Castells is a professor of urban geography at Berkley. He has written a number of books and articles about geography, the city, and the information society, including a three-volume analysis of contemporary capitalism, titled The Information Age. Garnham (2004, p. 165) refers to this as “the most sophisticated version” of the theory of the information society.

Castells' analysis involves economic, social, political, and cultural factors. I will focus on the economic, with a brief introduction to his analysis of space and the changing role of the nation state, and follow with an outline of some critiques of his work. Regrettably, this leaves much unmentioned, such as his theory of timeless time, of the social divides in modern cities and societies, or his examination of specific cases of social action in the context of what he calls the information city.

The Network Society

Castells (2000a; 2000b) claims that we are passing from the industrial age into the information age. This historical change is brought about by the advent of new information technologies – particularly those for communication and biological technologies. Society remains capitalist, but basis of the technological means by which it acts has changed from energy to information. This information is of central importance in determining economic productivity. Communications technologies allow for the annihilation of space and for globalization; the potential for rapid and asynchronous communication also changes the relationship to time. And, while he explains that networks are not a new form of social organization, they have become a “key feature of social morphology” (2000a, p. 5). This is because communication technologies, such as the Internet, allow for decentralization of operations and focusing of control, increasing the effectiveness of networks relative to hierarchical structures. Of business he writes, “[t]he main shift can be characterized as the shift from vertical bureaucracies to the horizontal corporation” (2000b, p. 176).

According to Castells, power now rests in networks: “the logic of the network is more powerful than the powers of the network” (quoted in Weber, 2002, p. 104). Some networks, such as that of financial capital, are global in scale. Networks also exist within and between businesses, where the organizational unit has shifted from being capability-oriented (e.g. accounting, human resources, etc.) to being project-oriented. Resources – including employees, consultants, and other businesses – are brought together to work on a particular project, then dispersed and reallocated when the task is complete. The ability of an actor in the network – be it a company, individual, government, or other organization – to participate in the network is determined by the degree to which the node can contribute to the goals of the network. This new environment requires skilled flexible workers: the organization man gives way to the flexible woman (Castells, 2000a, p. 12). This leads to a binary process of inclusion and exclusion from the network. The people at the bottom are those who, with nothing to offer the network, are excluded.

Capital and Labor

Castells distinguishes the terms “information” and “informational”. He says that information has been an essential component of all societies, whether capitalist or not. In the new “network economy”, information becomes a key factor in economic productivity. Today, for example, for example, the flow of capital into currencies, commodities, and stocks is based upon access to information about relevant topics, from international politics to climate change, weather predictions, and social trends. In that sense, the importance of information in contemporary society is not new. What is new, he claims, is the informational shift to the manipulation of information itself: the “action of knowledge upon knowledge itself” (Castells, 2000b, p. 17) is now the basis to increased productivity.

Castells argues that the operation of networks, particularly their ability to include or exclude actors on the basis of their ability to contribute to the goals of the network, has individualized labor: “the work process is globally integrated, but labor tends to be locally fragmented” (2000a, p. 18). The constantly-shifting “variable geometry” of networks individualizes labor. In consequence, the era of industrial-age conflict between production-based classes ends: individual workers no longer comprise classes, much of the power of the capitalists has shifted to those who manage the networks (for example, the fund manager has greater control than the investors whose money s/he manages), and the class of human capitalists has been replaced by the collective faceless capitalist of the network.

Despite the disappearance of capitalists and the proletariat, exploitation and differentiation remain. In Castells' analysis, labor is fundamentally divided into networked labor, which serves the goals of the network, and switched-off labor, which has nothing to offer the network and in the context of the network economy is non-labor (that is not to say their labor has no value – the realm of discarded labor is also the realm of criminal organizations outside the networks; Castells mentions that the networks can make “perverse” use of these masses of people). Networked labor is itself divided into two groups. Self-programmable labor – such as financial analysts, company officers, journalists – manages information; it is flexible and skilled. Its interests coincide with the goals of the network. Generic labor (including many workers in natural resource, manufacturing, and service industries, also minimum wage and sweatshop labor) is deskilled, interchangeable and disposable; for these people, the goal is simple survival so as not to be relegated to the class of switch-off irrelevant labor.

Flows vs. Places and the Role of the Nation State

Castells is an urban geographer, so his examination of space is central to his work. One of his key spacial characterizations of the information age is the “space of flows”. This is the domain of networks – of capital, of information, of business alliances, etc. He argues that “While organizations are located in places, . . . the organizational logic is placeless, being fundamentally dependent on the space of flows that characterizes information networks” (Castells in Nyíri, 2004, p. 23). This space of flows challenges what Castells calls the space of places, including regional communities and nation states. Networks lead to the “destruction of human experience” (Castells in Nyíri, p. 10): power is separated from political representation, production from consumption, information from communication.

The inclusion/exclusion logic of the network “switches off . . . people and territories dubbed as irrelevant from the perspective of dominant interests” (Castells in Nyíri, 2004, p. 7). This enforces domination: “[d]omination depends . . . on the simultaneous capacity of . . . elites to articulate themselves and disarticulate the masses” (Waterman, 2004, p.49). Groups may choose to develop their own networks with their own goals, but if they wish to interact with the dominant networks in society they must adapt to the goals of those networks.

In opposition to the space of flows is the space of places. While “the space of flows can be abstract in social, cultural, and historical terms, . . . places are . . . condensations of human history, culture and matter.” (Castells, 1990, p. 14). Hence, resistance to the space of flows of the networks arises in the form of communities oriented around places. These communities are often closed, based on fundamentalist or tribal identities. Those dislocated or excluded by the network society (such as unneeded labor) naturally gravitate to such identities of communal resistance – or are relegated to them: “elites are cosmopolitan, people are local” (Castells, 2000b, p. 446). Castells has argued (1990, p. 21) that for communal identity to be a site of democratic resistance communities must reach out and build links (network) with other communities of other cultures. However, his position has since shifted (Nyíri, 2004)1, and recent history has shown that the groups that isolate themselves from the network are often reactionary and exclusionary.

This clash between flows and places has consequences for the nation state. Sovereignty is ceded both upwards to the space of flows and downwards to regional and communal groups. States are caught in a bind. If they represent the identities of the communities within them, they isolate themselves from the network. If they obey the network logic, they cease to represent their populations; they become nodes in the network and surrender their sovereignty: “[n]ation states will survive, but not so their sovereignty” (Castells in Nyíri, p. 8)

Critiques of the Network Society

The first and simplest critique of Castells is that his depiction of the contemporary world is so familiar, even derivative (Webster, 2002, p. 115). Its propositions about the character of contemporary society seem commonplaces: the increasing importance of information and knowledge, the increasing speed and of financial and other transactions and the consequent destabilization, the growing gap between the knowledge-rich and the knowledge-poor, the sense that we are indeed in a time of social and technological discontinuity. However, this can also be seen as a success in capturing contemporary life.

More serious criticisms target Castells' analysis of the role of information, of production, and of the relationship between between informational labor and capitalism. These critiques call into question his claim that the present economic and social situation is a new age, rather than a continuation of industrial capitalism.

Modes of Production and Development

Webster (2004; 2002) and Garnham (2004a) have accused Castells of technological determinism. Castells distinguishes between the mode of production – which determines how production surpluses are distributed – and the mode of development, which drives productivity and thereby determines the amount of surplus production. The mode of development is technological: according to Castells, the industrial mode of development centered on energy; since the oil shock and stagflation of the 1970s capitalism has seized upon informationalism to produce growth. However, he does not believe that capitalism created the informational mode of development as a solution to this challenge, instead tracing it to “the autonomous dynamics of technological discovery and diffusion” (2000b, p. 59).

The theoretical problem raised by Webster is the relative autonomy of the mode of development. The mode of development influences social relations, but is not determined by capitalism and, according to Castells, may survive it: “the new economy . . . may well outlast the mode of production where it was born” (Castells, 2000a, p. 11). For Webster, this implies that the mode of development follows its own technological logic, is “in key respects, beyond the reach of politics” (2004, p. 17), and therefore despite Castells' protestations to the contrary, his theoretical basis – like that of Bell – is technologically determinist.

Garnham's (2004a) criticism centers on productivity. He asks whether what he calls the “thermodynamic” model of productivity, based on inputs and outputs of energy, can be applied to information, which he suggests is non-entropic2. Castells claims that the network economy relies on productivity increases made possible by information technology. But, says Garnham, there is little evidence of such productivity increases. Furthermore, he says, the “lack of a stable calculable relationship between the value of outputs . . . lies behind the historical difficulties in commodifying information” (Garnham, 2004b, p. 191). In other words, even if information output is increasing productivity, this activity is difficult for capitalists to measure and commodify, therefore it is difficult to integrate into a capitalist economy.

Alternatively, Garnham (2004a) says, productivity can be examined in terms of consumption and investment and the relations of production. Here, both he and Webster criticize the relatively autonomous role of the mode of development, emphasizing instead determining role of the mode of production. Technology is put to work at the behest of capital: “the informational mode of development is developed for and put at the service of a set of property relations and the goal of accumulation, not vice versa” (Garnham, 2004a, p. 174).

Capital and Labor

This downplaying of the role of capital also applies to Castells suggestion that much of the power of capital has been appropriated by informational (self-programmable) labor, particularly those actors who are at the juncture of different networks (Castells, 2000a, p. 16). As Webster (2002, p. 114) points out, this is very similar to the concept of a meritocracy, in which individuals have power in proportion to their knowledge, skills, or the value of their labor. This hierarchy of power therefore results, in Webster's words (p. 115), from “a 'natural' form of inequality that is supra-social, although of inordinate social consequence”. Again, the informational mode of development appears to be autonomous of – and in fact supersedes – the capitalist material base. A meritocratic model, of course, sidesteps the social and historical process – affected by factors such as wealth, education, and social relationships – by which labor is created: of how individual workers come to be self-programmable, generic, or irrelevant.

The problem with informational labor is more involved, however, for – as with other theories of the information society (Webster, 2002, pp. 115-119), it is a slippery concept to define. Castells includes such a wide diversity of tasks under the rubric of informational labor that it is difficult to accept the interests and values of these workers as a group. Webster provides the examples of journalists, stockbrokers, and surgeons. The relations between these occupations and the information they deal with is anything but consistent, and the group as a whole anything but homogeneous. (Is it level of education? communication? influence?) In fact, it seems that the other groups he mentions – commodity labor, surplus labor, etc. are far more homogeneous – in their work and in their common interests – than is informational labor.

Castells also claims that the information age is still a capitalist age (although he suggests the informational mode of development could survive the end of capitalism), but it is post-capitalist. In his model, capitalists are no longer the ruling class: their power has been usurped by the networks. The true logic of the system is the logic of the networks. But this is not new. Marx wrote (quoted in Garnham, 2001, p. 240):

The function fulfilled by the capitalist is no more than the function of capital – viz. the valorization of value by absorbing living labor – executed consciously and willingly. The capitalist functions only as personified capital, capital as a person, just as the worker is no more than labor personified.

Garnham's comment (2001, p. 240-1) is that “a given person or group can only be described as capitalist in those moments when s/he or they are acting in conscious and willed accord with the logic of capital accumulation.” Again, as both Garnham and Webster explain, the capitalists remain in control. The propertied class has better access to education, and its members are dominant in the top managerial positions which Castells claims are in control of the networks (Webster, 2002, p. 118).


First, the familiarity of Castells' analysis, and the connections he draws between the disparate phenomena and factors he examines, is valuable (here I have looked at the forces dissolving the sovereignty of the nation state).

Webster and Garnham present credible critiques of Castells' concept of an information age. The network society is more convincingly a development of industrial society than an entirely new construct. Capitalism is still the economic basis for society; the manifestation of capitalism's power in faceless networks is a phenomenon going back to Marx's original critique of the system in the 19th century. Meanwhile, the social changes wrought by information technology are not yet on the same scale as the intense period of upheaval that corresponded to the technologies whose effects were greatest in the first half of the 20th century, including the oil industry, the automobile, the aeroplane, electrification, the telephone, and nuclear physics. These criticisms affect not only the question of whether we are entering a new age, but also some of Castells' theoretical foundations.

Garnham's (2004a) wrote his criticism of Castells' treatment of productivity in 1998. The argument has since been weakened by more recent studies of the matter which have found evidence for significant productivity growth arising from the adoption of information technology. For example, Stiroh (2002) claims that annual productivity growth for the nonfarm business sector averaged 1.3% between 1973 and 1995, but the average rate has increased to 2.5% between 1995 and 2000. He credits this increase to information technology.

Castells does seem to be of two minds about the role of technology. On one hand, he describes it as enabling rather than determining: “The internationalization of the economy is only possible because of information technology. It is not information technology which pushes the process.” (1990, p. 9). His (2001) recounting of the history of the Internet also shows how that technology is a product of social forces, themselves shaped by the context of the capitalist society in which they were embedded. On the other hand, Garnham and Webster are correct that Castells seems to be elevating technology to a priviledged topic of analysis. In the context of a wider examination of society and the relations of production, his claims for the epochal impact of information technology seem suspect. However, I find that accusations of technological determinism go to far. Current theories of the social construction of technology do not reject technological determinism out of hand; theories of the social shaping and co-production of technology (e.g. Jasanoff, 2004; Mackenzie & Wajcman, 1999), for example, reject the binay distinction between technological and social determination, and instead examine how society and technology to shape each other.


1 This conflict between community tradition and timeless compressed space recalls Innis' distinction between time- and space-based media. Garnham (2004, p. 167) also notes Castells' debt to Innis.

2 In fact, information, energy, and entropy are intimately connected, as Claude Shannon illustrated in 1948 as part of his transmission model of communication. Whether this is relevant to Garnham's claim is beyond my the scope of this paper and of my understanding of information theory.


Castells, M. (1990). The infomational city: A framework for social change. Canada: University of Toronto.

Castells, M. (2000a). Materials for an exploratory theory of the network society. British Journal of Sociology Vol. No. 51 Issue No. 1 (January/March 2000) pp. 5-24.

Castells, M. (2000b). The rise of the network society (2nd ed.). U.S.: Blackwell Publishing.

Castells, M. (2001). The internet galaxy. U.K.: Oxford University Press.

Friedmann, J. (2004). Reading Castells: zeitdiagnose and social theory. In F. Webster & B. Dimitriou (Eds.), Manuel Castells, Volume III (pp. 152-164). London: Sage Publications.

Garnham, N. (2001). Contribution to a political economy of mass-communication. In M.G. Durnham & D. M. Kellner (Eds.), Media and cultural studies (pp. 225-252). U.S.: Blackwell Publishing.

Garnham, N. (2004a). Information society theory as ideology. In F. Webster (Ed.), The Information Society Reader. New York: Routledge.

Garnham, N. (2004b). “Information society” as theory or ideology. In F. Webster & B. Dimitriou (Eds.), Manuel Castells, Volume III (pp. 185-196). London: Sage Publications.

Jasanoff, S. (2004). The idiom of co-production. In S. Jasanoff (Ed.), States of knowledge: the co-production of science and social order (pp. 1-45). New York: Routledge.

MacKenzie, D. & J. Wajcman. (1999). Introductory essay: The social shaping of technology. In D. Mackenzie & J. Wajcman (Eds.), The social shaping of technology (pp. 3-27). U.K.: Open University Press.

Nyíri, J. C. (2004). Review of Castells, the information age. In F. Webster & B. Dimitriou (Eds.), Manuel Castells, Volume III (pp. 5-34). London: Sage Publications.

Stiroh, K. J. (2002). Information technology and the U.S. productivity revival: A review of the evidence. Retrieved November 27, 2005 from

Waterman, P. (2004). The brave new world of Manuel Castells. In F. Webster & B. Dimitriou (Eds.), Manuel Castells, Volume III (pp. 35-57). London: Sage Publications.

Webster, F. (2002). Theories of the information society. U.S.: Routledge.

Webster, F. (2004). Information and urban change: Manuel Castells. In F. Webster & B. Dimitriou (Eds.), Manuel Castells, Volume II (pp. 15-39). London: Sage Publications.