Notes on: DeLanda, M.  (2006) A New Philosophy of Society.  Assemblage Theory and Social Complexity, London: Continuum.

[important claims right at the end]

This is an exercise in social ontology, which is normally seen as realist—‘a commitment to the mind – independent existence of reality’ (1).  However, it is necessary to admit that human minds also have a role in maintaining social entities.  But they are autonomous from the conceptions we have of them.  The ‘real history and internal dynamics’ of social entities need to be studied adequately.  However it is clear the categories also affect behaviour when people become aware of them, but this shows that the referents of general conceptions can shift—there is a whole network of organisations and institutions which supports these categories providing a realist context [the example is hyperactive children and the institutional practices that classify them, 2].  These practices and networks must be seen as conception independent, unlike as in social constructivism.  They are assemblages, constructed by definite processes, and ‘in which language plays an important but not a constitutive role’(3).

Deleuze’s philosophy is the basis of assemblage theory, even though he and Guattari said little about it.  The concepts that might be used are dispersed throughout the work, and the style often does not allow ‘a straightforward interpretation’ (3).  However, there is no need to engage in excessive hermeneutic analysis.  Instead, we can reconstruct Deleuze, providing different definitions and arguments and ‘entirely different theoretical resources’.  If necessary, the results might be described as ‘”neo-assemblage theory”’ (4).

The idea will be to explain how wholes emerge that are not reducible to their parts, while avoiding the flaws of rival conceptions such as Hegelian totalities and other forms of essentialism.  The case studies to be chosen involve the links between micro and macro levels of reality, but avoiding the reductionism of rival approaches, which include the ‘phenomenological individualism of social constructivism’ (4), and the over-socialised notions of Parsons, Durkheim and Marx.  There are even those who wish to reduce everything to an intermediate level ‘such as praxis’ and this is the stance of Giddens.  There are also specific analyses at different levels between the micro and macro, where the ontological status is unclear—assemblage theory can be used to model them.  Assemblages can emerge from individuals, or from institutions, from communities, from organisations, cities and other entities.  Social complexity needs to be preserved, although the analysis is restricted to Anglo American examples.  Larger scale organisations are not meant to be seen in the geometric sense, but as extensive properties.  The empirical capacities of entities are also important.  The focus on ontology is justified since normally it is left implicit.  Thus ‘philosophers…  can greatly contribute to the job of ontological clarification’ (7). [if only!]

Chapter one.  Assemblages against totalities

The organismic metaphor is common, but unhelpful, since it works on the basis of analogies.  Spencer, and above all Parsons, have developed this metaphor most fully, but in connection with functionalism which has since been rejected. [even Mertons' much more concrete version?]  However, a general theory still persists that suggests that wholes displays some totality or organic unity, and that the parts gain meaning only by their relations to these wholes—‘relations of interiority’ (9).  Similarly, the part has no meaning on its own detached from the whole.  There are collectivities made up of selfsubsisting parts and relations of exteriority, but they do not count in this approach: they are mere aggregates etc (cf the distinction between a statistical aggregate and a social group).

Even Giddens remains influenced by this structure, which connects behavioural procedures, material resources and actual practice as integral parts rather than as a mere aggregate: ‘The end result of this is a seamless whole in which agency and structure mutually constitute one another dialectically' (10).

Without such internal relations, there can be no emergent properties for Hegel and others, but complex interactions between parts require heterogeneous combinations.  We need to consider the relations between parts as involving capacities to interact, which are limitless.  These capacities are not just constitutive parts of a whole.  This leads us to assemblages in Deleuze, 'wholes characterised by relations of exteriority' (10).  Parts may be detached from an assemblage and plugged into another, and the properties of wholes cannot be reduced to those of parts, since emergent capacities relate the parts producing an [empirical] synthesis.

Deleuze uses examples of symbiotic relations between plants and insects in ecosystems, not organisms.  Wasps and orchids must form necessary relations, but these relations are still contingent, based on empirical issues.  Elements are heterogeneous, although this might imply different values—in this way, more phenomena can be grasped as assemblages, including species and organisms themselves.

Assemblages are defined along one axis of materiality - expressivity, which can be combined or mixed.  The second dimension is territorialization - deterritorialization, referring to the degree of sharpness of the boundaries of an assemblage, the stability of its identity.  Again, this can vary over time.

Social assemblages show these qualities in face to face conversations, in interpersonal networks and hierarchical organisations, especially those in cities or nation states.  They have material components such as food, labour, buildings, machinery.  The expressive components are not just linguistic ones but include non-verbal behaviour, and matters such as reputation or image, or the nature of collective behaviour, or rituals that reinforce legitimacy of hierarchies.  Territorialization has a literal meaning in that interactions are always located in particular places with boundaries, but there are also non-spatial processes which increase internal homogeneity—social selections and sorting for example.  Deterritorialization arises from things such as the development of communication technology which avoids geographical boundaries.  Territorialization is important for producing permanent articulations of wholes [persisting across space—and time?].  There are also 'specialised expressive entities such as genes and words' (14) which similarly synthesise wholes.  Deleuze insists that all entities can express themselves, but human expressions are much more complex following the appearance of these forms of expression.  [DeLanda insists that atoms express themselves by creating particular patterns in radiation as in a spectograph].

Expression need not always be functional [the example given here is that the patterns produced by fingerprints have no biological function, just an important social one, but only once policing develops].  However, at certain thresholds, functionality increases dramatically—the development of genetic codes; the emergence of language.  Each of these can be seen as assemblages in their own right, but they produce 'a second synthetic process' (15) [also referred to as a double articulation, 15] —coding.  Social hierarchies can be coded in various ways such as traditional or rational.  However, ‘many social assemblages [unlike biological organisms] are not highly coded or highly territorialized'(15).

There can also be decoding in biology and sociology.  Animal behaviour need not be rigidly programmed, for example in their sense of territoriality.  Social decoding can arise from informal conversations where rules are suspended.  In neither case do codes define some essential qualities—they are simply components among many others.  DeLanda wants to demonstrate this by discussing language 'last and as a separate component' (16) in order to deny its centrality.

Assemblages always exist in populations which produce recurrent processes.  Assemblages then interact empirically with each other producing some emergent properties.  Larger assemblages may also emerge from smaller ones articulating together.  This implies that the macro - micro problems in sociology can be bridged [transcended or sidestepped really].  Thus markets emerge as concrete organisations made out of people and the goods they exchange.  Initially, they must be literally territorialized.  As urbanization intensifies, they become articulated together into regional markets, provincial markets and finally national markets.  The scale varies, and the articulation concerned is by no means simple or functional [but uneven and emergent].

Thus assemblages are made up of self subsistent parts articulated by exterior relations.  They have two main dimensions which help us see the roles played by components—material or expressive, stabilising or destabilising identity.  A third dimension can be added, involving 'processes in which specialised expressive media intervene' (19), which can produce flexibility or rigidity, coding or decoding.  All these are traceable to population characteristics, and these include still further synthetic processes.

Finally, it is important not to reduce causality to linear causality ('same cause, same effect, always') (19),  when speaking of the various mechanisms involved.  For one thing, social assemblages can involve reasons and motives, and the subjective components need to be explained.  However, it would be wrong to simply choose between subjective notions and linear causality. Linear causality has often been misleadingly connected with logical implication.  Instead, causal relations are productive – this is not the same as saying that one implies the other.  Causes can also involve complex entities, and changes in some components can turn the entity into a cause.

Linear causality usually involves atomistic events, but internal organisation can produce modified effects—low intensity ones, or amplified ones.  Here thresholds are important in producing effects.  Sometimes, they are low enough to register an effect every time a cause operates, but there are other states such as catalysis where different causes can lead to the same effect.  [The example is hormones stimulating growth when applied to the tips that the plant and inhibiting growth when applied to roots] (20).

The idea that causes always produce the same effects can be challenged by statistical causality.  Here it is a matter of acknowledging complex interfering events in combination with causes, not internal processes that modify.  Together, these processes lead us to reject the idea of the 'block universe'—'the world is a seamless web of reciprocal action…  A block of unlimited universal interconnections' (19).

Material components usually involve all these types of causal interactions, while 'expressive ones typically involve catalysis' (22).  Language usually plays a catalytic role, for example.  This assumes complex serial organisations such as the nervous system, but the catalytic role of language is so important that special explanations are required—reasons, motives.  Weber argued this in preserving room for both causal and motivated action.  Causal bits have been typically ignored lately, part of the drive to see everything is a text.  However, adequacy can involve assessments of causal relevance—when interacting with material objects, but also about making judgements about adequate performance (23).  Even reasons may look rather like causes in the case of traditional actions, although other types involved beliefs or desires.  Typically, however, both causal and meaning adequacy will be involved, 'complex mixtures of causes, reasons and motives' (24).  Sometimes this complexity is lost in reducing the meaning of action seen as rational—to match means to ends clearly involves all sorts of problem solving skills, including 'consideration of relevant causal events' (24).  Similarly, traditional routines may expressive congealed rational understanding if they lead to ‘successful causal interactions with material entities, such as domesticated plants and soil’.

Social explanations must draw on the full variety of causal interactions, including the notion of thresholds to explain why one individual might differ from another, and to work with probabilistic notions.  These are important with human beings who may be making intentional choices.  However there are also ‘collective unintended consequences’ producing statistical results (24).  With reasons, collective behaviour can be explained by the affects of socialization, but even this must be seen in a probabilistic terms: ‘the effects of socialization should always be pictured as variable and a proper object of study should be how this variation is distributed in a given population’ (25). 

[The same goes for capacities though. In Intensive Philosophy, these are seen as almost accidental or random, necessarily in order to make the case for emergence and haecceity. In social assemblages it is possible to see capacities as socially distributed though, as in Bourdieu’s various ‘capitals’ – this bit is missing so far?. NB DeLanda says below that Bourdieu overdoes the point with his notion of habitus. Deciding between them is an empirical matter of examining variations?].

C2 Assemblages against essences

Essences are introduced in different forms, as when general categories get reified—‘taxonomic essentialism as opposed to its platonic variety’ (26).  The example is the classification of creatures into genus, species and individuals, as in Aristotle.  The trick is to identify necessary differences or natural kinds.  Recent evolutionary theory shows the contingency and fuzziness of these essential differences.  Instead, each category is best seen as an individual entity, on different scales.  The argument can be applied to chemistry as well, replacing the notion of ‘hydrogen in general…  [with]…  the objective reality of large populations of hydrogen atoms’ (28).

Essentialism follows a flawed method by operating inductively.  Instead, the historical processes that produce the entities are required.  Any assemblage is produced by a specific history of territorialization and coding, and it is always open to destabilisation.  In this sense, assemblages are always ‘unique singular individuals’, differing only by scale (a flat ontology).  Individual people, communities and organisations must all be seen as individual entities (28).

Categories above that of genus, such as kingdom or phylum involve abstract plans which are non-metric or topological, ‘a space of possibilities (the space of all possible vertebrate designs, for example)’ (29).  Possibilities structured in this way, are better seen as possible capacities.  The possibility spaces can be studied as phase spaces and this structure described mathematically in terms of attractors, degrees of freedom and the rest (see Intensive Philosophy)   Possibilities may be highly constrained, as in the case of minimising energy.  Social spaces cannot yet be defined as formally, but they may have ‘a much more complex distribution of topological invariants (attractors)’ (29).  Invariants can also be described as universal singularities.  Distributions of these replace the notion of genera, while species can be defined as individual singularities.  The two kinds are separated by historical differentiation not logical differentiation.

Deleuze uses the notion of a diagram as a set of universal singularities, or the equivalent of a body plan, with possibilities and degrees of freedom.  National markets can be seen as an articulation of regional markets, themselves connections of local markets, with each category as an individual singularity.

Weber’s ideal types are the equivalent of diagrams [and the example is types of authority].  The ideal types are not essences but are best seen as ‘three universal singularities defining “extreme forms” that authority structures can take’ (30). The degrees of freedom correspond to issues like the extent to which the office is separated from the incumbent, and the degree of routinization of activities.

The analysis of assemblages is not simply a logical one, following from the discovery of necessary differences, but must involve ‘causal interventions in reality’ (31) [the examples given are physical damage to cells. Maybe pseudo-causal interventions in science labs also fit as implied in Intensive Philosophy? ].  Social analysis should discover actual mechanisms in operation [with actualisation used in the Deleuzian sense]. The analysis of phase space may be one of the resources which can be used, and quasi causal operators need to be understood as well as productive causes.  Quasi causes will be focused on in the book at the expense of actual causal mechanisms, but actual mechanisms need to be understood especially if they lead to emergence, and then to discussion of the macro and micro levels [an example of how the correct formulation of a problem is crucial].

In the first place, there may be more than two levels.  Micro and macro can therefore be used relativistically to describe components and emergence, at whatever spatial scale [there is an interesting note on page 127 denying the usual reified notions of macro and micro and citing a certain D Gerstein in support].  At each stage, the macro emerges from the micro, but this is not methodological individualism—there are not just reified rational individuals in assemblage theory, but populations and ‘subpersonal components’ (32).  Further, subjectivity is emergent and can become more complex in assemblages—for example it gets identified with formal roles in organisations, since ‘the emergent whole reacts back and affects [identities]’ (32). [Sounds like Bourdieu here on the ways in which selection processes make the rejected believe that they never wanted to join in the first place].

Relations between parts and wholes are not simple.  Wholes exist at a number of different scales, and parts can belong to different wholes.  Some networks cut across others, others combine together for political purposes such as collective bargaining.  There can be hybrids of interpersonal networks and institutionalised organisation.  The populations concerned vary as do the physical locations.

Wholes react back causally on parts, as Bhaskar has argued [a note on page 128 says that Bhaskar’s realism is ‘very close to Deleuze’s’, but differs from it in that Bhaskar is an essentialist, arguing that groups possess some real essence or nature].  In this way, wholes provide constraints and resources for parts, and ‘opportunities and risks’ (35) [I am still not sure that this formalist approach really gets to the mechanisms of matters such as class domination specifically].Even deterritorialized assemblages can affect their parts, but with a diminished capacity, as a ‘weak link’ [a contact with someone outside, linked to a reference about how to get a job] (35).  DeLanda goes on to talk about low and high degrees of solidarity [nearly Durkheim here]. Market places are also locations with lots of weak links, and economic constraints.

Wholes also exercise ‘causal capacities’ by interacting with each other, as in forming political coalitions, which again supply resources and constraints. These activities might be better explained as the activities of people, individual activists or officials?  However, there is also redundant causality, ‘many equivalent explanations of the process in question’ (37).  In some cases, for example individualist explanations might be unnecessary if similar outcomes emerge in different circumstances, say if changes in personnel fail to affect organisational policies, so that ‘the emergent properties and capacities’ of the organisation remain roughly the same after such a change’ (37).  This argument applies with assemblages at larger scales-- if changes in personnel lead to different outcomes we are justified in using individualist accounts, including reasons and motives.  [DeLanda assumes that this is usually not so].

Thus organisations have causal effects, and use people ‘as a medium of interaction’ (38), sometimes as networks or smaller organisations.  Again, causal redundancy can help sort out the importance of different components—if urban resources affect the conduct of a war more than the activities of the professional military, for example, urban centres become the agents, and the military the medium.

Emergence also has to be considered.  It is not just the original emergence historically, but the process that maintains identity, a constant struggle for territorialization against deterritorialization.  The tendency of populations to produce assemblages might not describe the specific historical process—most new organizations recruit from earlier organisations, so that there are already organisations in existence.  Also existing assemblages can produce new parts as they attempt to maintain themselves—so that networks and organisations can emerge from existing cities as well as the other way around.  [A brave attempt to avoid formalism, but there is much empirical complexity and possibly some social patterns here that require explanation, unless we are just picking up on some possibilities].

Finally, explanations can operate in different spatial scales, and the Napoleonic revolution in warfare [discussed in his book on the war machine] is an example, since it transformed local limited battles into national ones.  There were many causes, including political reforms, technological ones, Napoleon’s own genius and his influential networks.

So each entity is best considered as an individual, even large assemblages.  They all have an ‘objective existence independently of our minds’, but without invoking a notion of essence (40).  There is a need to carefully explain relations of parts to wholes.  Wholes are autonomous to the extent that they can causally affect parts and interact with each other.  This autonomy is limited by the distribution of universal singularities, however, the diagram at the virtual level.

There are also temporal scales to consider.  We might investigate these by asking whether, for example, organisations take longer to change than people, or rural ones longer than urban.  Some changes may be set by interactions between assemblages, such as the slow emergence of rational bureaucracies, not explained by ‘decisions by individual persons’, for DeLanda (41), which are therefore causally redundant.  Even strategic planning can operate in a different temporal scale to individual decision-making, since a large amount of resources have to be mobilised first, and different sorts of inertia overcome, requiring large networks and alliances of people: the commitment to change must exceed that of particular individuals.  There are also external pressures such as technological change, which can have an uneven impact on departments inside organisations, and produce resistance or negotiation, which will itself depend on authority structures—so even if individuals are committed to change, it might not happen on an individual timescale. There are lags. There are also discretionary powers among agencies [this reminds me very much of Colebatch on the difficulties faced by national policies when implemented].  Again, this is seen in terms of normal processes of adjusting objectives to political realities.

There is also the problem of social order [rendered here as the persistence of institutions beyond a human life].  Do assemblages have some natural life span of their own, and is this affected by size?  DeLanda thinks there is no simple relation.  Some networks do not endure after the people comprising them die off, while some communities can survive a change of personnel.  Some organisations are short lived [the example given is restaurants!, 44], while cities are typically long lived.  However ‘most social assemblages larger than people do tend to outlive them on average’ (44).  This can arise from overlapping generations and recruitment, and also semantic information through language—‘specialised media of expression’, special assemblages (44).

Such assemblages are special because they can offer ‘variable replication’, featuring ‘populations of replicators’, which produce learning, and the capacity to operate at different spatial scales simultaneously [actually the examples are best when they refer to genes, 44-45].  However, these special assemblages should not give rise to some essentialist argument, as when language mediates all experience [and DeLanda specifically mentions social constructivism here, 45, with a reference to Berger and Luckmann].  This is a reification of linguistic categories, despite the constructivist argument that general categories are really stereotypes: there is still an ‘ontological assumption that only the contents of experience really exist’ (46).

C3 Persons and networks

[A real eye opener for me, raising all sorts of questions about the origins of subjectivity that I had not thought about since reading JS Mill years ago.  Unfortunately, this one involves a diversion into the works of Hume, which I have not yet read]

Actual persons are best understood as assemblages, with sub personal components.  We need to investigate these in order to fully develop an ontological model, although much will remain to be explained.  [At this point, Delanda seems to have begun the usual drift into theoreticism.  The point now is to find a theory that will agree with an assemblage approach—‘a plausible model of the subject which meets the constraints of assemblage theory, that is, a model in which the subject emerges as relations of exteriority are established among the contents of experience’ (47).  Before, the point was to explore ways in which assemblage theories might fit social sciences, but now we’re looking for approaches that will support assemblage theory, which is now not to be tested but to be taken for granted].

This will require us examining empiricism, the argument that sense experience is the basis of all knowledge.  Deleuze read Hume slightly differently, as an account of subjectivity, one which precedes the linguistic and transcendental arguments of Kant.  What we need is a collection of distinct sense impressions.  Ideas derived from these, in Hume, are replicas of impressions, with lower intensity.  Impressions can be classified as visual, oral, olfactory and tactile, and so can the passions ‘from pride and humiliation to love and hatred’ (48).  These are separate and heterogeneous and have to be articulated.  It is a different approach from that which sees impressions as examples of general [linguistic] categories.

A unity between the impressions and ideas arise through association.  DeLanda intends that these should be seen as exterior in the same sense as the actions of causes.  Association between ideas can then be understood as the actions of particular operators—grouping by contiguity, relations of resemblance, constant conjunction (‘the habitual pairing of causes and effects (48)).  These operators produce wholes with emergent properties.  They are also genuinely exterior, not defined in terms of the properties of the ideas themselves.  They are shared among all humans according to Hume, again not in an essentialist way but as a series of species properties.  These only look natural because they are slow to change, or identified with an isolated species in reproduction terms.

This formal model leaves out contents, though, such as the actual choice of ends and the role of passions or habits.  Nevertheless, the human subject is seen as a pragmatic one.  Subjectivity has been explained as emergent, and as both in terms of causes and personal motives or meanings.  As with all assemblages, there are material/expressive and territorialization/deterritorialization processes of work.  [DeLanda seems to be searching round rather for examples, and following rather formalist procedures.  It seems driven by the need to find examples to fit the models rather than the other way about.  For example the material aspect means bodily mechanisms including neurological ones, labour in producing associated links.  Like all discussions of the body, I’m not sure if there’s a strong case here involving biological determinism, or just the routine acknowledgement that humans have bodies.]  The expressive aspects include the non linguistic and linguistic components, remembering that linguistically represented ideas are in fact expressions of impressions.  Territorialization is done through habit—‘for Hume…  A more powerful force sustaining the association of ideas than conscious reflection’ (50) and various habits or routines to maintain personal identity.  Identity can be destabilised by things such as madness, sensory deprivation and so on].

There is also a positive aspect of changing identity through acquiring capacities and skills.  These permit new assemblages, like those with machines that one has learned to master.  These skills are more flexible and adaptive that might be seen with a rigid linear causality (but not if we allow non linear causes).  Linguistic components are important, but not all-constituting.  After all, all languages are relatively late developments with humans [highly debatable here].  Other animals may also possess the ability to match means and ends without language.  Language is best seen as augmenting these prior forms of intelligent behaviour.  Only language makes possible complex combinations and therefore complex ideas [in other words most of the things that distinguish humans?] DeLanda thinks that this is still compatible with the need for exterior relations, because of the characteristics of language, especially if its grammar can also be traced to non linguistic forms [and there is a reference to Harris, and to his own discussion in Nonlinear History].

In particular, language shapes beliefs, a process of bringing idea ‘closer to…  Impressions’ (51).  This adds to their intensity.  Again, particularly important characteristics of language are denied by this definition of intensity: it is the intensity of belief that motivates people not semantic content, as seen in desires [no particularly Deleuzian definition here?].

There are still problems, but at least we have given a compatible account of empirical subjectivity.  Nevertheless, interactions between people are also very important—‘ephemeral assemblages’ especially face to face conversations (52).  Goffman has done much to show how for example additional layers of identity such as public faces emerge in conversations.  Goffman apparently has an exterior notion that has non linguistic aspects such as glances and gestures as unintended expressions of states of mind and body [supported with a reference to interaction ritual].  Goffman’s discussion of embarrassment also show the limits of linguistic management—the situation has a momentum and intensity of its own requiring repair.

There is an attempt to fit the two axes to conversations—material resources in copresence [the argument gets a bit stretched because attention and involvement are also seen as material resources, 54].  There also technological inventions in communication [still seen as resources compared to some fundamental face to face encounter].  Words are clearly important, but so are non-verbals, especially if one is deliberately constructing an image [the universalisation of American middle class man].  Physical borders in space and time territorialize, and there are special places where people are agreed to converse.  Deterritorialization can be produced by anything, including embarrassment, a special interest for Goffman.  Technology is effective only in ‘forcing participants to compensate for the lack of copresence in a variety of other ways’ (54)—DeLanda needs to read Deleuze on cinema!].

The language going on there should be seen as both semantic and pragmatic.  The latter is arguably the more important for Goffman.  At least it will be useful to distinguish conversations where the semantic content is the most important from those where it is of little consequence [shades of Schutz’s ideal types here?] The semantic dimensions do offer more potential for puffing an identity.

Interpersonal networks emerge from routinized conversations.  Again these are best understood in terms of exterior relations, including patterns of links and how links are given properties by factors such as gender or race.  There are also emergent properties in any network, however, with properties such as frequency and reciprocity.  Densities are particularly important—a densely connected network ‘has the capacity to store local reputations and…  to deter potential cheaters’ (56).  The stability of a network can also indicates its emergent and non personal qualities.  This produces the solidarity of communities, expressed as personal reasons and motives such as ‘altruism…  [or]…  strict calculations of reciprocity’ (57).  [Getting a bit Parsonian here? Can the pattern variables be far?]

Relationships have a material basis as well, involving expenditure of time & resources and also physical relations like looking after other people’s kids.  There also various non linguistic activities which build solidarity and trust.  As usual ‘actions speak louder than words’ (57) [this seems to be the way in which DeLanda weasels around the issue of the importance of language: he simply treats non verbal communication as not language?].  Interpersonal networks may be (de)territorialized and conflicts can sharpen boundaries and develop community identity.  Solidarity can therefore lead to exclusion limits on members’ autonomy [Durkheim needed here].  Social mobility and secularisation go the other way.  So do communication technologies that enable dispersal [all very formalist]. 

Linguistic components proper include shared stories and categories, with reference to an historical sociologist Tilly.  Narratives simplify the solidaristic processes, leaving out long-term or hidden aspects, including those related to unintended consequences.  Solidaristic stories have particular roles at the margins of communities.

Political activity in occupations can produce political alliances such as social movements.  Since these involve a struggle, ‘a movement typically breeds a counter movement, both of which should be considered component parts of the overall assemblage’ (59).  [The danger of the banalities of conflict theory, or the notions of the functional nature of conflict lurk here]. Tilly does a lot of work, especially on the ways in which conflict resolution developed in Britain, involving claims to legitimacy.  Successful claims also involved material aspects such as the ability to organize mass rallies, or to effectively control them.  Enduring organisations emerge, ‘specialised association’ (61).  However, there are also ‘protest cycles’, for Tilly, as new opportunities and threats emerge.  Sometimes these lead to effective polarisation and revolutionary change.  Much of the struggle is over categorisation and classification, principally because these attract different rights and obligations [in other words, there is a material basis to social constructions, 62].  Classifications are about material resources as well as semantics.

Social classes are even larger assemblages, which arise following differential access to resources and constraint.  There are social processes which rank groups.  Tilly again has his own take, denying any basic divisions and seeing classes as political groups [parties for Weber].  Bourdieu is seen as offering an excellent example, based on cultural as well as economic resources, a distinction which ‘corresponds, roughly to the one between material and expressive resources in assemblage theory’ (63).  Distinction is dynamic, and proximity is an important issue.  These are seen as relations of exteriority, and Bourdieu undertakes statistical analysis to connect material position and lifestyle.  ‘Dispositions’ look compatible with assemblage theory on subjectivity, but Bourdieu then introduces the notion of habitus which is almost automatic, and which minimises all the differences between the motivations, reasons and causes of behaviour.  Habitus ensures an unthinking submission to order, as punters make a virtue out of necessity.  The habitus even controls who may make conscious choices, and thus becomes an underlying master process.

This is where DeLanda disagrees, and says there is no need for a master process, and the submission to order cannot be taken for granted but must be explained ‘in terms of specific enforcement mechanisms’ (65).  These may include a density of networks, or modern practices of compliance.  Bourdieu sees economic and cultural capital structuring social space in general, but DeLanda argues for analysis of concrete social entities [in a very abstract way!].  Resources are emergent, and there are also genuinely external ones such as technological or natural resources.  Bourdieu makes his mistake because he believes in ‘the linguisticality of experience’ (65) [with a reference to Language and Symbolic Power].  Tilly and DeLanda insist that groups are real not just phenomenological, based on actual concrete practices [Bourdieu too surely?  A bit of erratic scholarship from Delanda here? I think the quote he cites from Language ( p.105)... derives from an introduction to a section on the phenomenal subjective perception of categories, which are then explained by differential resources in the naming of the worlds etc, just as DeLanda himself does]. They arise when categories gain economic significance, sometimes through a self fulfilling prophecy.  Nevertheless, boundaries are ‘contingent and precarious’ (67) threatened by social mobility, or technological innovation [idealism here, and Bourdieu is needed to see how cultural qualities define and regulate these threats].  We also need a good account of political organization and governments.

[DeLanda’s notes referring to the habitus suggest that an assemblage reading sees the habitus as a topological space including habits and routines that make up individuals, ‘the space of possibilities for different combinations of habits and skills’ (133).  But surely the point is that this habitus is already deeply structured by social class and social reproduction, that class is the main attractor.  Other possibilities might exist in principle, and this is surely discussed by Bourdieu, or implied at least when he refers to dominant culture as ‘arbitrary’.  Nevertheless, thinking of other combinations might be a very useful thing to do.] 

C4 Organisations and governments

This concerns those organisations with authority structures.  As before, they have to have resources to enforce obedience, and components to develop their legitimacy.  Weber’s three ideal types are useful, as long as we remember that actual organizations may be a mixture of the three.  Rational legal forms have come to dominate in modern states.  A particular feature of them is that they acquire behaviours independently of the people holding the office in question, thereby qualifying as actors in their own right as argued before.

The usual expressive roles are available, based on individuals or symbols.  DeLanda acknowledges that the very rationality claimed by an organisation could be ceremonial, especially where outcomes are complex [the examples are mental health clinics, legal agencies and schools].  Mostly, this is acknowledged in acts of obedience, but disobedience must also be occasionally dealt with by punishment.  This leads to Foucault on the regulation of bodies, their dispersal in space and time, the way they are monitored, ceaseless inspection and recording and so on.

There can be boundary disputes over jurisdiction.  There can also be crises of succession, producing phenomena such as the routinization of charisma.  Written records are important, but mostly in terms of logistics ‘not the type of writing that lends itself to endless rounds of hermeneutic interpretation’ (74) [this is a further demarcation of ‘straightforward language…  The material form of writing documenting relatively simple facts’, which permits the anti linguistic case]. Group beliefs are also important, sometimes integrated as a discourse, again with a reference to Foucault.

How do such organisations interact and form hierarchies and networks?  Assemblage theory uses the same techniques allowing for differences of scale—so that ‘the collective unintended consequences of intentional action becomes more prominent at larger scales’ (75).  The resources required involve specific transactions of exchange, for example and these may be asymmetrical.  Much dependence on the size of the resource and its particular importance—scarcity, the extent to which it can be substituted by another resource, the extent to which it can be controlled and how.  Such control offers a material basis for hierarchy, even if legitimate authority is absent [making linguistic resources secondary again].

Organisations can cope or resist dependency by vertical integration, producing large self sufficient organisations, as in U.S. car manufacturers.  They can also network via interlocking directorships.  Small companies can agglomerate, sometimes by locating in regions where there are accumulations of skilled labour or other resources.  Firms and suppliers can cooperate via consultation.  Often, companies display mixtures, but there are some extreme cases.  Silicon Valley is an example of a region featuring agglomeration, while the Route 128 region in Boston has competing integrated corporations.  Both forms of relationship feature expressive and material qualities.  Again, communication between firms may be ‘not a matter of semantics…  but assessments of strategic significance or importance’ (80) [a further restriction of linguistic to mean semantic—this seems similar to Weber, where purely rational action requires no interpretation and so on.  There may also be a lurking problem with the idea that bureaucracies are actors in their own right—if so, who is the speaker in the expressive activities?  Lower down, page 80, he says that decision-making also raises the problem of just how linguistic expressive behaviour of organisations is.  Here, DeLanda seems to be talking about economic pressure as a form of communication.] Competition and corporation need to be managed in order to avoid threats to common resources—thus a certain amount of solidarity between organisations is important.

Decision-making does involve attitudes and beliefs about the activities of others, but implementing strategies depends on resources, and often requires causal understanding of material processes.  Understanding flows of money can also be important.

Territorialization issues include the role of local geographies in constraining mobility.  Territorialization also has an internal dimension and this depends on the size and autonomy of the firm.  Associations between similar firms in an industry can also be important, as can professional and worker associations.  Turbulent environments can deterritorialize, including rates of innovation, organisational inertia, adaptability, and who is affected.  Contracts can be important, and firms may switch between sales and employment contracts when dealing with other firms. There are also different contractual forms.  Contracts further require the legal system and government organizations.

There are additional problems in applying assemblage theory.  There are many forms of organisations.  It is convenient to focus on the federalist form as in the USA.  As before, reified generalities are not useful, including the term ‘the state’—we need to distinguish between formulation and implementation of policies, for example, and these two wings may have a complex relationship.  Similarly, there is a need to distinguish the democratic and bureaucratic organizations, and a further separation between politics and administration.  Again there are problems of relationships here, sometimes arising from ‘expertise asymmetries’ (86).  Expertise can also shift between bureaucracies and regulators.  There is a further distinction needed between territorial entities and specific hierarchies: changes in government might not affect all the relationships between cities regions and provinces [however it is complex DeLanda tells us since political agencies commonly do administer territories].

It is also necessary to isolate assemblages from  populations or coalitions, on the basis of endurance.  Relationships between organisations need not change the identities involved.  Hierarchical assemblages need an expressive role to legitimate authority and the material role focuses on enforcement.  Voting systems as material components can also have effects of their own, which can include ceremonial rather than technical effects.  Bureaucracies depend instead on technical legitimacy, ‘disinterested expertise’ (89).  There is also a need to regulate experts in the interests of fairness—by appointing key personnel, controlling resources or altering constitutions to include monitoring.  Professionalisation of personnel is an important source of autonomy.

Crises of various kinds can produce deterritorialization, such as coups d’état, or succession crises, or splits within one governing body.  Linguistic resources include the use of unambiguous wordings, written constitutions and laws, and the extent of codification.

Interactions between hierarchical assemblages can be disrupted by a situation such as war.  These require mobilisation of the population, negotiation with critics inside and outside, concessions to citizens.  There is also the need for symbolic expressive activities to increase legitimacy and links civilians and armed forces.  These can vary according to their populism.  Threats of war can territorialize and increase solidarity.  Civil wars and revolutions do the opposite.  A revolutionary assemblage requires a population with rising expectations followed by deprivation; a struggle between dominant groups and challenges; a display of vulnerability by government, as in the inability to deal with a crisis (93, drawing on an obscure source).  Governments, by contrast with populations, tend to be concerned with issues such as sovereignty over particular territories, and this introduces spatial aspects affecting assemblages.

C5 Cities and nations

Some assemblages are bound to a definite spatial location, like cities or nation states.  This helps us understand the role of geographical space, both as constraining and as permitting the expression of important differences, such as the ones between city and country.  [You have to admire DeLanda here, for bringing in all sorts of trendy new links with sociology].

Giddens tries to understand the effects of space, physical territories, but he does overdo the determinism: spaces also need expressive components to territorialize them, and human beings do have some choice in how they respond to spaces.  Nevertheless spaces, and buildings, often survive particular human occupants, and this alone provides constraint in the form of routine.

Buildings constrain interactions through offering different kinds of connectivity.  Technological changes alter connectivity, for example, the design of skyscrapers around metal frames.  [DeLanda says there is an obvious connection between the development of vertical office blocks and stratified bureaucracies, and the height of your office reflected your status in new ways].  Clearly buildings also offer symbolic forms of expression to split sacred and profane, for example [Bourdieu’s work on the Kabylian interior springs to mind here].  Normally, traditions endure, but when fashion develops, frequent change occurs.  DeLanda thinks that fashion is impelled by the need to preserve class distinctions around the aristocracy—quite Bourdieusian (98), with no preservation of open possibilities? New disciplinary regimes have affected distributions of people in particular buildings, as in Foucault on the prison.

Buildings exist in collectivities, larger assemblages, such as neighbourhoods or city zones.  Underground infrastructures also connect buildings.  Exteriors become expressive, and again private forms are driven by ‘ inter-class competition’ (100).  Town planning can attempt deliberate forms of connectivity [as in social mixing in Milton Keynes].  Occupational specialisation can also territorialize, and the opposite for institutionalised segregation.  Here, rates of geographical mobility and economic factors such as land rent [or rateable values in the UK] can have effects [including changes in the use of buildings, as in gentrification], as can increased transport networks which permit the growth of suburbia.  Early urban sociology developed notions of city zones, although the concentric ring model did not fit traditional European cities or changes.  There can be, for example, several city centres or cores.  There have been various influences on the location of retail sectors as well

Discussions of changing land use imply a larger assemblages—populations.  An early difference between town and country was important.  This boundary has changed with population growth, and with patterns of shifting populations, including people who live in both.  Characteristic activities in rural areas require different sorts of resources, such as more land per individual, while cities usually dominate the countryside economically.  Concentration of resources can be responsible for a dynamic of urban growth.  There are also geographical regions, providing a ‘range of objective opportunities and risks’ (105).  The technological characteristics of forms of transport had an effect, for example the building of railroad tracks influencing the development of bead-like suburbs around stations. The decoration or silhouette of buildings can have an expressive effect, as ‘kind of visual signature’ (15), and this is sometimes deliberately modified, for example for tourists.

Population growth and characteristics clearly affect the development of settlements, including the density of cities.  More generally, assemblages like cities display ‘symmetry breaking events, as each town confronts centripetal processes, like the capture of population…  and other resources, as well as centrifugal ones, like congestion, pollution, traffic’ (108).  There may be a single optimal state, but usually rapid sequential change is apparent.

For some geographers, central places are the important elements of urban hierarchies – market places or churches, regional markets based on towns, jails or schools, the more specialist services found in the larger centres.  These hierarchies are based on resource dependencies. Ports and their connections have been particularly important in developing connections between cities of equal status, in a network, and responding to economic changes and specialisation.  Ports often make up for their own resource limits by trading, often over long distances in the only practicable way.  These contacts added to the cosmopolitanism. Expressive elements vary in terms of increasing complexity for land-locked cities as well, until we get to metropolitan centres which are also culturally very diverse.

Territorial states and nations can simply be seen as larger scale assemblages, but they are also integrated differently.  Initially, cities resisted national integration, but some particularly large ones formed the nucleus of the nation and eventually became capital cities.  National integration often involved military intervention as well, and this also has the effect of implying concentration of resources and tight boundaries.  Increased warfare finally ended autonomous city states, and permitted the application of new notions of sovereignty.  The physical location of nation states has always had military implications.

Nation states come to specialise in new activities such as public finance and economic growth, including the development of national debt and credit.  But resources clearly play a part including natural resources, which can also affect the growth of national markets based on transport.  Various kinds of national controls can then be expressed, including the development of national building styles, sometimes in symbolic architecture such as the European ‘Grand Manner’—e.g. ‘uniform facades acting as frames for sweeping vistas culminated with an obelisk, triumphal arch, or statue’ [Think central Paris] (115).  Standard languages are also required, with a consequent national education system. There are physical controls, including objects such as the Maginot line.

Nations are deterritorialized by migration and trade, and internal factors such as remnants of independent city states or colonies.  Maritime networks also resist integration, as in the relative autonomy of Amsterdam.  Such links across borders can produce transnational states, or at least world economies—there is still some doubt about their internal coherence, and this is certainly enough to question Marxist reductionist processes such as world systems theory.  It is necessary instead to see world trade as sets of assemblages

There are no seamless totalities, but assemblages that retain a certain ‘relative autonomy’ and ‘ontological independence’ (119).  This resists reductionism into micro and macro levels.  It also permits different social science contributions to be valued, from understandings of face to face to those of international trade.  The different conceptions are best seen as ‘a chorus that does not harmonise its different components but interlocks them while respecting their heterogeneity’ (119).

back to Deleuze page