cannot neglect to note the specific character of springs, which are of course inextricably linked to rivers and were often the main focus of worship, yet have a number of characteristics that separate them from the course of the river itself. Indeed, the subject of springs is such a compelling and well-attested one, that it certainly deserves a separate treatment. Sadly, as my interest lies primarily with the boundary formed by the river's course, I can only offer a brief overview here of the matter of sacred springs.34

Water had an important role to play in the cult activities of many temples and deities, even those with no obvious thematic connection to rivers, springs, or the sea. This importance is to be found mainly in its use as a cleansing agent. A number of myths related by Ovid in his Fasti,

invaluable as ever for the student of Roman religion, will serve to illustrate this phenomenon. In the first, the poet tells the tale of Rome's foundation, beginning with the rape of Rhea Silvia, a Vestal virgin, by Mars.35 The reader is introduced to Silvia, carrying a pitcher, as she approaches a river.36 Although the details of her task remain elusive, Ovid informs us that she is there “[seeking] water to wash sacred things”.37 This detail is not important for the tale that follows; it would seem that Ovid simply felt the need to provide a narrative reason for Silvia to be alone by the waterside. It is precisely because it is used in this way, however, that we may infer that the act of drawing river water for ritual purposes would appear to his audience a plausible and otherwise unremarkable one. It seems likely, then, that river water performed a purificatory function in the rites of Vesta in reality as well.

A second example provided by the Fasti is a rather more obscure myth concerning three

33 See “River gods”, below.

34 Aldhouse-Green (1999) is a thorough and accessible exploration of one such spring. 35 Ovid, Fasti 3.9-78 (English transl. A.S. Kline 2004).

36 The fact that it is a river or stream, rather than a source or well, is evidenced by the use of the word ripa, river bank, in line 13.

(17)

closely connected constellations known as the Raven, the Snake, and the Cup, and apparently intended to explain these names.38 According to this myth, Apollo is preparing a feast for Jupiter, and gives his servant, the raven, a cup and orders it to run an errand for him. However, the hungry raven is distracted by a fig tree, waiting patiently for the fruit to ripen and forgetting his orders. When the bird finally returns to Apollo it carries a snake in its beak, blaming the creature for the delay. Apollo sees through the lie, of course, and curses the raven; as a reminder, the three

constellations are placed in the sky. Important for my purposes here is the errand Apollo sends the raven on: he orders the bird to “bring a little water from the running stream”, “so nothing delays the sacred rites”.39 We are not told the exact nature of these rites, but apparently they could not

commence without river water, and a similar purificatory use seems probable. It is again the offhanded manner in which the use of water in ritual is referred to – it is used as little more than a plot device – that demonstrates how natural and commonsensical such a use must have seemed to Ovid's audience, and how ubiquitous it must have been.

A number of rituals associated with the Tiber and dating from the earliest period of Roman history also involve the purification of ritually unclean things, but in an opposite manner.40 Firstly, at the end of the Vestalia, a festival in honour of Vesta held in June, the waste from the Temple of Vesta, accumulated throughout the year, was deposited into the Tiber and swept away to sea. Ovid mentions this practice as well.41 Secondly, an obscure ritual was performed annually in either March or May, involving the ritual offering of Argei, effigies in the shape of men, apparently representing the people of Rome, to the Tiber. The effigies were collected from the various sections of the city in a procession, and thrown off a bridge into the river in a solemn ceremony presided over by the chief priests. The precise meaning and background of this rite is unclear. The Greek historian Dionysius of Halicarnassus, writing around the time of Augustus, reports that the ritual was still being performed in his day, and attributed its origin to a myth of Hercules coming to Rome, abolishing human sacrifice, and instituting the annual offering of these effigies to the Tiber as a replacement.42

Victims required cleansing before the ritual of sacrifice could commence, as did people who, prior to their participating in a ritual or entering a sacred place, had been ritually 'contaminated'.43 Not just any water would do for this sacred purpose, however; in order for it to function as a purificatory agent, it had to be 'alive'. That is to say: it had to be running water from a natural

38 Ovid, Fasti 2.243-266.

39 Ovid, Fasti 2.249: ne quid pia sacra moretur. 40 Campbell (2012), p 141 for both of these rituals. 41 Ovid, Fasti 6.226-227, 6.711-714.

42 Dionysius of Halicarnassus, Roman Antiquities 1.38 (English transl. E. Cary 1937-1950). Ovid, Fasti 5.622-660 also refers to this ceremony.

(18)

source, as opposed to channels, aqueducts or cisterns.44 We have seen this evidenced already in the myth of the raven, as Apollo explicitly instructs the bird to fetch “water from the running stream”.45 Similarly, Tacitus, in writing of a restoration of the ruined Temple of the Capitoline Triad under the emperor Vespasian, mentions that prior to the traditional dedicatory sacrifice, the entire temple site was sprinkled with “water drawn from fountains and streams” by the Vestal virgins.46 It appears, then, that not all water was by definition sacred; in itself it was a lifeless substance much like any other. Only when it flowed of its own accord through the landscape or from the earth was it seen to be infused with divine and cleansing power.

Incidentally, this association of water with ritual purification is also recognizable, in a more abstract sense, in a topos common to many myths about river deities and other gods:47 a hero drowning in a river or dying near it and being reborn as a god. A famous example of this

phenomenon is the deification of Hadrian’s lover Antinous, who drowned in the Nile. Less well-known is a story told of Aeneas, who vanished after a battle near the river Numicius and was

presumed dead. Some of his compatriots believed he had ascended to godhood and built a shrine for him, “the father and god of this place, who presides over the waters of the river”.48 In these myths, a flawed, mortal human being is purified and cleansed of his imperfections by the river, and turned into a divine entity free from ritual contamination. In a recent article, Rabun Taylor has argued that this cleansing occurs simultaneously on a personal and a collective scale: in a sense, the hero is sacrificed to the water in order to purify the entire community.49

The use of water in cleansing rituals was almost certainly the primary function of rivers and springs in Roman cult practice. Aside from its use to wash away ritual contamination, however, running water was also popularly believed to have healing properties. Cicero famously wrote, apparently quoting an old proverb, that “as long as a man is at the waters he is never dead”.50 The Greek orator Aelius Aristides, who lived during the second century CE and suffered from poor health throughout his life, was a frequent visitor to rivers and springs. In his Sacred Tales, a chronicle of his continuous quest for treatment for his various illnesses, he mentions no less than fifteen such visits.51 Aelius seems to have placed a particular trust in the curative effects of river bathing. Often, specific rivers were recommended to him for this purpose, and Aelius explicitly offers homage to the river Aisepos in Anatolia. This was clearly not just the behaviour of a lone

44 Wissowa (1912), p 219. See also note 3 on the same page for additional references. 45 Ovid, Fasti 2.250.

46 Tacitus, Histories 4.53 (English transl. C.H. Moore 1925-1937). 47 For river deities, see the following paragraph.

48 Dionysius of Halicarnassus, Roman Antiquities 1.64.

49 Taylor (2009), with several other examples of this phenomenon as well.

50 Cicero, On the Orator 2.274 (English transl. J.S. Watson 1860): quamdiu ad aquas fuit, numquam est emortuus. 51 Aelius Aristides, Sacred Tales (English transl. C.A. Behr 1981-1986).

(19)

eccentric: to Aelius' dismay the springs at Cyzicus, also in Anatolia, were highly crowded.52 It must be admitted, however, that Roman opinions varied on the cause of these healing properties. For many, they were obviously tied to the sacred nature of the water or the divinity associated with it. It was, after all, precisely its life-giving powers that linked water to the gods in the first place.53 This popular belief is clearly demonstrated by the votive objects often found at springs, which will be discussed below. For Aelius Aristides, too, the curative properties of rivers and springs were connected to the gods and spirits inhabiting or watching over them; he shows genuine piety in thanking the rivers, nymphs and associated deities. Moreover, his aquatic treatments were divinely inspired by Asclepius, the god of medicine, who repeatedly appeared to him in a series of dreams.54 The Republican scholar Varro, too, writing two centuries earlier, notes that “many sick persons are wont to seek water from the spring” of Juturna, a Roman water-deity famed for her healing powers.55

On the other side of the argument we find the more scientifically-minded Vitruvius, who argues that the different qualities of rivers and springs are caused by the varying nature of the soil; some minerals were poisonous, while others were healthy.56 Pliny, in his Natural History, takes a similar stance;57 he believes also that the temperature of the water is key to its healing properties: lukewarm springs and rivers were especially potent.58 It is crucial to note, however, that neither author actually dismisses the divine nature of these waters,59 or their seemingly supernatural effects, and Pliny acritically relates several stories of apparently miraculous healing connected with springs and rivers, usually associating specific waters with specific ailments.60 The testimony of men like Pliny, a member of the military and political elite, and a personal friend of emperor Vespasian, demonstrates that it would be far too simplistic to argue that the belief in the supernatural curative power of rivers and springs was a purely local affair for those of lower social backgrounds.

Up to this point, I have for the most part treated rivers and their springs as interchangeable. In reality, the situation was somewhat more complex. Most springs were of course connected to rivers by virtue of being their source, but in many ways they were religious sites in their own right, more so than any other location along the river's course. In Roman Gaul, cult places related to rivers were most frequently located at the spring, with the crossings and confluences of rivers only a

52 Aelius Aristides, Sacred Tales 5.11-13. Cited in Campbell (2012), p 337-338. 53 Derks (1998), p 141.

54 Aelius Aristides, Sacred Tales 2.17, 2.18, 2.45, 2.51. Cited in Campbell (2012), p 337-338. 55 Varro, On the Latin Language 5.71 (English transl. R.G. Kent 1938).

56 Vitruvius, On Architecture 8.3.1-28. Cited in Campbell (2012), p 338-341. 57 Pliny, Natural History 31.28.

58 Pliny, Natural History 31.2.

59 Pliny, Natural History 31.2 tells us that rivers and springs “augment the number of the divinities”. 60 Pliny, Natural History 31.3-12.

(20)

distant second and third in popularity.61 The situation in other parts of the Empire appears to have been very similar; as Servius, a fourth- to fifth-century CE grammarian, wrote in a commentary on the Aeneid: “all springs are sacred.”62 In some instances, the worship at these springs was clearly directed towards the river itself, as in the case of the sanctuaries at the source of the Seine and Marne, where inscriptions were found dedicated to the goddesses Sequana and Matrona,

respectively.63 In most cases, however, offerings were made more generally to the nymphs of the place, or the fons (spring) itself. In these cases, it might be said that the spring was a religious site in its own right.

The worship of or at springs in Rome dates back to the archaic period. The city itself was home to a number of wells and springs, and from the earliest times each was associated with a certain deity; the Camenae, for example, who watched over a spring near the Porta Capena, or goddesses like Carmenta or Juturna (who I have mentioned once already). Over time, it seems a certain generalization took place, while these spring deities for the most part faded into obscurity or transformed into something with a broader applicability. The Camenae soon came to be identified with the Muses, imported from Greece, and lost their uniquely aquatic nature. So did Carmenta, who became a goddess of childbirth, while Juturna became a healing deity and was even granted her own temple.64 At the same time, the fontes deities of specific spring were mostly replaced by the Greek nymphs.65

It is to these nymphs, then, that most offerings at springs were dedicated. The religious practice associated with springs was mostly votive in character: individuals gave offerings to the water to repay the deities for services provided or miracles worked. In some places a spring was marked by a sanctuary or even a temple; remarkably, this was not just the case for the sources of major rivers like the Seine,66 but often for much smaller and seemingly unimportant brooks as well, as was the case at the large complex found at Genainville, France.67 In most cases, however, a spring was not marked by any building, and can only be identified as a cult place due to the often huge amounts of votive objects found in or around the water. Coins were often tossed into the spring as a simple offering, much like people today throw coins into fountains for good luck.68 More elaborate objects were also offered, though: votive altars were set up, and sculptures of the

61 Derks (1998), p 138.

62 Servius, Commentary on Virgil's Aeneid 7.84. Cited in Campbell (2012), p 128.

63 Sequana: CIL 13 2858, 2861-2865, 11575. See also Aldhouse-Green (1999). Matrona: CIL 13 5674. 64 Coarelli (2007), p 275-280.

65 See Wissowa (1912), p 219-221. 66 Aldhouse-Green (1999). 67 Mitard (1993).

(21)

nymphs or other deities were submerged in the spring.69

Lastly, many springs, especially hot springs, became the focus of a somewhat more secular popularity and were visited by thousands for their curative properties. These spas, known to the Romans as simply aquae, could be found all over the Empire, but with a high concentration in Italy, Greece, and Asia Minor.70 Springs seem to have been more strongly associated with healing than rivers – the preference of Aelius Aristides for rivers, noted earlier, appears to be somewhat of an anomaly in this regard.71 The presence of a nearby spa could be an immense boost to a town's economy, and in some cases the springs became the heart of a tourist industry; the hot springs at Baiae in Campania, for example, were a highly fashionable destination for the Roman elite in the late Republic at least.72 Clearly, curative springs were a prominent and popular feature of the Roman landscape, as is also indicated by the high number of aquae marked on the Peutinger Map – even in the fourth century CE, they were apparently thought worth mentioning amongst the cities of the world.73

In the following paragraph, we shall turn to the subject of rivers as gods in their own right, but before we do so it is important to note that other, more well-known deities were on occasion associated with rivers as well. In some instances the connection between the god and the water is obvious, as in the case of Asclepius, the god of medicine, whose cult was often linked with springs or rivers for their healing properties. Significantly, the Temple of Asclepius in Rome was built on Tiber Island, where even today a hospital can be found. The connection is rather less obvious for four rings found in the Rhine near Mainz, clearly thrown into the river as a ritual act. The

inscriptions on the rings identify them as gifts dedicated to Heracles, Mercury, and Mars.74 Why these gods should receive these gifts through a river is unclear to me, but perhaps this curious fact will serve to illustrate the sacred power possessed by all rivers, independent of its own divinity.

River gods

As we have seen, rivers, their sources, and their waters were held by the Romans to be sacred. But for a great number of these rivers, the religious significance went one crucial step further: these rivers were themselves gods, and although they are only bit players in most of the better-known myths and rarely have temples dedicated to them, their worship was very much a genuine part of lived religion all throughout the Empire. In this paragraph I shall sketch out some general

69 Derks (1998), p 138-139.

70 Campbell (2012), p 348-351 provides a map, though it is incomplete. 71 Campbell (2012), p 346.

72 As evidenced by Horace, Letters 1.15 (English transl. C. Smart 2004).

73 Allen (2003). In total, 55 spas are marked on the map, of which almost half are in Italy. 74 See Wegner (1976), p 78. CIL 13, 10024 207, 10024 13a, 10024 13c, 10024 18a.

(22)

characteristics of these deities and give a brief overview of their appearance and function in mythology and iconography, thereby illustrating some of the subjects and themes with which they were commonly associated. Subsequently, I shall turn to their actual worship and demonstrate how this phenomenon was shared across many layers of society.

To start with the basics, both river gods and goddesses are attested, although the ratio between the two genders seems to have varied from region to region. In Greece, for example, virtually all river deities were male, in contrast to the exclusively female nymphs usually associated with springs.75 In Gaul, on the other hand, the distribution of genders was somewhat more even: river goddesses were perhaps more common, although male deities were certainly not unknown, either.76 All river gods could, of course, be called upon to provide safe passage across the water, assure a bountiful harvest on its banks, and usually to cure the ill and ensure good health for

everyone. Not all were exactly alike, however; certainly in the mythological and literary spheres the specific characteristics of individual river deities are clearly established. Although rivers might be respected for their size and strength, these attributes were not necessarily directly proportional to the prominence of the river god.

The worship of rivers as gods was not a uniquely Roman phenomenon. On the contrary: this worship, in one form or another, seems to have been virtually universal among the peoples of the ancient Mediterranean and western Europe. Greek stories of river gods abound,77 and the Nile had held a divine status in Egypt and abroad for thousands of years. Moreover, archaeological finds indicate that the Celts inhabiting Gaul and Britain, as well as the Germans across the Rhine, worshiped river spirits too, although the details of their religious beliefs are unclear.78 As a consequence, when the Empire spread out across the region, Roman and local religious views on rivers could often be reconciled with very little trouble. It should also be noted in this context that the Romans did not usually rename the rivers of the lands they conquered, opting instead to simply adopt a Latinized version of their (the river's, and its associated deity's) local name.

These common cultural threads were not, of course, entirely coincidental. In part, at least, they were causally linked: Roman conceptions of river deities were strongly influenced by foreign cultures, most prominently of course by the Greeks.79 This influence made itself felt as early as the seventh century BCE, in the time of the kings, through the colonies of Magna Graecia in southern

75 Brewster (1997).

76 Sequana (Seine) and Matrona (Marne) were female, see note 64 above. Rhenus (Rhine) was male, though, and shall be discussed in detail in Chapter III, “Pater Rhenus”.

77 For a thorough and often entertaining overview of the river gods of Greece and their associated myths and cults, see Brewster (1997).

78 Campbell (2012), p 139-140.

(23)

Italy. Before this influence took hold, the Romans had worshipped the numina of rivers, in similar ways perhaps as the Celts and Germans, but did not anthropomorphize them.80 The fundamental manner in which river gods were portrayed in Roman art and literature – and hence in the popular imagination as well – was borrowed almost wholesale from the Greeks.81 Of special importance in this cultural exchange was the Acheloös, foremost among the rivers of Greece.Images of this river god were exported to Etruria in the archaic period, where they proved to be so popular that local manufacturers began to copy them.82 It seems likely that the Romans in turn adopted the imagery of the Greek river gods from the Etruscans.

Fig. 2: Roman statue of a river god holding a cornucopia, Naples. Reproduced from http://www.theoi.com/Gallery/S36.3.html.

So what did Acheloös look like? He, like other Greek river gods, was originally portrayed in sculpture and on coins as a bull with a man's head. From the fifth century BCE onwards, though, their features became gradually more human, turning first into a young man with bull's horns, before shedding the bovine aspect entirely and adopting the form of an older, bearded man starting in the second century BCE. Finally, following the example of statues of the Nile produced in Alexandria, the iconographic type that would become the standard in Roman times developed: that

80 Campbell (2012), p 128-135. 81 Campbell (2012), p 153. 82 Meyers (2009), p 241-243.

(24)

of the naked or semi-naked bearded man reclining, often resting on an urn from which water flowed, and holding one of various attributes in his hands (see Figure 2).83 A second type that remained popular in the Roman period was apparently developed by the Etruscans: the disembodied head of a bearded, long-haired man, originally with bull's horns, although these too disappeared eventually (see Figure 3).84 These images are often found around springs or fountains, with the mouth of the river god sometimes acting as the fountainhead, thereby giving a literal image to the notion that the appearance of running water was a divine gift.85

The association between river gods and bulls is a very old one,86 and finds its clearest expression in a famous Greek myth, recalled also by Ovid in his Metamorphoses.87 In this tale, Heracles fights the mighty river Acheloös for the hand of his future wife Deianaria. During the struggle, the river god takes on the shifting forms of a man, a serpent (perhaps reflecting the meandering course of the river), and a bull. In this same myth, incidentally, we can find the origins of another iconographical trope, that of the river god holding a cornucopia, symbolizing its life-giving powers; it was supposedly created from Acheloös' horn, broken off during the fight. Though the bovine features of river gods gradually vanished from artistic depictions over the centuries, they were not forgotten. In later Greek and Roman literature, rivers were with some regularity described as “horned”.88 Virgil, for example, refers to the “gilded horns” of Eridanus (the river Po).89 Several centuries later, the Gallo-Roman poet Ausonius in his Mosella still gave horns to the Rhine.90

It is rare, however, for a river god to play a prominent part in Greek or Roman myths. As a survey of the religiously significant rivers of Greece demonstrates, although these deities are mentioned in myth with some frequency, they usually play only a peripheral role. They are often passive observers of the events of the story, or receive little more than a name-drop, often as the father or ancestor of a hero or maiden, lending these mortals an air of divinity.91 Even among the barbarians such claims were made; according to the late Republican poet Propertius, the Belgic chieftain Virdomarus “boasted he was born of the Rhine itself”.92 This tendency is also given an odd twist in the Aeneid, where the human Turnus, foe of Aeneas, is brother, rather than son, to the water goddess Juturna.93 When river deities do take action, it is virtually always in the romantic sphere;

83 Gais (1978).

84 Meyers (2009), p 241. 85 Vollkommer (1994), p 35.

86 See also, for example, Homer, Iliad 21.237 (English transl. A.T. Murray 1924). 87 Ovid, Metamorphoses 9.62 (English transl. B. More 1922).

88 Vollkommer (1994), p 4 deals with the specific case of the Rhine. 89 Virgil, Georgics 4.371 (English transl. J.B. Greenough, 1900).

90 Ausonius, Mosella 436-437 (English transl. D. Parsons 2003). See also Chapter III, “Pater Rhenus”, below. 91 Brewster (1997) provides countless examples.

92 Propertius, Elegies 4.10 (English transl. A.S. Kline 2002). 93 Virgil, Aeneid 12.59.

(25)

they chase water-nymphs or the beautiful girls who visit their banks, often against their will. Of the Greek rivers, the Acheloös and Alpheios especially were known as notorious womanizers and appear to have fathered dozens of children.94

Fig. 3: Limestone river god mask, Cologne. Reproduced from Vollkommer (1994), p 34.

Their appearance in myth and iconography serves to mark out a number of subjects, domains, and themes with which river gods were commonly associated.95 The most prominent of these is without a doubt that of life and health. River gods were life-giving forces, as their waters brought fertility to their banks, their representations in art carried the cornucopia, and their mythical personifications impregnated women and fathered heroes and kings. Healing and health are related themes, as is ritual purification. The opposite side of the coin, death, as well as rebirth, was ascribed to river gods as well, as the topos of the dying hero being reborn as a god through the intervention of the river, briefly discussed in the previous paragraph, demonstrates. In addition, the power of prophecy was frequently ascribed to river gods; secret knowledge of both past and future was also part of their domain.96Sadly, this subject falls beyond the bounds of my interest in this thesis.

Because of their relative obscurity – few temples were built for them – one might expect that

94 Brewster (1997), p 9-14, 80-87.

95 See Taylor (2009) for an essay on many of these themes, their appearance in myth, and their interrelation. 96 Taylor (2009), p 30-31.

(26)

river gods were primarily mythological figures or decorative elements in iconography, but this was certainly not the case. The worship of river gods was very much a part of lived religion. It was certainly so in Homer's time; the Iliad contains several casual references to sacrifices to river gods. In his tale, the Trojans were said to regularly sacrifice bulls to the river Scamander, close to their city, as well as “casting single-hooved horses while they yet lived into [the river’s] eddies”. The same river god also had a priest.97 The Greeks evidently had similar customs. At one point, Nestor, wisest among the Greeks, tells the tale of a battle in Greece in his youth. Before the fighting, sacrifices were made to Zeus, Poseidon, Athena, and Alpheius, the god of the “sacred stream” near which his army had been camped.98 In a third passage, Achilles speaks of the promise his father had made to the river Spercheus, if the god ensured Achilles’ safe return: to sacrifice fifty rams on his altar, and that Achilles would offer him a lock of his hair.99 The river god apparently failed to keep his side of the bargain, and Achilles did not return.

Of course, Homer, who lived somewhere around the eight century BCE, was ancient history by the time of the Roman Empire. The evidence provided by Pausanias, who travelled extensively during the second century CE and wrote a Description of Greece, which has a lot to say on the subject of rivers, demonstrates that the worship of river gods was still very much alive at this time. Indeed, the custom of cutting off one’s hair as an offering to a river, as in the tale of Achilles, was still practiced in places.100 In the city of Psophis, situated on the river Erymanthus, there was even a full-fledged temple to the river god, which contained, according to Pausanias’ brief description, an “image”, a cult statue, of the god. He goes on to explain that “the images of all rivers except the Nile in Egypt are made of white marble, but the images of the Nile […] they are accustomed to make of black stone”, implying that many other such white marble statues existed across Greece.101

When the Romans came to Greece, they respected these local river rites and even shared in them; Sulla, the Roman dictator who campaigned in Greece in the early first century BCE, halted his march for a moment to give sacrifice to the river Cephisus.102 This is not to be wondered at, as the rivers in Italy, the Roman heartland, were apparently worshipped in much the same ways, although our sources are often scarce. A popular Italian river cult centred around the Clitumnus river in Umbria, which is known to us primarily from a letter of the younger Pliny, dating to the start of the second century CE.103 In this letter, Pliny heartily recommends a visit to this river, which

97 Homer, Iliad 21.130-133, 2.523. 98 Homer, Iliad 11.725-729. 99 Homer, Iliad 23.144-150.

100 Pausanias, Description of Greece 8.41.3 (English transl. W.H.S. Jones & H.A. Ormerod 1918).

101 Pausanias, Description of Greece 8.24.12.

102 Plutarch, Sulla 17.3 (English transl. B. Perrin, 1916).

(27)

boasted not only a sacred spring but also “an ancient and venerable temple, in which is placed the river-god Clitumnus”. Pliny clearly believed in the god’s power, adding that “the prophetic oracles here delivered sufficiently testify the immediate presence of that divinity”. Nor was he alone: there were “numberless inscriptions upon the pillars and walls […] celebrating the virtues of the fountain, and the divinity that presides over it”.104

Of course, not all river gods were equal. Some were certainly worshipped more prominently than others. We have already seen that the Acheloös was considered the greatest river in Greece, and the Nile also was renowned the world over.105 As might be expected, the rivers nearest to Rome, the Tiber foremost among them, took centre stage in the Roman world. Certainly in the earlier periods of their history, before the Empire had expanded too far beyond their banks, local river gods played an important role in Roman religion, although Cicero records that even in his time, the first century BCE, “the augurs’ litany includes as we may see the names of Tiberinus [god of the Tiber], Spino, Almo, Nodinus, and other rivers in the neighbourhood of Rome”.106 A festival was also held for the Tiber and the fishermen who made their living off the river in June, with games taking place on the Campus Martius.107

According to Georg Wissowa, the great historian of Roman religion, the cult of Tiberinus actually grew out of an earlier worship of Volturnus, a god of rivers and water in general, who had its own priestly college and another festival, the Volturnalia.108 At some point, this deity came to be associated with two specific rivers: the Volturno in Campania, which adopted its name, and the Tiber, for which Volturnus took on the epithet “Tiberinus”. This god of the Tiber had yet another religious festival in its honour, taking place on the 8th of December, the date of the founding of a sanctuary to “Father Tiberinus” on the Tiber Island.109 In addition to this worship of the river itself, a number of rituals were connected to the Pons Sublicius, a bridge over the Tiber; these I will discuss in the next paragraph.

Finally, an incident taking place during the reign of Tiberus in the first century CE and reported by Tacitus demonstrates that the divine status of rivers was more than just a story to the Romans, and that the religious reverence of rivers could even influence political decision-making. Tacitus tells us that a discussion arose in the senate over a plan to prevent the Tiber from flooding

104 For more on Italian river cults, see Wissowa (1912), p 224-225. For an extensive analysis of the evidence for

the Rhine, another major river, see Chapter III, “Pater Rhenus”, below.

105 Hesiod, Theogony 340 even names it first among the many river god children of Oceanus and Tethys.

106 Cicero, On the Nature of the Gods 3.52 (English transl. H. Rackham 1933).

107 Ovid, Fasti 6.235-240.

108 Wissowa (1912), p 224-225. The same theory is repeated by Wegner (1976), p 102.

(28)

(which happened regularly) by damming some of its tributaries, thereby weakening its current.110 In response to this plan, deputations from towns located on these rivers came to Rome to protest the plan. Their reasons were partly pragmatic, as their livelihoods depended on the water, but they had reservations about the religious implications of the plan as well; “the faith of their fathers”, which included “hallowed rituals and groves and altars to their country streams” should be respected. It was also deemed improper that the mighty Tiber should “flow with diminished majesty”. Although Tacitus dismisses such concerns as “superstition”, the plan was subsequently scrapped.

It certainly seems that the worship of river gods was primarily a local affair, and this is not surprising. Divine as they were, rivers were bound to their particular environment and could not be expected to hold much sway in other parts of the world. We should certainly not assume, however, that it was a matter only for the lower classes and beneath the notice of the elite. Some members of the elite, like Tacitus, were clearly sceptical of the divinity of rivers, but others, like Pliny or Aelius Aristides, embraced it, and we have seen that Tiberinus/Volturnus was part of the pantheon of the official state religion. Votive inscriptions dedicated to river gods, nymphs, or springs, were set up by men and women from every layer of society, from slaves to senators.111 The worship of rivers as gods permeated the Roman world, and it is crucial to remember this religious background when discussing the role of these rivers in any other context as well.

River crossings and religious boundaries

It shall by now be abundantly clear that when most Romans looked at a river, they saw more than we do. To Roman eyes, rivers were far more than mere geography; they were sacred, infused with divine power to bring life and death, even to tell the future, and they were deities in their own right. So what did it mean to cross an entity of such power, taking a boat over the stream or building a bridge? The fact that such a crossing involved interacting with a god is acknowledged and

illustrated nicely by a scene on the Column of Trajan, erected after the emperor's conquest of Dacia and telling the tale of that campaign. Near the start of the frieze spiralling upwards along the column, we see the river god of the Danube in human form, rising out from his waters and looking on with apparent approval as the Roman armies cross over his stream (see Figure 4).112 In this paragraph, I shall elaborate on the ritual and religious significance of these river crossings.

Once again starting in the realm of myth, perhaps the most dramatic examples of the ritual aspect of crossing a river are provided by the legendary rivers of the underworld, Styx and

110 Tacitus, Annals 1.79. Cited in Campbell (2012), p 130.

111 Campbell (2012), p 133-134.

(29)

Acheron.113 These terrible waters, of course, formed the ultimate boundary in the Greco-Roman cosmology: the one between the realms of the living and the dead, and to cross them was in that sense clearly a deeply meaningful act. Moreover, this crossing could not be affected without the proper rituals. The tale of Charon the ferryman, who demands payment in the form of a single coin, an obol, before allowing the dead to cross, is well-known. Proper burial of the body was another ritual necessity; in Virgil’s Aeneid, the shade of Palinurus, the helmsman of the Trojan fleet who drowned before reaching Italy, is left to wander the banks of the river of the dead until his body has been found and buried.114 In myths such as these we find the ultimate expression of the religious significance of crossing a river, and of the difficulty of doing so without the proper ritual

preparation.

Fig. 4: The crossing of the Danube on the Column of Trajan. Reproduced from Lepper & Frere (1988), plates VI-VII.

Of course, these are just stories in the end, and though myths can definitely tell us something about the beliefs of the people who told and shared them, they are no substitute for the real thing. There is certainly evidence to suggest, however, that these exact same patterns were followed in actual religious observance. The great historian Herodotus tells us that during the Greco-Persian War of 480-479 BCE the Persians under Xerxes “sought good omens” before crossing the river

113 Brewster (1997), p 15-19, 69-71.

(30)

Strymon by sacrificing a number of white horses.115 The Greeks knew this custom, too: some years earlier, around 494 BCE, Cleomenes, King of Sparta, had marched on Argos and reached the river Erasinus. The king offered sacrifices to the river, but in return he was given omens “in no way favourable to his crossing”. He therefore chose to find another way around, and was even a good enough sport to honour the hostile river “for not betraying its countrymen”.116 Even earlier, the archaic Greek poet Hesiod in his moralizing poem Works and Days had advised his brother never to cross a river without praying and washing his hands in the stream. If he did not, the gods would be “angry with him and bring trouble upon him afterwards”.117

Once again, there seems to have been little difference between Greeks and Romans in this regard. The Romans themselves similarly sought omens and offered sacrifices before crossing a river. According to Cassius Dio, writing in the third century CE, Crassus crossed over the Euphrates in 54 BCE to go to war with the Parthians in spite of unfavourable omens, and went to his death.118 As we have seen, some decades earlier Sulla, also leading an army into battle, made sacrifices to the river Cephisus before crossing it.119 Another interesting example is the Petronia, a tiny stream running through the southern part of the Campus Martius in Rome before joining the Tiber.120 Though crossing it would not have been a difficult task, this small brook apparently carried great religious significance as one of the boundaries of the city, even into the Imperial era. According to the obscure second-century CE grammarian Festus, if a magistrate wished to cross it to conduct business on the Campus Martius, he had to take the auspices first.121 Elsewhere in his (fragmentary) text, in fact, he states that this ritual was required for the official crossing of any river whatsoever.122 It may very well be that this rule was not always followed to the letter, but its appearance in this text does express a fundamental belief in the divinity of rivers and the danger and ambiguity associated with crossing them.

The building of bridges across a river seems in many ways to have been a similarly religious act, and a number of traditions dating from the earliest period of Roman history remind us of this fact. In fact, the first bridge ever built over the Tiber in the seventh century BCE, the Pons

Sublicius, was associated with a number of rituals; according to Plutarch, the bridge was originally maintained by the priesthood, and “sacrifices of the greatest antiquity and the most sacred

115 Herodotus, Histories 7.113 (English transl. A.D. Godley 1920).

116 Herodotus, Histories 6.76.

117 Hesiod, Works and Days 739-740 (English transl. H.G. Evelyn-White 1914).

118 Cassius Dio 40.18.

119 Plutarch, Sulla 17.3-4.

120 Campbell (2012), p 16.

121 Festus, On the Meaning of Words L 295.24: Petronia amnis est in Tiberim perfluens, quam magistratus

auspicato transeunt, cum in campo quid agere volunt. Cited in Holland (1961), p 18-20.

Rivers as boundaries in the "Empire without end": An ancient perspective on a modern debate