Introduction: the echo chamber
Donald Trump's victory in the US presidential elections in 2016 and the vote for Brexit by 17.4 million people in the United Kingdom in the same year are just two recent examples which demonstrate the unprecedented power of online social networks to effect political and ideological change. The role of social media was crucial in both these seismic political events, and the psychological and emotional micro-manipulation of individuals through social media channels has been revealed as the new, dark side of politics through the exposure of the practices of British consultancy firm Cambridge Analytica and their even shadier clients.1 The rise of Facebook, Twitter, and Instagram has fundamentally shifted the way that politics is done, but also the way that human social life is done: we have moved into an era of global online social networks that have the capacity to incorporate widely geographically separate individuals into close-knit groups and communities that bolster and support them and their ideological stance on the world, while the companies that operate the system simultaneously gather and manipulate the rich data freely given by individuals into agglomerated information about the ideological proclivities of different social groups. From such a combination comes the power to market thingsâand more importantly, ideasâto certain individuals, through emotion: drawing on both an individual's feeling of belonging in a group of like-minded people and on the fears they have about the boundaries of those groups. We are witness to and enmired in our own personal echo chambers, whether this concerns fears about vaccinations, trans rights, or imminent ecosystem collapse.
In network terms, this is homophily, the tendency for like to attract like. We live in times dominated by homophilic online social networks, and by the powerful emotional perception of the edges of those networksâwhere âourâ group comes into contact, and conflict, with âtheirâ group. Such issues around group behaviour have always been at the heart of the study of human history; however, it is only in recent years that archaeologists and historians have had the technological capacity to analyse intra- and inter-group behaviours as forming through the structures and dynamics of interconnected social networks. Partly as a result of the introduction of social media into our daily lives, archaeology and history have seen a huge increase in networks as part of the vocabulary used to talk about and think through the study of the past. Over the last fifteen years, network methods and language have increased in popularity and scope in archaeological and historical applications,2 making network approaches to the past one of the strongest emerging research areas in our disciplines. Subjects of study range widely from the interactions of prehistoric Kuril islanders in the north of Japan to the Pueblo cultures of the American Southwest; and from the material remains of the Epipalaeolithic in the Near East to those of the late pre-colonial period in the Caribbean.3
As a central premise, network analysis is concerned with how entities relate to other entities. Such entities are most often called nodes, which are connected via edges, ties, or relationships. What is classified as a node ranges from brain cells to individual people to types of artefact to households to nation states and everything in between; and the ties that connect them range from neural, marriage, type of material, consumption practices to trade, and many more.4 Archaeology and history have a more limited range of material than social science, but there is still considerable diversity of data under study using network methods by archaeologists and historians. This diversity reflects also the diversity of scholars with different disciplinary backgrounds: some scholarship has drawn on physics and mathematics to push forward the side of theory that derives from complexity science;5 others have focused on the power of computing and agent-based models (ABM) to develop approaches which work to recreate possible past network scenarios in silico, and make inferences from similarities between their archaeological dataset and the computer-generated patterns.6 Others have been inspired by the work of Bruno Latour in particular, adopting actor-network theory (ANT) to challenge the way we see relations between people, things, and institutions and to re-envision the world as created through these relations rather than existing a priori.7 And others have derived their methods from sociology and social network analysis (SNA), where the relationships between people and how these relate to social structures are of primary importance. This last approach has perhaps been most effective for those studies with rich historical datasets, such as bodies of letters or papyrological data,8 which have had the capacity to examine past social networks between individuals.
Although there is diversity in what is studied, most archaeological uses of network methods rely on similarities between material culture assemblages as proxies for the strength or frequency of social relationships, such as trade, thus elucidating mostly regional interconnections between sites.9 Within this body of work and across network science more broadly, the âsmall world networkâ10 has had a huge impact: a network structure featuring strong social groups or clusters that are linked together by random âweak tiesâ.11 This network structure, and specifically the presence of these random weak ties between the strongly bound clusters, enables the easy transmission of certain kinds of information between separate groups, or clusters. These weak ties are therefore described as possessing strength, and the small world network has been observed in many different situations across time and spaceâranging from the colonising world of the Archaic Greeks to the spread of disease epidemics in modern America.
However, there are issues with âcomplexâ information transferâsuch as new religious beliefs, behaviours, or normsâacross such weak social ties, precisely because of their weak social integration within a group structure. Weak ties possess little emotional strength, little power to persuade. For complex information transfer that requires trust or reciprocity in order to be accepted and spread, we need to look instead at the social ties that carry emotionâwhether this is romantic or familial love, friendship, respect, admiration, or fearâand which bind individuals into meaningful relationships and communities: the strong ties that form human beingsâ closest and most important social bonds of family, friendship, and close association through geographical neighbourhood or occupation. Strong ties are central to our social understanding of ourselves. Today, we can have strong tie relationships that are global in scaleâfor example, family members on the other side of the world, and technology that enables us to maintain those relationships regularlyâbut in the past, strong tie relations were more localised. Of course, a sense of community or group identity was possible across longer distances, with some aspects of strong emotional social ties maintained by repeated trade or mobilities, the sending of letters, and by religious practice,12 but at a general level, attachments to place, social norms and behaviours, language, profession, and so on which help to form and maintain strong tie clusters will have been more local in character.
How then, did new ideas or innovations that require introduction through trusted strong tie social bonds travel beyond the local to cross geographical distance? Because the transmission of complex behaviours across great distances can be seen materially in the archaeological record, in the case of certain aspects of âglobalisationâ,13 or the adoption of new ways of doing things, such as technological innovations in the production of pottery or metallurgy, or the spread of new religious beliefs. Weak ties that connect local clusters are the obvious answer in network terms, but they cannot provide the whole answer in social terms. The chapters gathered together in this volume showcase the importance of looking beyond weak ties to focus on the role that strong social ties have in the transmission of complex information which, in order to be accepted and lead to social change, require individuals to be bound together by relationships that are structured through mutual trust, memory, and reciprocity. They highlight the vital role of places of gathering in the process of information transmission, in this case sanctuaries; the role of narrative and storytelling in creating a sense of community even across geographical space; and the role and control of social systems in order to facilitate or stifle new information transfer.
The strength of weak ties and the small world
To understand the importance of strong social ties more fully, we must start with weak ties, and why they have been seen as so socially powerful: why they are understood as structurally strong. Mark Granovetter coined the term âthe strength of weak tiesâ in the title of his seminal paper,14 because sociologically, people's relationships or social ties are characterised by their strength or weakness: where strength is ascribed to close, repeated, and regular contact between individuals, with whom they form local groups or clusters; and conversely, weakness is ascribed to acquaintances or people that are not so well known, acquaintances who are part of different social groups. Strong ties exist between dense groups of friends and family, which can also include geographical neighbours and colleagues, binding them into a reciprocal social network, where social cohesion is marked by trust, repetition of interaction, shared memories, and shared emotions. In sociological terms, these individuals have been argued as forming âclosed triadsâ, where all members of the triad know each other.15
By contrast, a âweak tieâ in a social network describes the connection that exists between people who are otherwise integrated into different tightly knit social networks, and despite the fact that they usually represent a relationship between the two people that is infrequent, casual, or temporary, research has shown time and again how effective these relationships are at spreading certain kinds of information. This is because the network structure of the tightly knit strong tie social groups of which people are generally part is localised: the people with whom we interact most are likely to have access to the same kind of information as ourselves.16 A person who is a weak tie connector to another social cluster on the network is able to transfer information from their social group into our own. In network terms, these weak tie connections between different strong tie social clusters have a powerful cumulative effect on the structure of the network as a whole: they connect the separate, localised strong tie social groups together and enable information to transfer from one part of the network to the other with relative ease. These ties, although relationally weak, are structurally strong.17 This phenomenon has been described as the âsmall worldâ. The weak ties between clusters in a small world network enable local information to have global, that is, full network, reach. It is the weak ties which enable this network quality of a small world: they bring the capacity for âshort path lengthâ, the network term which describes the ability to directly access other disparate clusters.
The power of weak ties has been amply demonstrated through real-world examples. The spread of disease is a particularly good case in point, where the arrival of one contagious individual into a strongly tied social group can cause the infection of pretty much everybody. This is a âsimpleâ contagion, that is, the process of exposure, infection, and spread is not dependent on anybody having an opinion about the process, or a choice about whether they are infected. They just catch the virus (though wearing a mask and staying socially distant helps). Granovetter's example was more sophisticated, exploring instead the way in which information about the labour market was transferred, and where individuals in a group with fewer weak social ties were seen to be at a disadvantage in terms of knowing about job openings at the right moment. If we are ...
