Introduction

We live in a networked world. The concept of a network of friends, of businesses or indeed of computers is pervasive in our conversations, newspaper articles or business plans. For many, the increasing importance of innovation and adaptation to turbulent environments is changing the nature of interaction with other organisations and as a response we increasingly encounter more networked inter-organisational relationships such as alliances, partnerships, clusters and communities of practice. These organisational forms often involve interaction between numerous individual organisations such that the flows of information and resources between them are complex. As a result these networks of organisations are becoming a dominant organisational form in the 21st century (Cravens & Piercy, 1994).

For many business sectors, the development of networks of organisations may be new or novel. For example, 'Just In Time' manufacture, which requires a network of suppliers working together, has been in place since the early 1980s (Huson & Nanda, 1995). In comparison, tourism has always been a networked industry and the usual description of tourism as a fragmented and geographically dispersed industry belies a pervasive set of business and personal relationships between companies and managers in businesses such as national tourism offices, hotels, attractions, transport, tours, travel agents and restaurants. It is this network of relationships that allows the tourism industry to deliver its product and to overcome the problems of fragmentation. Therefore it can be argued that the tourism industry provides the ideal context for study of networks.

The network concept is based around relationships between entities such as organisations or people (termed nodes), and the properties of networks studied by researchers relate to the structure of these relationships. The study of networks may be considered to have a number of paradigmatic characteristics (Wellman, 1988: 82) focusing on:

• Structural advantages and constraints on behaviour.

• The discovery of groups through their relationships rather than a priori allocation to categories.

• The overall structure of multiple relationships in a group rather than that between a particular pair of alters (in the language of network analysis, a particular node is identified as 'ego' and those nodes that ego has relationships with are termed 'alters').

One consequence of this approach is that it makes problematic the classical economic concept of a market as a homogeneous collection of identical suppliers and buyers. Instead, studying networks presupposes that the individuals do not act in isolation and with perfect information, but that the behaviour of individuals is profoundly affected by the pattern of relations that they may (proactively) develop. In studying networks the focus therefore is on relations rather than attributes, and on structured patterns of interaction rather than isolated individual actors. A second implication is that the fundamental basis for the study of networks is different from other areas which study the attributes of people or organisations. Instead, network analysis studies relationships (Knoke & Kuklinski, 1991).

Definition of a Network and Network Analysis

Originally, the concept of a network was a metaphor for the complex interactions between people in the community. However, with the development of quantitative approaches the concept of a network became formalised and related to mathematical theory. In graph theory a network is a:

finite set of points linked, or partly linked, by a set of lines (called arcs) ... called a net, there being no restriction on the number of lines linking any pair of points or on the direction of those lines. A relation is a restricted sort of net in which there can only be one line linking one point to another in the same direction, i.e. there are no parallel arcs. (Mitchell, 1969: 2–3)

Transferred into sociology, a network is defined as a specific type of relation (ties) linking defined sets of persons, objects or events (Mitchell, 1969), and the sets of persons, objects or events on which a network is defined are called actors or nodes. Thus a network consists of a set of nodes, and ties representing some relationship between the nodes. Today, there are many definitions of a network but as pointed out by Jarillo (1988: 31), many have been developed by applying this basic definition to new areas such as the study of organisations where, for example, Gamm (1981) defines a network as a system or a field comprised of organisations and inter-organisational relationships.

Given this definition of a network, network analysis (or social network analysis) is an approach and set of techniques used to study the exchange of resources among actors such as individuals, groups, or organisations (Haythornthwaite, 1996). Because of this focus on relationships, the techniques used to analyse networks differ substantially from mainstream statistical methods that demand independent units of analysis. Network analysis therefore uses a set of integrated techniques to draw the patterns of relations among actors and to analyse their structure. The analysis is conducted by collecting relational data and organising it into a matrix and calculating various parameters such as density or centrality.

Network analysis has increased in popularity through the 1990s as an analytical framework, encouraged by the emergence of theories of society that emphasise relationships and integration. This is due in part to the effects of globalisation, which encourages alliances and linkages across organisations and nations, and to the greatly enhanced ease in communications encouraged by the wide diffusion of information technologies. In business and economics, network analysis represents a new organisational paradigm, drawing upon the competencies-based theories of the firm, where relationships shape and constrain organisational performance.

Within the tourism literature, the use of the concept of a network appears logical and delivers a number of useful outcomes for the analysis of tourism destinations and organisations....