Just like vulnerability, social capital is a term currently widely used and discussed (but only recently also in hazard research: Dynes 2002; Nakagawa and Shaw 2004; Kirschenbaum 2004; Bohle 2005; Pelling and High 2006). What is more, the concept “has become one of the most popular exports from sociological theory into everyday language”, despite the fact that it “does not embody any idea really new to sociologists” (Portes 1998, 2).
Although only rarely reflected upon, the concept of social capital stems from at least two distinct strands of thought: sociology of social inequality and political sociology. The first conceptualisation goes back to Bourdieu (1986; similarly Coleman 1990, 302) who conceived social capital as “resource of individuals”. The second and much more influential perspective, which emphasises the role of social capital as collective asset, is mainly connected to Putnam’s idea of (not) “bowling alone” (Putnam 1993 and 2000).5 Bourdieu (1986, 248) defines social capital as the “aggregate of the actual or potential resources which are linked to possession of a durable network of more or less institutionalised relationships of mutual acquaintance and recognition”. These resources are based on the affiliation to one or several social groups. It is both the quality and quantity of these social relationships and the resources (further social, but also economic and cultural capital) which can be mobilised via this network which makes up the social capital of an individual. This is an important difference to Putnam who conceptualises social capital as a collective good of a community indicating its respective level of “civicness” (for a critical appraisal: Portes 1998, 18–20).
Despite all the differences, in both conceptualisations social networks play a crucial part. Social networks form an important nexus between the individual and social structures. Therefore, network analysis is interested in the “in-between”, i.e. in the structure, quantity and quality of social relations as units of analysis (Burt and Minor 1983; Schenk 1983; Pfenning 1996). In the context of floods and other hazardous events, one might assume that social networks function as resources for information, material compensation, emotional support and physical help and are something exclusively “positive”. However, network theorists provide ambiguous hypotheses concerning the actual role of social networks in different situations. There is, first of all, the “strength-of-weak-ties” hypothesis (Granovetter 1973, 1983) which holds that heterogeneous social networks—resting in various social and local contexts—have more and in particular more diverse information about a certain topic (in its original application referring to labour markets and getting a job) than a dense network consisting of persons who are similar in various socio-economic and socio-demographic dimensions. With respect to coping with floods and their consequences, a variety of information channels (hence: networks of weak ties) might help an endangered person to assess a hazardous situation more appropriately than a network built upon strong ties. Then, also the coping behaviour might be more adequate.
But, secondly, there is also evidence for the “strength of strong ties” meaning that dense networks of people in a similar situation are exploited as a resource. Frequently interacting (i.e. densely connected) persons are more likely to share similar information, attitudes and beliefs (with a similar approach: contagion theory; Scherer and Cho 2003). The most prominent examples in this respect are networks of innovation (Burt 1987) or—from the realm of urban sociology—the emergence of ethnically segregated neighbourhoods in big cities and of ethnic entrepreneurship which built upon the strong ties of kinship and cultural-linguistic similarity, respectively (Portes 1998, 12–3). When transferred to floods, on the one hand such networks might be obstructive in the immediate pre-phase of an extreme event since they could hinder the reception of diverse and possibly even ambiguous information.6 But, on the other hand, they are able to create an immediate flow of resources in the entire period of a disaster (information, physical and emotional support, economic capital etc.).
Without denying older traditions in disaster research which strongly focused on communities (Barton 1969; Erikson 1976; Couch and Kroll-Smith 1991; Mitchell 1996), there are some good reasons for dealing with social networks (and social capital) instead of focussing on communities in their ambiguous meaning of being both locally based and socially constructed. Kirschenbaum (2004, 96) points out that traditional community-based approaches usually defined their object of research by taking physical and geographical borders as a matter of fact instead of referring to subjectively defined borders and cross-local networks.7 But regardless of whether communities, social capital or social networks are in the focus, it is apparent that most disaster research is interested in the recovery phase and the effects the disastrous event has on social cohesion and community relations (Beggs et al. 1996; Sweet 1998; Nakagawa and Shaw 2004). Only a few authors deal with the role of social networks and social capital in earlier stages (Barton 1969; Hurlbert et al. 2000; Kirschenbaum 2004).
In this report, social capital will be used in a non-romantic manner (which is one of the criticisms related to Putnam). Thereby, we will follow principal conceptual ideas of both Bourdieu and Putnam, hence taking into account social capital as an individual resource (i.e. related to the various social networks a person creates and belongs to and the economic, social and cultural resources they provide) as well as a collective asset (i.e. a community resource for which trust and shared norms are basic requirements).
At this point we also want to introduce our notion of local knowledge. Usually, in the discourse on natural disasters it is agreed upon that this form of knowledge is a valuable resource for mitigating the impact of a hazard, since the local population developed specific strategies over time for coping with crises (Blaikie et al. 1994, 64–9). We will incorporate this dimension into our analysis, by focusing on the constitution of this form of knowledge in the interaction with the physical as well as the social environment. In this respect, local knowledge is a form of knowledge, which was developed and tested in the local environment and which is therefore held as highly reliable and accepted. However, the operationalisation of “local knowledge” by means of a standardised questionnaire is hardly possible in a meaningful manner. Therefore we approach this dimension via social networks and their spatial arrangements suggesting that exclusively or predominantly locally based networks continuously create and recreate local knowledge.
Social networks as defined above predominantly refer to informal ties people have to friends, neighbours and kin. However, in the context of a disaster threatened residents usually have to deal also with representatives of organisations, such as fire brigades, municipal authorities, the Red Cross, the police, the army etc. Therefore, when analysing trust (e.g. as regards information announcing a disastrous flood about to come) and the like, also the distinction between formal and informal networks according to Matthiesen (2005; with a slightly different terminology) makes sense. Formal (Matthiesen: “hard”) networks are “strategic cooperation structures within formal-institutional structures and systemic functions, with clearly defined strategic goals, explicit benchmarking processes (milestones) and […] with a defined end (death of network)” (ibid., 10). In the following, all those governmental and non-governmental organisations are subsumed that are part of official disaster protection efforts. The network has a clearly defined beginning (in Germany for example Warning stage 1), a clearly defined end (termination of the disaster declaration) and encompasses such different institutions as the regional government, the municipality, the police, the army, in Germany the THW (Technisches Hilfswerk; Federal Agency for Technical Relief), as well as non-governmental organisations such as the local fire brigades and various aid agencies (Streitz and Dombrowsky 2003). Informal (Matthiesen: “soft”) networks consist of family-members, friends, neighbours and colleagues. They are defined, above all, by “intensified communication processes and shared tacit/explicit components of knowledge” (Matthiesen 2005, 9). Hence these networks are more or less identical with the social capital as defined above.