A Risk Society

Risk defines modernity. We see it in almost every aspect of life, from disease to climate change, accidents to crime. Risks such as these pose a threat to our health, our safety, our finances and our mental health. The German sociologist Ulrich Beck has argued that Western civilisation is nothing other than a ‘risk society’, in which not only are our lives shaped by the calculation of probabilities but the number of risks has multiplied in line with innovations in technology.Footnote ¹ Be it nuclear bombs or the threat of environmental catastrophe, Beck’s world is packed full of the dangers created by scientific advances – except that they are no longer seen as advances. The old narrative of science as a force for good has been fundamentally weakened by innovations that threaten human existence itself. People have come to distrust the motives – as well as the statistics – of experts. As the Coronavirus pandemic showed clearly, experts disagree, often fundamentally, and, indeed, make errors in their predictions. Why should ordinary individuals believe and trust in experts when they can agree on neither the data nor its interpretation?

Paradoxically, therefore, greater scientific knowledge has resulted in greater general uncertainty. For Beck, this is a sign of a ‘reflexive modernity’, where respect for traditional sources of knowledge, such as scientists, has declined and left a void and a desperate feeling of anxiety and insecurity. Devoid of reliable and trustworthy sources of information, individuals are left to seek out their own knowledge about the nature of the risks they face in their lives and have turned to all manner of ‘alternative’ experts in a search for reassurance, ranging from internet conspiracy theorists to the new certainties of various counter cultures. The globalised economy has generated an extraordinary array of life choices for the individual to make, but such choice has generated yet more fear. Taking the wrong course exposes individuals to risks and represents a failure on their part to understand the implications of their actions, a failure which only adds to their sense of anxiety. Not surprisingly, Beck argues, reflexive modernity has sought cover by taking control of all aspects of life and attempting to eliminate risk from ordinary life. Taking any form of unnecessary risk becomes an act of ignorance – an immoral refusal to understand the dangers implicit in a particular course of action.

Beck, however, sees risk as the product of modern capitalism. By definition, therefore, he denies that the concept existed in the pre-modern world. Risk represents what Beck calls ‘a systematic way of dealing with hazards and insecurities induced and introduced by modernization itself’, as opposed to ‘older dangers’, or what Giddens calls ‘inclement nature’.Footnote ² ‘Human dramas’, argues Beck, such as ‘plagues, famines and natural disasters’ and ‘the looming power of gods and demons … differ essentially from “risks” in my sense since they are not based on decisions, or more specifically, decisions that focus on techno-economic advantages and opportunities and accept hazards as simply the dark side of progress’.Footnote ³ Pre-industrial hazards were ‘strokes of fate’, no matter how large and devastating they were, and they were attributable to an outside ‘other’ and could be blamed on gods. Whereas modernity, in this view, has introduced a whole range of mega-risks, unlike anything seen before, in the pre-modern world ‘danger’ offered a sufficient range of vocabulary because an intuition existed of the possibility of future harm without there being much desire, need or ability to quantify it. Or as Joffe explains: ‘The incalculable threats of pre-industrial society are turned into calculable risks in industrial society, in line with the modern project of promoting rational control in all spheres of life.’Footnote ⁴

Beck’s work has been highly influential in the field of risk studies and clearly has important implications for any proposed analysis of risk in the Roman world. And Beck is certainly not alone in seeing the ancients as passive victims of fate. The social theorist Tony Giddens also sees modernity as living ‘after tradition’, and ‘is essentially to be in a world where life is no longer lived as fate’.Footnote ⁵ In this view, whatever disasters happened to humanity in the past were the work of an unseen divine or other supernatural power. They were events that were not, Beck concludes, ‘politically charged’. We need think only of accusations that Nero deliberately caused the Great Fire of Rome in 64 ce to see that the reality in Antiquity was different. But, to be fair to Beck, as a social theorist, the ancient world has never been of much relevance to his work. Elsewhere he does appear to accept that risks have always been present, but he argues that the nature of modern risks is of a different order: the consequences of nuclear war are unfathomable and far greater than anything humanity has faced before.

As one writer has noted, ‘risk is quite unique in the quantity and extent of research that draws upon it’.Footnote ⁶ But it is perhaps this denial of risk’s existence in the pre-industrial world that explains why the subject has been of limited interest to scholars of Antiquity. The work that has been done is mostly on the Greek world and has been focused on peasant culture or on the use of divination to understand the future. Garnsey examined the frequency of food crises in Antiquity and how peasants and emperors responded to this risk. Gallant looked at the strategies adopted by peasant farmers in response to climatic variability and fluctuations in harvests. This included the cultivation of social relations to build up potential support networks. Grey adopted a similar approach towards the later empire. These books contain only brief discussions of what constitutes risk and how the concept might be applied to the ancient world, understandably preferring to focus on the specific practices they are examining. Grey’s study of the effect of the eruption of Vesuvius contains a more detailed and useful discussion of the concept of risk in relation to that calamitous event.Footnote ⁷ Beard’s article on Roman aleatory culture also contains useful observations on how Romans dealt with and even embraced risk.Footnote ⁸

Eidinow has published a popular book on the subject of luck, fate and fortune, which shows, in an accessible way, how ancient understandings and discussions concerning these concepts are still relevant to the modern world of risk. Beerden includes a more detailed discussion of the concept of risk in her work on divination in the ancient Greek world, although this follows a narrow, statistical approach, seeing the concept as a quantifiable uncertainty and therefore inapplicable to Antiquity. For her, divination was how the ancient Greeks dealt with their unquantifiable uncertainties. A similar approach is taken by Eidinow, who sees risk as ‘quantified certainty’ and therefore argues it cannot be applied to Antiquity.Footnote ⁹ Both these approaches provide excellent insight into how the practice of various forms of divination allowed the Greeks to try to manage the uncertainties in their life.

There is an increasing body of work dealing with the substantial threats that the Romans confronted. My own work looks at how the Romans thought about and dealt with disasters as a whole and the strategies they used to try to prevent them from happening, while there have been various studies that examined particular events, such as the Great Fire of 64 ce.Footnote ¹⁰ Scheidel and Harper have both provided comparative historical studies of plagues and pandemics and, in Scheidel’s case, war.Footnote ¹¹ Interest in the risks posed by climatic conditions and changes in Antiquity has also grown rapidly, with various attempts to interpret the historical data.Footnote ¹² The Justinianic plague has seen substantial new research, including an increase in analysis of new forms of genetic evidence.Footnote ¹³ There is also an ongoing debate over the seriousness of the impact of both this plague and ancient climate change.Footnote ¹⁴

There have also been many studies into various aspects of more specific risks that Romans faced. These have included demographic studies into life expectancy, given the significant threats of disease and malnourishment, and how these relate to the assumptions at work in Ulpian’s Life Table of annuities. There have also been studies into a wide variety of ancient phenomena, ranging from military logistics, to legal aspects of uncertainty and maritime loans, all of which reveal something of how the Romans understood future danger. These are all examined in more detail in later chapters.

This book looks at how the Romans understood, thought about and dealt with risk. It sets out to challenge the views of Beck and Giddens in a number of ways. It argues that risk is a useful term to apply to Antiquity and that the Romans did not simply see themselves as passive in the face of fate. It argues that the Romans did display some understanding of risk and took a variety of steps to help manage it. And it argues that modernity’s attitude to risk should not be seen as entirely unique. Pascal’s 1654 discovery of probability represented a significant step forwards in the understanding of uncertainty. But it was a shift along a spectrum, not a sudden change from darkness to enlightenment, from total ignorance to knowledge.

Probability and Risk

Pascal’s understanding of the stable relative frequencies of certain chance events meant that the future could be calculated and, for some, represents ‘the underlying essence of risk’.Footnote ¹⁵ Once it was understood that there is a one-in-six chance of a dice throw producing a six, then this could be used to predict how later dice rolls would turn out. It is worth remembering that this is a model that will not necessarily correspond to the actual future. It may instead happen that the next three dice rolls all produce sixes, but the likelihood of this unusual occurrence could also now be calculated (as 6 × 6 × 6 = 1/216) and used to plot a normal distribution chart, known as a bell curve, which captures the likelihood of the various combinations of throws occurring. Pascal’s discovery had significant implications for how uncertainty was seen, in that it could now be quantified, for decision-taking about what courses of action to take in the future, and indeed people’s relationship with the future. Some aspects of the future were now mathematically knowable (which is not the same as knowable).

Interestingly, in the two centuries before Pascal’s discovery, the notion of risk also emerged across Europe as a way to denote situations of potential damage to seaborne cargoes. We can interpret both as part of a societal shift towards a more calculating worldview, according to which people were trying to quantify unknowns and calculate the likelihood of certain outcomes and therefore make decisions about what was the best course of action to take. The term entered English during the 1660s from the French risque, itself derived from the Italian riscare (to run into danger), formulated from the medieval Latin risicum. As for the original etymology, as the Oxford English Dictionary notes, the origin of risk is ‘much debated’. One theory is that the term risk comes from the Latin resecare (to cut off), from which the Spanish risco, cliff, derives, which obviously posed a threat to shipping; or that it comes from the Icelandic ráðask (meaning something like ‘to decide to launch an attack’), a military term introduced into Latin following Norse attacks on the European continent; another possible source is the Arabic root rizq, meaning ‘sustenance’, ‘income’ or ‘fortune’, originally derived from the Persian rozik, ‘daily bread’; or another is the Greek rhiza, meaning ‘root’ or rhysis, ‘deliverance’. Whatever the ultimate source, it is clear that all these origins contain a sense of undertaking actions where there is the potential for both benefit and harm.

As for what the term risk has come to mean in the modern world, there is no simple consensus or accepted definition. In its simplest form, risk is about the future. Since the future is unknowable, risk is also about uncertainty and the inability of humans to be sure of the consequences of their actions. It represents a lack of knowledge, since there can be no risk when there is certainty. Or, to put it another way, risk cannot exist in a predetermined world, a world of fate.Footnote ¹⁶ In a broad sense, therefore, we can see risk as relating to events where something is at stake and where the outcome is uncertain.Footnote ¹⁷ When dealing with probabilities, therefore, risk can be described as a neutral term, concerned merely with mathematically calculated losses and gains. In Frank Knight’s classic book, Risk, Uncertainty and Profit (1921), risk was defined as an objective quantity that could be obtained by calculation according to the factors relevant to the outcome. By contrast, uncertainty was something more subjective and judgemental that could not be worked out mathematically.Footnote ¹⁸

The Knight approach was later adopted in various areas of risk management and finance. In finance, risk was seen as representing volatility, which could be calculated based on the previous price behaviour of an asset. The more an asset’s price moved the greater its volatility and risk. The capital asset pricing model took this further and argued that, for investors to attain higher returns, they had to accept higher risks.Footnote ¹⁹ One benefit of this approach is that, whereas the term risk tends to highlight the downside, volatility is a more neutral term that reflects the fact that risk can produce both good and bad results. The problem is that it is a predictive model that can estimate future outcomes based only on what has happened in the past. When events do not turn out as expected, the consequences can be dramatic. The collapse of the highly leveraged Long-Term Capital Management (LTCM) in 1998 and the 2008 Great Financial Crisis both showed that any calculation of risk in this way always includes a large element of qualitative assessment or reliance on past behaviour as an indicator of the future.

There are, then, limits to the calculation of risk based on probabilities. The calculation can be based only on a priori knowledge, as is the case with dice, where there is a limited range of possibilities, or it can represent a statistical probability based upon previous events, such as is done in insurance.Footnote ²⁰ In real-life problems, the number of relevant factors is often so large as to mean that no a priori knowledge is usable. Also, real life tends to throw up unexpected or even unimagined outcomes. I sat on my Cambridge college’s committee that examined the risk register and we never even considered the possibility of a pandemic. That is not a criticism but simply a statement that it is extremely difficult to plan for extreme events, let alone for what former US Secretary of Defense Donald Rumsfeld notoriously called ‘unknown unknowns’. These unimagined risks have also been termed ‘Black Swans’, in reference to the discovery of such animals in Australia by Europeans to whom it had never occurred that such birds might have existed.Footnote ²¹ Whereas Knight’s approach argued that uncertainty differed from risk because it was not reducible to numerically definite probabilities, Black Swan events showed that many situations threw up eventualities that had never even been imagined as possibilities. The financial calculations of the Nobel Prize–winning economists of LTCM generated a comforting sense of being in control until something unexpected blew them out of the water. In such a context, where there are so many variables and unknowns, risk calculations should not be seen as objective evaluations but more accurately as estimates, or ‘judgements in the context of uncertainty’. They represent more of a subjective probability. Indeed, any risk that is fully measurable does not in reality represent an uncertainty, in that the future outcomes are known, and so there is no risk of any alternative scenario occurring. The boundary between risk and uncertainty has therefore become blurred at best.

Risk has always been a term used mostly of negative outcomes. It is therefore about danger and has become widely used to represent specific dangers themselves (as in, ‘we face a number of risks’). Some people prefer to call these dangers ‘hazards’ as a way of differentiating from the act of risk calculation. So, a hazard can be seen as ‘a set of circumstances which may cause harmful consequences’, whereas risk is ‘the likelihood of it doing so’.Footnote ²² This can be taken further, and risk can be calculated as the size of the hazard multiplied by exposure to it, where exposure means the extent to which the victim can be affected by the hazard.Footnote ²³ The problem with this approach is that ‘hazard’ is a term that is rarely used as a noun in modern English, and its use attempts to force specific, technical meanings onto the term ‘risk’ when this has become widely used in a more general sense. Some might like to define risk as a scientific concept, but most calculations of risk involve the assessment of expectations of future behaviour based on knowledge and experience of the past, and knowledge and experience can be obtained by both formal and informal means. Tradition, custom, rules of thumb, estimates, judgements based on practice – all are ways that lay people can make risk assessments based on their own experience. Great bodies of scientific data may give a sense of objectivity compared with such common-sense calculations, but events such as the pandemic have highlighted the fact that such data are far from straightforwardly objective.Footnote ²⁴

One of the problems of trying to define ‘risk’ is that it serves only to limit what has become a broad term, used in a variety of ways:Footnote ²⁵

1. 1. the chance or possibility of a usually negative event happening

2. 2. danger or the cause of danger

3. 3. the probability of an event happening

4. 4. the size of the negative impact.

This variety of usage does not simply reflect a recent lack of precision, since the term has always been employed in a wide variety of ways since its introduction more than three centuries ago. The reality is that the term ‘risk’ is now used in all manner of technical and everyday contexts, but that frequency is itself a measure of how important a concept it has become in the modern world.

Despite this breadth, there are some shared implications in the use of the term. The first is that it suggests that the future can be changed – not necessarily controlled completely but influenced and to some degree altered by analysing the factors affecting possible outcomes and making judgements accordingly. As we have seen, some have argued that this represents a fundamental shift from the pre-modern past, in that the future is now manageable by humans rather than supernatural forces, a question to which I return. All risk concepts share this element in common: a belief in the distinction between reality and possibility that means there is no room for fatalism.Footnote ²⁶

The second implication is that, since risk refers to the future, it in some sense exists only in the imagination. As we have noted, even Pascal’s probabilities represent solely a model of what will actually happen when dice are rolled. If risk is a model for the future, then any more complicated calculation than dice-rolling is going to involve various assessments about what is acceptable or desirable as an outcome. This, by definition, involves a variety of value judgements. Risk, it therefore becomes clear, is a matter of perception and how it is perceived will contain a moral dimension.

Culture and Risk

While uncertainty can be seen as an ‘objective feature of the universe’, at least as far as human experience goes, risk is in the eye of the beholder.Footnote ²⁷ Every society has a unique set of fears about the future that it prioritises over others. Different risk attitudes can be adopted by different societies to the same underlying uncertainty. The actions a society takes to alleviate these and the degree to which it treats them as acceptable depend on various cultural factors. In this context, it is evidently problematic to treat risk as an objective, technical concept. Instead, what constitutes a risk reflects a range of ideological, structural and social-psychological elements in a given social situation. Societies can in these ways be seen as revealing themselves by how they deal with dangers.Footnote ²⁸ They have what can be described as specific risk cultures.

This approach was influenced by the work of the anthropologist Mary Douglas.Footnote ²⁹ Douglas did not deny the reality of the underlying dangers: ‘this argument is not about the reality of the dangers, but about how they are politicized’. She argued that there is always a moral dimension to risk-taking and that a failure to follow societal norms resulted in the victims themselves being blamed, especially in the modern world with regard to sexual behaviour and drug-taking. In this way, real dangers are used to give ‘automatic, self-validating legitimacy to established law and order’.Footnote ³⁰ This moral and political dimension also affected the allocation of resources within society towards certain types of risk reduction. All such allocation decisions were driven by whether society saw the risk as sufficiently threatening to make it worth the extra investment.Footnote ³¹ Risk, Douglas argues, has become a key idea for modern times because of its uses as a ‘forensic resource’ that has a meaning ‘consistent with the political claims in vogue’. The term ‘danger’, she argues, ‘does not have the aura of science or afford the pretension of a possible precise calculation’.Footnote ³²

It is possible to take this approach to understanding risk a stage further. Influenced by Foucault’s theories of social discourse, risk perceptions are seen as the source of hazards themselves. The ways in which a society constructs risk represents a discourse, which concentrates on certain forms of uncertainty as being especially problematic. This narrative can harness and amplify risks to aid governmental control. Risk is in this view closely linked to power because it is often those in authority who have the ability to decide what constitutes a danger. The underlying hazard is itself seen as a cultural product that is invoked in a risk discourse to justify decisions regarding what constitutes risk, dangerous behaviour and the threatening individuals who engage in it. A risk has to be recognised as such, which will always involve an act of human perception, and such perceptions can be based on a variety of sources, including scientific data, everyday experience or religious beliefs. When different groups fail to agree on what constitutes a risk, it is not because they are interpreting the data differently, it is because different data matter to them.

From this viewpoint, risk does not exist in reality; rather, it represents a culturally embedded way of ordering reality that renders it into a calculable form.Footnote ³³ It is a way of representing future events so that they can appear manageable. The significance of risk does not therefore lie with risk itself but with what it gets attached to. This anthropological approach sees all societies as developing reflexive mechanisms for the processing of perceived dangers, from fears that the sun will not rise again to natural disasters, potential catastrophes that are sought to be averted by various rituals. The concept of risk itself developed in a specific historical context, and represented ‘a new way of understanding the future and what harms or possibilities it might hold’.Footnote ³⁴

Early modern Europe, from 1500–1800 ce, was a place where the future was no longer seen as the realm of fate or fortune but one of calculable probabilities. The period, therefore, saw the development of a variety of ideas based on probabilistic reasoning and the possibility of acquiring more certain understanding of how future events would unfold and so guide decision making. It was an era where new kinds of political and economic organisation arose, such as central banks and insurance markets, which reflected this new approach to the future. The twentieth century saw an expansion of risk discourses into many other fields of expertise such as sciences, social sciences, medicine, and law, making it seem unique to what Beck and Giddens term ‘advanced modernity’. Beck has little to say about the emergence of the concept of risk, but he does see risk calculation as symptomatic of modernity: ‘induced and introduced by modernity itself’.Footnote ³⁵ From a cultural standpoint, this represents a particular kind of cognitive activity focused on the assessment of potential outcomes rather than the discovery of anything real underneath. As Garland says: ‘Risk begins where certain knowledge ends. Claims about risk are, literally, uncertain knowledge claims – impressionistic guesses, informed estimates, and probabilistic predictions about a future that cannot fully be known.’Footnote ³⁶

Risk is a multivalent modern term. Many studies of risk now adopt a broader, less realist approach to risk, seeing it as inseparable from the perspective of the viewer. Others retain a technical approach based on statistically quantifiable factors. As Cambridge University’s Professor of the Public Understanding of Risk, David Spiegelhalter, notes, ‘as it has gained in popularity, it has lost the sharp edges of definition’.Footnote ³⁷ In practice, people tend to adopt whatever approach suits their aims. Actuaries, corporations and economists are more likely to use a statistical approach, whereas historians and anthropologists favour a more cultural definition.

Risk in Rome

How then am I defining ‘risk’ for the purposes of this book? I will definitely not be adopting the narrow technical approach, since this would mean that almost no risk analysis would be possible with regard to Rome. I adopt the popular and informal definition suggested by David Spiegelhalter, that risk is ‘anything to do with situations when “bad” (or “good”) things may, or may not happen’.Footnote ³⁸ The crucial elements are simply that there is uncertainty and that the outcomes may be ‘nice or nasty’. This last point is an important one. Most people instinctively see risk as a negative concept, focusing on the possible downside (indeed, it is noticeable that Spiegelhalter puts the ‘good’ in parentheses). But risk can also be seen as something that will generate significant variation in any outcome. Embracing risk can be a way for the individual to get richer, to improve his or her social status and generally move up in the world.Footnote ³⁹ As I am interested in the ways in which risk was perceived and dealt with in the ancient world, I am alert to the discourses that were created around a very different set of risks than we would accept today, many of which related to religious ideas. Douglas was also right to see that risk involves a political blame game against those who engaged in what is seen as risky behaviour.

Perhaps more importantly, I restrict my approach to looking at areas where some element of calculation can be found in the Roman material, primarily relating to the period of the early empire. By this I do not mean numerical calculation of probabilities, since that did not exist, but areas such as law, religion, finance and seaborne trade, where the Romans also display various methods of probabilistic thinking, generally based upon experience and observation of the past, which show that they were concerned with trying to plan for an uncertain future. Lawyers, sailors and priests were the experts in ancient Rome. Ordinary Romans also had to make various assessments and judgements in the course of their everyday lives, and they relied more on the inherited wisdom of the past to reduce and control the risks they faced. These cultural tools were more communally calculated but still gave room for individual interpretation in their application.Footnote ⁴⁰ These are all the focus of this book.

This might seem a long way from the statistics generated by ‘objective’ risk studies. But what this introduction has tried to bring out is that a more cultural approach to risk moves away from this numerical approach, which sees risk as unique to the modern world. Looking at the various ways in which the Romans dealt with uncertainty serves to narrow the gap between us and them in that it shows us all trying to find ways to cope with the uncertainties we inevitably face in life. But it also widens the divide. If thinking about risk is always contingent upon historical and local contexts, and that knowledge is constantly contested and disputed, then we can expect that Roman ideas about uncertainty will be almost unrecognisable from our own. We must also remember that the ‘Romans’ were never a simple or static group. I am mainly concerned with arguing that various levels of understanding of risk can be found in the ancient sources, which relies heavily on the texts of the elite and evidence from the city of Rome itself, but I hope it will also become apparent that the inhabitants of the vast Roman empire represented a diverse set of peoples inhabiting many different circumstances. They all faced different types and degrees of risk and dealt with them in many different ways. Social status, in particular, dictated an individual’s capacity to make decisions based on their own assessments. Slaves may have understood the risks of their situation but could not do much about them. Male members of the propertied class, by contrast, could consciously decide whether or not to invest in safe investments, such as land, or gamble on maritime trade.