Abstract
This article offers an in-depth analysis of the diffusion model of science popularisation. It reviews criticisms against the model and shows that they do not warrant its rejection. It argues that the diffusion model has elements, hitherto neglected, which can facilitate a better understanding of popularisation. Viewing popularisation as the diffusion of knowledge is beneficial because it enables us to: (1) pinpoint the origins of popularisation and trace its historical continuity; (2) explain why science requires continuous popularisation; (3) understand why the values that popularisers promote are not arbitrary; and (4) define more precisely the role of popularisers.
1. Introduction
The systematic study of the history of science popularisation began in earnest in the 1970s (Cooter and Pumfrey, 1994; Gregory and Miller, 1998). Since then, scholars have offered a variety of interpretations of the nature and purposes of popularisation, predominantly taking ‘externalist’ and ‘genealogical’ approaches – that is, focusing on the cultural or political contexts in which popularisation takes place, the underlying motivations of popularisers and their audiences, the power structures within which popularisation is entangled and so on. Overall, the aim has been to emphasise that science popularisation is a ‘social practice’ that evolves through history and bears close links with other ‘social practices’ and to unmask the factors that have shaped the reasoning of those involved in popularisation.
While acknowledging the positive contribution of ‘externalist’ and ‘genealogical’ approaches, this article argues that the understanding of the history of science popularisation can also benefit from an approach which more minutely examines the nature of the scientific knowledge that popularisers are trying to communicate. This calls for an openness towards the idea that science popularisation is a process of the diffusion of scientific knowledge, an idea which supporters of ‘externalist’ and ‘genealogical’ historiography have criticised. 1 This article reviews the criticisms that scholars have raised against the diffusion model (as this idea has become known) and shows that they do not warrant its outright rejection. 2 It then proceeds to show that viewing popularisation as the diffusion of knowledge offers valuable insights into the process of science popularisation.
2. Literature review: Doing away with a diffusion model of popularisation
The scientific meaning of the term ‘diffusion’, which is also the meaning on which all other meanings are loosely based, is that diffusion is a movement of matter or energy from an area of high concentration to an area of low concentration. Comparing popularisation of science to a diffusion process suggests a movement of scientific knowledge from a place where that knowledge is in high concentration (i.e. within circles of individuals with scientific expertise) to a place where the concentration of that knowledge is low (i.e. among individuals without scientific expertise – the wider public).
In one of the earliest studies on the historiography of science popularisation, Richard Whitley described the diffusion model as too simplistic, because in his view, it reduced popularisation to a link between an elite group of knowledge producers and an incompetent mass of ‘atomistic receivers’ who ‘passively internalise knowledge’ (Whitley, 1985: 4). According to Whitley, the relationships between scientists and lay audience are so ‘changeable’ and ‘variable’ that ‘no single type of connection can be assumed to be general’ (Whitley, 1985: 11). He further criticised the diffusion model for assuming that popularisation plays no role in the generation and development of scientific knowledge. For Whitley, once we acknowledge that all scientific knowledge ‘depends upon particular social relationships and collective judgements’ and that facts are ‘socially constructed cognitive objects’ which ‘become established through negotiations’, then, it becomes clear that popularisation, being itself a social process, plays an integral part in the production of natural knowledge (Whitley, 1985: 11–12).
Along similar lines, Hilgartner argued that a diffusion model was inadequate because it underappreciated ‘the ambiguity and flexibility of such concepts as “genuine science”, “popularization”, “appropriate simplification” and “distortion” and consolidated an ‘epistemic hierarchy which ranks scientists above such actors as policy-makers, journalists, technical practitioners, historians and sociologists of science, and the public’ (Hilgartner, 1990: 533–534). The diffusion model, to him, also appeared to have underlying political uses – a problem that new, more adequate approaches had to address.
Another influential piece described the diffusion model as ‘confused’, ‘insufficient’ and ‘positivist’ (Cooter and Pumfrey, 1994: 248–251). Inherent in the model was a deeply inadequate notion of a ‘cultural lag’ – deeply inadequate since according to the authors ‘popular culture can generate its own natural knowledge which differs from and may even oppose élite science’ (Cooter and Pumfrey, 1994: 249). The authors further argued that ‘‘Successfully popularized’ natural knowledge might take on very different meanings within popular culture from those intended by its popularizers’ (Cooter and Pumfrey, 1994: 249, original italics). In their view, a ‘less static’, interactive approach was needed, one that captured the two-way dynamic within the popularisation process; acknowledged the autonomy of ‘popular science’ from ‘learned science’ and emancipated lay audiences from the status of mute and passive recipients (Cooter and Pumfrey, 1994: 252).
Drawing on the work of Andrew Cunningham, Jan Golinski, Simon Schaffer and other influential historians of science, J. R. Topham, argued that the ‘diffusionist notion of “popularization” was ‘utterly inadequate as a characterization of the actual processes of scientific communication’ (Topham, 2000: 560). It had rightly been ‘rejected as a historiographical principle’ because it was itself a product of ideological boundary work which took place in scientific circles in the late eighteenth and nineteenth centuries (Topham, 2000: 560). Portraying popularisation as the diffusion of knowledge, Topham argued, was a crucial element of the attempts of scientists in this period to transform science into an esoteric practice of knowledge production to which only a small group of ‘self-fashioned’ ‘experts’ had access, ‘producing’, in turn, a lay audience which did not have any say in the validation of knowledge and which over time was being rendered ‘increasingly passive’ (Topham, 2000: 561). As the ‘expert-lay audience’ relation was itself an ideological creation that was being maintained by the scientific community in order to warrant the legitimacy of their knowledge claims, a model like the diffusion model which affirmed this relation could not be used by historians to describe what popularisation actually was but could only be used to describe what the scientific community wished us to believe about the nature or purpose of popularisation. The job of the historian should be to analyse the processes through which concepts such as ‘scientific experts’ and ‘lay audience’ were being constructed, rather than focus on the relations between these concepts (Shapin, 1990).
The concerns raised in these earlier works were further articulated in a presidential address by Peter Bowler and a plenary speech by James Secord.
In his address, Bowler argued that the diffusion model was especially inadequate when studying popularisation in the eighteenth and nineteenth centuries, since this was
a period when the distinction between professional and amateur was meaningless, and when those practising science had to respond to the interests of an audience which extended far beyond the ranks of the handful of specialists who could appreciate the most esoteric aspects of what was being studied. (Bowler, 2006: 165)
The only context in which a diffusion model might be somewhat appropriate, Bowler argued, would be in the popularisation of science in the twentieth century, when the professional-amateur distinction became ‘more clearly demarcated’ (Bowler, 2006: 165).
James Secord challenged the diffusion model indirectly. He questioned the integrity of the concept ‘popular science’ partly because it carried a ‘diffusionist baggage’ – ‘To label something unequivocally as popular science can be seen as tantamount to saying that it is “not science” or even a kind of pseudoscience parading as the real thing’ (Secord, 2004: 670–671). Under the weight of this ‘diffusionist baggage’, a term like ‘popular science’, if used, would only serve to strengthen the boundary between scientific experts and non-experts, whereas the job of the historian should be the opposite – to critically analyse and deconstruct that boundary (Secord, 2004: 670–671).
3. Addressing the criticisms of the diffusion model
It is worth making two points before I go on to address the criticisms of the diffusion model. First, while the idea of a diffusion model has been around since the beginnings of the systematic study of the history of science popularisation, to my knowledge no actual body of historical scholarship exists that has taken a diffusionist approach of the kind I describe here. What the scholars cited above criticise is not actual historical work but merely a view of science popularisation – the ‘culturally-dominant view’, as Hilgartner put it – which has allegedly been held by scientific communicators and the general public (Hilgartner, 1990: 519, my italics). 3 Second, these criticisms are not part of a balanced scholarly assessment of the features of the diffusion model; in fact, to my knowledge, no such assessment exists. They all appear in studies on popularisation whose primary aim has been to advance alternative interpretations at the expense of any attempt to understand the diffusion model in its own right. They point out the limitations but do not consider any potential strengths that the model might have or any potential benefits that its application might bring. Overall, while the criticisms of the diffusion model reflect valid historiographical concerns, I aim to show below that when considered alongside any benefits the model might bring, they do not warrant its categorical rejection.
(I) A common argument against a diffusion model which views popularisation as the spread of scientific knowledge is that it is too simplistic (recall Whitley’s argument above). Scholars have pointed out that science popularisation has taken many different forms over the last few centuries (e.g. Fyfe and Lightman, 2007; Lightman, 2007). A diffusion model, they argue, is therefore unfit for the purpose of understanding the multifaceted nature, social character and rich history of popularisation.
That science popularisation has evolved since the seventeenth century is indisputable. Throughout its evolution, however, popularisation has also retained some core features. Among them are (1) the presence of scientific content in works on popularisation and (2) the attempt to make that content intelligible. By focusing on these enduring features, the diffusion model helps us determine what is and what is not science popularisation. 4 It also enables us to analyse the role that the nature of scientific knowledge plays in shaping popularisation (see below). As we shall see, taking the view that science popularisation is diffusion of knowledge does not lead to the reduction of popularisation to the transmission of scientific content or entail that we see popularisation exclusively in those terms. It does not represent a crude oversimplification of a complex phenomenon but provides a vantage point from which to see things more fully and more clearly.
(II) The understanding of popularisation as the diffusion of knowledge has also been criticised for allegedly promoting the idea of a ‘scientific elite’ (recall Cooter and Pumfrey’s argument above). While it is indeed the case that a diffusion model distinguishes between two groups – those with acquired expertise in scientific subjects and those without – there is nothing in the distinction per se that endows scientific experts with the status of ‘an elite’. The distinction that a diffusion model allows us to make is not between groups that ‘possess’ knowledge and those who do not, but between those who are in the business of generating scientific knowledge and those who are not. If anything, a diffusion model serves to emphasise the notion that knowledge is no one’s exclusive possession and that it is, in principle at least, within everyone’s reach. It helps us analyse how knowledge can spread, even in the case of the most complex scientific ideas. A diffusion model, therefore, represents a potentially democratic, not an elitist view of popularisation while at the same time acknowledges that producing natural knowledge is difficult and resource-intensive.
(III) A related criticism of the diffusion model is that it denigrates the status of popularisers and assigns them and lay audiences the passive role of mere recipients. Historians instead argue that popularisers and lay audiences ought to be viewed as integral to the creation of scientific knowledge; at the centre of a thriving popular, public, ‘artisan’ or ‘indigenous’ science which may not only carry its own meaning but may actually oppose the science found in the work of scientists (Fyfe and Lightman, 2007; Lightman, 2007; Stewart, 1992). According to this view, popularisation cannot be viewed as a process of diffusion because the end product – ‘the popular science’ – is different from the science found in original scientific works.
There are many instances in history when, for example, amateur naturalists and astronomers have furnished science with useful ideas or data, or when works of science popularisation, even works of science fiction, have suggested fruitful avenues for research and/or stimulated scientific advance (a useful overview of ‘citizen science’ projects going back to the nineteenth century can be found in Mahr and Dickel, 2019). 5 However, collecting empirical data which is then handed in to be collated or to be analysed in a laboratory, or imagining the technology of the future, or demonstrating scientific knowledge to facilitate learning is not, strictly speaking, doing science. It may begin the process, but for ideas and observations to become part of established knowledge, they must enter the scientific discourse, and at this point, some engagement with professionals and experts is usually required. Doing science involves, among other things, making sense of data in light of acquired specialised knowledge in the field, or developing a seemingly crazy science fiction idea into a scientific hypothesis and designing an experiment to test it. Also, generally speaking, scientists produce scientific reports, while popularisers of science produce scientific stories. 6 Scientific stories make existing knowledge learnable and meaningful to non-scientists but unlike scientific reports most of the time they do not add to the corpus of scientific knowledge (see below for further details). While the diffusion model assigns scientific experts the role of producers of knowledge, it does not deny that popularisers or lay audiences may themselves come to possess some degree of expertise; neither does the model denigrate the role that popularisers or lay audiences play in the development of science more generally; on the contrary, as we shall see, it helps us define their role more precisely.
(IV) The view of popularisation as the diffusion of knowledge has also previously been dismissed as presentist, the main reason being that the professional category of the ‘scientist’ emerged only in the twentieth century (recall Bowler’s argument above). However, we should not confuse the professional/amateur distinction with the expert/non-expert distinction. The distinction between an expert and a non-expert is a distinction in skill, while the professional/amateur is a distinction in the public and/or institutional recognition of that skill. Therefore, the late arrival of the professional category of the ‘scientist’ is irrelevant here. What matters is that individuals specialising in the empirical and mathematical study of nature were already present in the seventeenth century and working as part of learned associations, such as Accademia dei Lincei, Accademia del Cimento, The Royal Society and the Académie des Sciences, at a time when we also witness the first attempts at science popularisation. Although their professional status is different from that of modern-day scientists, just like modern-day scientists they were clearly working towards a better understanding of nature.
(V) Another criticism often directed at the diffusion model is that it cannot be reconciled with the objectives of present-day historiography. While the diffusion model takes for granted the existence of experts, non-experts and the boundary between them, histories of science popularisation generally tend to question it by focusing (1) on the ways in which categories such as ‘experts’ and ‘lay audiences’ are socially constructed and (2) on the analytical deconstruction of the ‘boundary’ between them (recall Topham’s, Secord’s and Shapin’s arguments above).
This social constructionist approach to the study of popularisation would be incompatible with a ‘diffusionist’ approach only if social constructionism is taken to an extreme, that is (1) if one ascribes to the view that not merely the notion of ‘expertise’ is socially constructed but that the skills of the expert themselves are a social construction and (2) if one assumes that the analytical deconstruction of the boundary between experts and lay audiences entails its actual eradication. In either case, there would be nothing left to analyse with the help of a ‘diffusionist’ approach. It is my understanding that scholars generally do not subscribe to this extreme version of social constructionism and that, therefore, there is room for an approach to the study of science popularisation which proceeds from the recognition of the reality of the unequal distribution of scientific knowledge in society. While existing approaches address the social construction of the notions of ‘expertise’ or ‘lay audiences’, thereby advancing the understanding of the relationship between popularisation and the cultural and political context in which popularisation takes place, I will show below that a ‘diffusionist’ approach helps us understand something different but equally important – the relationship between popularisation and the nature of scientific knowledge.
(VI) Another criticism of the diffusion model is that it does not account for the rhetorical function of popularisation materials. In contrast, many proposed alternatives place the rhetorical function of popularisation at the forefront, focusing on how popularisers talk about science and analysing popularisers’ implicit or explicit attempts to persuade the public into taking a particular view of science (Golinski, 1989; Jarvie, 1990; Lievrouw, 1990; Mazzotti, 2004; Stewart, 1992). 7
Works of science popularisation do, of course, serve a rhetorical function. However, as I mentioned above, to even count as a work of science popularisation, a text, a lecture or a visual display must also fulfil another, one might say more mundane function – communicate scientific content. Allowing for a view of science popularisation as the diffusion of knowledge opens up an opportunity to examine the scientific content in works of popularisation as historical evidence in its own right and study the efforts of popularisers to inform and educate. 8
(VII) The diffusion model has also been dismissed as a product of a particular political ideology dating back to the late eighteenth/early nineteenth century. In this period, publications and educational institutions frequently described their popularisation efforts as efforts to stimulate the ‘diffusion of knowledge’. Historians and sociologists of science popularisation have seen in these descriptions an attempt to introduce a view of expert-audience relations that catered for the social and political interests of scientists and science popularisers at the expense of those of lay audiences; to ‘create and maintain a large, passive, lay audience for esoterically validated scientific knowledge’ (Topham, 2000: 561). A view of popularisation which has been used by scientists and popularisers as a political weapon, so the argument goes, should not be used by historians as an analytical tool.
However, a closer look at the way in which the term ‘diffusion’ was used, particularly but not exclusively in the late eighteenth and early nineteenth centuries, suggests that another interpretation is possible. In this period, the term ‘diffusion’ was most often used either (1) metaphorically, to denote ‘progress’ in a very similar way to which, say, the term ‘Enlightenment’ was used, or (2) literally, as a collective description of the initiatives that were being undertaken to spread scientific knowledge as widely as possible, such as the proliferation of cheaper printed materials and the establishment of libraries and institutions for the education of the working classes. 9 This suggests that the term ‘diffusion’ carried a different type of political baggage than that suggested by Topham and others, namely, that in the mind of the popularisers, diffusion of knowledge was part and parcel of Enlightenment ideals such as the pursuit of improvement, equality, tolerance, education and happiness. While a view of science popularisation as the diffusion of knowledge is clearly not value-free, it can nonetheless be used in historical analysis at the very least because it can help us trace and evaluate the efforts that were being made in pursuit of Enlightenment ideals, however flawed these ideals might turn out to be.
(VIII) As I mentioned briefly in the introduction, historians and sociologists of popularisation often emphasise the role of personal and political motivations in the work of popularisers (Berman, 1978; Lightman, 2007; Ring, 1988; Turner, 1980). They correctly point out that the diffusion model cannot adequately deal with the role played by such motivations. However, while there is historical evidence that the popularisation of science has been used for personal gains and political purposes, it would be too cynical, and indeed historically unjustified, to assume that popularisation is exclusively an act of manipulation and political control. The scientific content of popularisation is in itself historical evidence that above and beyond any personal or political motivations, popularisers aim to educate. The diffusion model therefore opens up the possibility of investigating an aspect of popularisers’ work which has hitherto been largely neglected – popularisers’ motivation to make constantly evolving and difficult knowledge of the natural world learnable.
(IX) One of the challenges that scholarship on science popularisation currently faces is that the existing interpretations of popularisation do not form a coherent whole. The study of science popularisation, to borrow Secord’s description, has grown into a ‘fragmented’ field with a ‘rich array of research that somehow adds to less than the sum of its parts’ (Secord, 2004: 660). 10 In the 20 or so years that have passed since these remarks were made, a ‘solution’ does not appear to have been found.
The fragmentation of the scholarship on popularisation is concerning not merely because it is having a negative effect on scholarship; it is concerning because it fragments a historical process which has enjoyed an unfragmented existence throughout history. Science has been popularised continuously since the seventeenth century. It comprises a diverse corpus of works but its very continuity provides clear evidence for the presence of an intellectual process which has been sustained by factors not tied to specific individuals and their motivations, to local contexts or to particular time periods. While adopting a diffusion model approach is unlikely to resolve the problem of fragmentation, I will show below that it can help us better understand one of popularisation’s core features – why it is continuous rather than sporadic.
To summarise: the currently dominant historiographical approaches have helped scholars identify the ways in which the popularisation of science has been shaped by cultural factors, political interests and individual personalities. However, the wide adoption of these approaches has left a number of important questions unaddressed. This article does not call for the ‘resurrection’ of some old-fashioned ‘diffusionist’ view of popularisation but for a fresh approach that helps us address these questions. Given that no satisfactory alternatives have been proposed that can help us explain what the diffusion model (as described here) can help us explain, it is our scholarly duty to consider its usefulness more carefully.
4. What’s a diffusion model useful for?
‘All models are wrong but some are useful’ (Box, 1972: 202)
I offer below an in-depth analysis of the diffusion model of science popularisation, which I believe is the first attempt at such an analysis. While the popularisation of science is not solely or merely about spreading knowledge, viewing popularisation as the diffusion of knowledge helps us understand some of the core features of the process, while also allowing for variation between historical periods and across cultural and political contexts. For the purposes of keeping the article to a manageable length, the analysis will focus on the popularisation of scientific knowledge through printed materials.
Science popularisation has existed continuously throughout history
Science popularisation has a long, rich, complex but also continuous history – there has never been a period since the seventeenth century when established and breakthrough scientific ideas have not been popularised. Adopting a diffusion model of popularisation helps us account for this continuity. By emphasising the essential features of modern scientific knowledge, the model helps explain what makes popularisation necessary and continuous rather than contingent and sporadic.
(I) One fundamental feature of modern scientific knowledge is novelty. The history of modern science shows unequivocally that over the last four centuries, science has continuously and systematically produced genuinely new knowledge, starting with the ‘explosion’ of new discoveries and inventions during the period commonly known as the Scientific Revolution (Hall, 1989; Deutsch, 2012; Wootton, 2015; Weinberg, 2015). Natural scientists at that time were very much aware of the fact that their new approach to the study of nature was capable of producing new knowledge – it is not merely rhetoric that the natural philosophy that Brahe, Kepler, Galileo, Boyle and Newton produced was called, including by contemporaries, the ‘new science’ and that the titles of many important publications from that time emphasise novelty (e.g. Tartaglia’s A New Science (1537), Kepler’s Astronomia Nova (1609), Bacon’s Novum Organum (1620), Galileo’s Two New Sciences (1638) and Boyle’s New Experiments Physico-Mechanical (1660)).
And so, when we examine the science popularisation record from the seventeenth and eighteenth centuries, we find that, invariably and uncoincidentally, new discoveries, new inventions and newly discovered principles dominate the field. The poet and playwright Ben Jonson popularised Galileo’s discoveries to the English court in 1620 in a theatrical performance called News from the New World Discovered in the Moon (Johnson, 1692 [1641]). Many books on popularisation covered the principles of astronomical observation (Bonnycastle, 1786; Ferguson, 1768; Harris, 1719). Lectures on mechanics and motion were given by all major popularisers in Britain (Desaguliers, 1734, 1744; Walker, 1799). And, of course, a great deal of science popularisation appeared on Newton’s work (Ferguson, 1756; Martin, 1754). The historical record of popularisation in the following centuries shows the same trend. In the nineteenth century other sciences, such as chemistry, geology and biology also became the focus of popularisation due to the discovery of new chemical elements, ‘deep time’ and the mechanism of natural selection in evolution. The discovery of a whole new subatomic world and the formulation of a new physics of space-time preoccupied much of popularisation efforts in the twentieth century. The ability of science to produce genuinely new knowledge is therefore one of the things that makes a continuous popularisation of science necessary. 11
(II) Another fundamental feature of modern scientific knowledge that helps account for the continuous need for popularisation is its cumulative but undogmatic and, ultimately, provisional character. Why don’t popularisers ever run out of materials to popularise? It is not because they are ingenious spin doctors. History shows us that it is because science is a revisionary, never-ending process of knowledge generation that enables us to understand nature ever more deeply and with ever greater precision. Newton’s physics has been fundamentally revised in the twentieth century, the atomic theory has evolved dramatically since John Dalton, and Darwin’s theory of evolution through natural selection has been absorbed into the ‘Modern Synthesis’. This is largely why new popularisation works are needed as time passes and why nineteenth- even twentieth-century, popularisation works on astronomy, chemistry and biology differ from those produced today.
(III) Another feature of modern science that makes continuous popularisation necessary is the inherent difficulty of scientific knowledge. As the Lady in Benjamin Martin’s (1759) dialogue confesses: ‘Philosophy, I mean the Knowledge of natural Things in general, is what I should be greatly pleased in the Study of, were it not so difficult a Science’ (Martin, 1759: 1). Richard Proctor’s bestselling Half Hours with the Telescope (1868) begins with similar sentiments: ‘The student of astronomy is often deterred from telescopic observation by the thought that in a field wherein so many have laboured, with abilities and means perhaps far surpassing those he may possess, he is little likely to reap results of any utility’ (Proctor, 1868: 1). Often the source of difficulty in studying science is mathematics and the aim behind popularisation works throughout history has been to explain concepts and discoveries by non-mathematical means, such as ordinary language or experimental demonstration (e.g. Bonnycastle, 1786; Clare, 1735; Desaguliers, 1734, 1744; Ferguson, 1756, 1768; Martin, 1754; Nicholson, 1782; Smith, 1738).
If science is difficult why do non-scientists even bother trying to learn about it? In exceptional cases, it is the difficulty which makes science attractive. For example, as historian Jonathan Rose reports, the Cornish carpenter George Smith (b. 1800) ‘read every sort of book’ that came his way but he particularly liked mathematics because it was ‘slow reading’: ‘A treatise on algebra or geometry, which cost but a very few shillings, afforded me matter for close study for a year’ (Rose, 2001: 372). There are, however, deeper incentives that make science reading an attractive proposition in spite of the difficulties. Science is difficult but it is also fundamental, both in the ideal sense of addressing the ‘biggest questions’ about the universe and human existence and in the material sense since its principles offer the only way we know of making technological advances. Thus, while the inherent difficulty of science may deter people from learning it, its fundamental relevance earns people’s continuous attention.
(IV) Another feature of scientific knowledge that necessitates continuous popularisation is the nature of the scientific report, the standard form of communication in modern science. A scientific report necessarily restricts itself to describing the results of scientific investigation. Finding purpose or a deeper meaning in those results and discussing at length their wider implications beyond science itself, must be done somewhere else. That ‘somewhere else’ was, and continues to be, primarily in works of popularisation – in the ‘scientific stories’ where issues of morality and the value and meaning of scientific pursuits can be freely discussed. For example, in the eighteenth and nineteenth centuries, works in science popularisation were used as a vehicle to promote the virtues of learning and religious observance (e.g. Adams, 1789; Arbuthnot, 1701; Hack, 1832; Martin, 1737, 1749; Miller, 1857; Wesley, 1763). While the view that science draws people closer to God persisted till much later, in the nineteenth century, alternative arguments about the secular virtues that would accompany the attainment of scientific literacy were also gaining prominence (e.g. Huxley, 1866, 1868; Maxwell, 1990 [1856], 1871; Royal Institution of Great Britain, 1854; Youmans, 1867; Huxley, 1895). In the twentieth century, popularisers and scientists writing for the public grappled with questions about the actual and potential misuse of science which came to the fore after breakthrough discoveries in physics allowed physicists to build weapons with previously unimagined destructive power (Bronowski, 1956; Oppenheimer, 1948; Russell, 1952). And, in the twenty-first century, issues with the public understanding of the nature of science, science literacy and trust in science have increasingly come to the fore in the literature on popularisation (e.g. al-Khalili, 2022; Oreskes, 2019; Tyson, 2022).
The existence of fundamental characteristics of modern science that make continuous science popularisation necessary, rather than contingent, is clearly accounted for by the diffusion model. While it is true that the popularisation of science is not only a matter of making scientific content learnable, it is also true that we cannot hope to understand and appreciate the nature and role of popularisation without explaining why the learning of science requires mediators in the first place. The answer, as we have seen, cannot be adequately found in the personal motivations of popularisers or the political context of their work but relates directly to the fact that science is new, constantly evolving, difficult and amoral knowledge about nature.
The unequal distribution of scientific expertise and the origins of science popularisation
Scientific work requires cooperation, perseverance, imagination, technical ingenuity and mathematical skill. In other words, doing, not just learning, science is difficult and depends on a gradual, cumulative and collective acquisition of expertise, the passing on of that expertise from one generation to the next and the critiquing of that passed-on expertise by the succeeding generation. As time passes and as the rate of scientific development increases, the amount of knowledge and technique that one needs to master and the amount of work that one must do before one acquires enough expertise to be in a position to contribute to scientific knowledge also increases. The clear expert-non-expert distinction on which the diffusion model rests is therefore not artificial but, on the contrary, stems from reality. It is not a distinction that we use to judge some people superior in the ability to understand and study nature and others inferior, but to acknowledge the fact among people with similar abilities some go down the path of scientific education and work, and some do not. A diffusion model, therefore, provides an accurate reflection of the dynamics of intellectual work in a modern, industrial, heavily science-dependent society.
By affirming the expert-non-expert distinction, the diffusion model provides a further advantage – it allows us to account for the origins of popularisation in the period of the Scientific Revolution. To explain how, we need to consider what conditions bring about diffusion. First, let’s take as an example the diffusion of energy. The laws of physics tell us that as long as there is a difference in the amount of energy, or an energetic slope, between two areas, energy will flow from the area where energy is abundant (we can call this area ‘a source’) to the area where it is less abundant (we can call this area ‘a sink’). What matters in a diffusion process is how steep that energetic slope is. The steeper the slope, the more energy diffuses.
The period of the Scientific Revolution can be described as a period in which the intellectual slope (akin to an energetic slope), ‘linking’ natural philosophers involved in the ‘new science’ and the rest of the public, increased to an unprecedented degree. On one hand, this was a period in which a wave of new knowledge swept the fields of mathematics, physics and astronomy. However, at the same time, it was a period of a growth of ignorance in the sense that with the discoveries of new empirical facts about nature, most people found themselves in a situation in which they did not yet know what a small minority was discovering or inventing. This was not because most ordinary people at the time were uneducated or lacked access to education, although this was the case in most places. It was because genuinely new and difficult ideas and fundamentally different kinds of ideas were being produced by scientists at an unprecedented rate. Better explanations of natural processes followed – explanations which were previously unknown, unintuitive and difficult to grasp. It is this great disparity in knowledge generated by the new way of doing science that explains why systematic efforts to popularise started when they did and have not ceased to this day. Popularisation did not emerge at the time of the Scientific Revolution by chance – it had to emerge if there had indeed been a revolution in knowledge. The large corpus of works on science popularisation is itself material evidence that a Scientific Revolution unfolded and that cognitive progress in natural science was taking place.
The values popularisers promote are not arbitrary
Most popularisation texts contain: (1) a preface or an introduction in which popularisers declare the purpose of the work and explain how they intend to achieve that purpose and (2) a main part dealing with their scientific subject. For example, here’s how John Bonnycastle prefaces An Introduction to Astronomy in a Series of Letters from 1786:
The principle object in view [. . .] has been to avoid, as much as possible, all complicated mathematical principles and calculations [. . .]. [. . .] The great object in view was to unite truth with perspicuity, and to give a general idea of the operations and phenomena of nature [. . .].
The ‘purpose’ of the work, Bonnycastle continues, is to give ‘general instruction’ (Bonnycastle, 1786: iv–v). We would have reasons to doubt his words, if in the pages that follow Bonnycastle did not go on to explain the movement of the planets, the structure of the Solar system, basic facts about the Earth’s geography and the principles of optics. Bonnycastle’s book does indeed offer clear evidence for ‘general instruction’, for the diffusion of scientific knowledge, however, successfully or unsuccessfully. The same aims are to be found, declared and pursued, in many other printed works by popularisers from the eighteenth century to the present day, including J. T. Desaguliers, Benjamin Martin, James Ferguson, Adam Walker, Jane Marcet, Margaret Bryan, Mary Sommerville, John Tyndall, Richard A. Proctor, Robert Ball, James Jeans, Richard Feynman, Carl Sagan, Brian Green and others. 12
There are, of course, popularisation works, whose aims clearly go beyond that of instruction. Science popularisation is about making science learnable to a wider audience, but it is also about making scientific knowledge meaningful. While these two aims often intertwine in the same texts, there are examples throughout history where speculation about the wider meaning (rather than the ideas) of science is at the forefront of popularisation works. Notable examples include J. T. Desaguliers’ allegorical poem The Newtonian System of the World, the Best Model of Government (1728); Benjamin Martin’s eulogy A Panegyrick on the Newtonian Philosophy (1749) and T. H. Huxley’s lay sermon On the Advisableness of Improving Natural Knowledge (1866), Jacob Bronowski’s Science and Human Values (1956), Carl Sagan’s The Demon-Haunted World (1996) and Jim al-Khalili’s The Joy of Science (2022). Does the fact that these works, and many others like them, were not written to teach science per se, mean that viewing popularisation of science as the transmission of knowledge is misleading? No, it does not. Making science learnable and making science meaningful are not processes which take place independently; on the contrary, explaining to people who lack scientific expertise why they should learn about science is an important part of teaching them what science is. Moreover, the interpretations and moral messages that popularisers promote in their works are not arbitrary, but conditional upon the nature of science and its cognitive power. For example, Desaguliers would not have been able to recommend the structure of the Solar system as described by Newton as an analogy for good government if that structure had not been shown to be essentially correct and mathematically sound. Benjamin Martin would not have been able to make a strong case for nurturing scientific literacy among the population in his Panegyrick, if science had not shown itself capable of widening the scope of natural knowledge and describing nature accurately, thereby, for Martin, elevating the human spirit and bringing humans closer to God’s creation. Huxley and Bronowski could not have made out a viable case regarding the value of scientific education if science operated in a dogmatic and authoritarian way and lacked the cognitive power to discover facts about the natural world, some of which also serve to improve the material conditions of human life. And Sagan and al-Khalili would not have been in a strong position to recommend to the wider public to adopt the values of science – values that include commitment to fight bias and accept uncertainty, to respect evidence and to change one’s mind in light of evidence – if the history of science itself did not contain compelling evidence of the cognitive progress that science has achieved thanks to the adoption of these values by the scientific community. Historians cannot adequately study the values promoted in works on popularisation without factoring in first the progressive nature of the scientific knowledge in which those meanings are rooted. A diffusion model, which proceeds from this assumption, is therefore a useful way of looking at popularisation, enabling us to analyse not only the knowledge but also the values that popularisers aimed to promote.
The role of science popularisers
The physical process of diffusion does not involve active transportation – if a slope exists between ‘the source’ and ‘the sink’, diffusion will follow. If that is the case, then how can we use it to model popularisation as a process in which popularisers play an active role?
Diffusion cannot take place if there is a barrier between the area where matter (or energy) is in high concentration and the area where that concentration is low. If a barrier exists and is removed, then diffusion follows. This is precisely the role that popularisers play in popularisation – they remove, or at least attempt to remove, the barriers to understanding that exist between scientists and lay audiences, so that a process of diffusion of knowledge can take place. As I explained above, such barriers exist because science is genuinely new, difficult and amoral knowledge. How popularisers attempt to remove these barriers, what means or strategies they use, varies from one historical period to the next depending on the audience, context and available technology. However, their role remains the same throughout history. And if they succeed in helping those without scientific expertise to grasp the ideas that only those with scientific expertise are capable of producing, then knowledge is successfully transmitted – diffusion has taken place.
For example, the popularisation of the work of Isaac Newton in the eighteenth century was so prolific and diverse that it gave rise to a whole corpus of popularisation works loosely described as ‘Newtonianism’. Notable examples include Henry Pemberton’s scholarly exposition of Newtonian mechanics A View of Sir Isaac Newton’s Philosophy (1728); J. T. Desaguliers’ allegorical poem The Newtonian System of the World, the Best Model of Government (1728); Francesco Algarotti’s six-dialogue classic Sir Isaac Newton’s Philosophy Explain’d For the Use of the Ladies (1739) and James Ferguson’s Newtonian bestseller Astronomy Explained Upon Sir Isaac Newton’s Principles (1756). Here we have four different works – they target different audiences, adopt different presentation styles, place Newton’s discoveries in different contexts and offer different interpretations of these discoveries. However, in all four works, the substance and the scientific ideas being communicated and interpreted are the same. Where popularisers played an active role was in deciding how best to overcome the cognitive barriers presented by the physics and astronomy of Newton; otherwise, they would not be able to convey their intended meaning. The great differences between the four works can be largely attributed to differences in these decisions.
If scientists produce knowledge and popularisers play an active role in choosing how to transmit that knowledge, what about lay audiences? Where do they fit in the diffusion model? Since the seventeenth century, we have witnessed rapid scientific and technological progress. This progress is neither inevitable nor irreversible and is only possible because scientists have established procedures that enable them to produce accurate empirical knowledge and because the conditions exist for wider society to appreciate and support the pursuit of science. Knowledge, the diffusion of which is popularisation’s core function, helps lay audiences to make more informed, and thereby more democratic, choices with regard to the place of science in society. While a diffusion model assigns the role of knowledge producers to scientific experts, it also assigns an active role to lay audiences in sustaining and supporting scientific endeavours.
5. Conclusion
Writing in 1987, the statistician George Box advised his students: ‘Remember that all models are wrong; the practical question is how wrong do they have to be to not be useful’ (Box and Draper, 1987: 74). Historians of popularisation will do well to consider his advice. No model of popularisation can give us a full understanding of the process. The diffusion model itself leaves out much of the granular detail of popularisation in different contexts and time periods. However, the rejection of the diffusion model has been a move in the wrong direction. The model is the vital test for determining what is and is not science popularisation. If a work intended for popular consumption attempts to convey scientific knowledge accurately and honestly then whatever else it does over and above this, it is science popularisation. What is more, the model succeeds in capturing the fundamental features of the popularisation of science – as a long-term historical and intellectual process – better than any other alternative on offer. It helps us understand what precise role popularisers play, why the values popularisers promote are not arbitrary, why popularisation came to exist when it did and why it has been needed ever since.
Footnotes
Acknowledgements
I wish to thank Ronald Wilson for reading the manuscript of this article and for his valuable editorial contributions. I also wish to thank Lindsay Paterson and John MacInnes for their helpful feedback and insightful comments.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: For the research towards this article, I received funding from the British Academy (Grant number: PF19/100118).
