Reflexivity as Non-linearity
Scott Lash
U
LRICH BECK'S theory of `reflexive modernization' presumes the
existence of two modernities: a `first' or `simple modernity', and the
modernity at stake in this section of this issue of Theory, Culture &
Society, which is a second, or `reflexive' modernity. I think it makes sense
to understand the first modernity as comprising predominantly a logic of
structures. Then the second or reflexive modernity, if we are to follow
Manuel Castells (1989), involves a logic of flows. Beck's notions of unin-
tended consequences, of ever-incomplete knowledge, of not irrationalism
but a rationality that is forever indeterminate is comfortable in the logic of
flows. Beck's chronic indeterminacy of risk and risk-taking, of living with
risk is much more of a piece with, not the determinacy of structure but the
partial, the elusive determinacy of flow.
So we need to ask ourselves, along with Ulrich Beck, what can reflex-
ivity mean in an age of flows? What can reflexivity mean in the transition
from industrial to the second, informational modernity? Beck's notion of
reflexivity, it seems to me, stands rather in contrast to apparently similar
notions in the work of Jürgen Habermas and Anthony Giddens. This is
because it has a markedly different genealogy. In many respects both
Habermas and Giddens wrote from dissatisfaction with the structural func-
tionalism, the linear systems theory of Talcott Parsons, and the dominance
of Parsonsian sociology in the post-war decades. Both Habermas and
Giddens had affinities with Marxism. Both featured a stress on the import-
ance of agency in contrast to structure. This is perhaps the key to under-
standing reflexivity in the second modernity. It has little to do with the
reflection of structure on agency. It equally has little to do with the (partial)
determinacy of agency by structure. Indeed, it is not about reflection at all.
First modernity reflexivity was a matter of reflection. Indeed Habermas's
communicative action might be paradigmatic, of not reflexive but `reflective
modernization'. Second modernity reflexivity is about the emergent demise
of the distinction between structure and agency altogether. It is not about
Theory, Culture & Society 2003 (SAGE, London, Thousand Oaks and New Delhi),
Vol. 20(2): 49­57
[0263-2764(200304)20:2;49­57;032618]
reflection at all. I also do not think it is predominantly about reflex. Second-
modernity reflexivity presumes a move towards immanence that breaks with
the dualism of structure and agency. Bruno Latour (1993) has wonderfully
captured this immanence in his work. Only for Latour, whose work, for me,
is as much anthropological and philosophical as it is sociological, this is
more or less an ontology. It is the way things are and have always been. It
is the modernity `we never were'. It is pre-modern, modern and contem-
porary. Beck's view and my view tend to be more traditionally sociological:
in the tradition of Marx, Weber, Durkheim and Simmel, who thematized
social change. For us this change from the dualism of reflection to this
immanence or monism of reflexivity is not an ontology, but instead the
defining thematic of the second modernity.
Back to Habermas and Giddens and the idea of reflexivity in terms of
the increasing independence of agency from structure.1 This was the
position from the late 1970s when the two theorists came to hegemonic
positions in German and Anglo-Saxon sociology. Beck comes effectively
from another generation. He came to a position of predominance in the late
1980s in Germany. This was an era that had defined itself against Habermas
and corporatist and Marx-influenced social democracy. Indeed, Beck came
to prominence among a generation, for whom no longer Habermas, but
Niklas Luhmann was the guiding figure. It is significant that Giddens's two
most influential younger colleagues and co-workers have been John
Thompson and David Held, two Habermas-influenced sociologists who have
produced an impressive corpus of work. Beck, in contrast, came to promi-
nence against the grain of Habermas and with Luhmann. Surely, in the early
years of the present millennium, the outbreak of dogmatic Luhmannism in
Germany is something Beck feels distinctly uncomfortable with. Beck is
clearly not a Luhmannianer. Yet a generation of German students have come
to maturity in an ambience that is given shape to by both Beck and
Luhmann, and this ambience is characterized no less by conflict than by
convergence of their respective conceptual frameworks.2
At stake in this, and the defining thematic for me of this article, is a
decidedly non-linear notion of reflexivity. In the first modernity, the
modernity of structure, society is conceived as a linear system. Talcott
Parsons's social system is such a linear system. Linear systems have a single
point of equilibrium, and only external forces can disturb this equilibrium
and lead to system change. The reflexivity of the second modernity presumes
the existence of non-linear systems. Here system dis-equilibrium and
change are produced internally to the system through feedback loops. These
are open systems. Reflexivity now is at the same time system de-
stabilization. Complex systems do not simply reproduce. They change. It is
the `chaos' or noise of the unintended consequences that leads to system
dis-equilibrium. Beck does not use this sort of language, but this sort of
non-linearity is at the heart of reflexivity in the second modernity. It breaks
with the linearity of simple modernity. Indeed, it might be fair to suggest
that Parsonsian systems-linearity was, in many respects, the other side of
50 Theory, Culture & Society 20(2)
Habermas's linearity of agency.3 By the same token, Luhmann's second-
modernity non-linearity of system finds its parallel in Beck's non-linear
blend of agency/structure.
In this sense also, the individual of the first modernity is reflective
while that of the second modernity is reflexive. The idea of reflective belongs
to the philosophy of consciousness of the first modernity. And, to be fair,
Habermas was one of the first to note this. To reflect is to somehow subsume
the object under the subject of knowledge. Reflection presumes apodictic
knowledge and certainty. It presumes a dualism, a scientific attitude in
which the subject is in one realm, the object of knowledge in another. Beck's
work, from the very start, has presupposed a critique of such objectivist
knowledge, a critique of such dualisms, be they Cartesian or Kantian. Beck's
(1974) very first book, well before Risk Society, stemming from his doctoral
work, addressed issues of knowledge and science. This work was already
implicitly phenomenological, breaking with assumptions of the sub-
ject­object epistemology of the Enlightenment and positivism. For Beck, as
for phenomenology, the knowing individual was already in the world with
the objects of his/her knowledge. This subject could only grasp a certain
portion of the object, in connection with what Husserl called the subject's
`attitude', in Beck's case the interest-constituted attitude of the knower.
Thus the objectivity of simple-modernity knowledge is replaced by the
intentionality of knowledge in the second modernity. This intentionality is
again at centre stage in Risk Society, now tied up with the ecological prob-
lématique. Science and industry, for all their claims to objectivity, and to
being somehow objective and outside of the world, are indeed in the world
with their own proper interest-constituted intentionality. The problem here,
although it is at the same time its saving grace, is that what is intended
leads to the most extraordinary unintendedness, to side-effects, to unin-
tended consequences.
The Cartesian subject of simple modernity, of Descartes' Metaphysi-
cal Meditations, is reflective. So is the Kantian subject of determinate
judgement. Beck often describes today's non-linear reflexivity in terms of,
not the `I think therefore I am', but instead in terms of `I am I'. `I think,
therefore I am' has to do with reflection. `I am I' has more to do with reflex.
And Beck often indeed works from the contrast of `reflex' with reflection.
Reflexive, he argues, has more to do with reflex than reflection. Reflexes
are indeterminate. They are immediate. They do not in any sense subsume.
Reflexivity, Beck notes, is characterized by choice, where previous gener-
ations had no such choices. What Beck often omits to say is that this choice
must be fast, we must ­ as in a reflex ­ make quick decisions. In the second
modernity we haven't sufficient reflective distance on ourselves to construct
linear and narrative biographies. We must be content, as Ronald Hitzler
(1988) has noted, with Bastelbiographen, with bricolage-biographies in
Lévi-Strauss's sense. We may wish to be reflective but we have neither the
time nor the space to reflect. We are instead combinards. We put together
networks, construct alliances, make deals. We must live, are forced to live
Lash ­ Reflexivity as Non-linearity 51
in an atmosphere of risk in which knowledge and life-chances are
precarious.
So what is at stake? The second modernity and its non-linear reflex-
ivity are a result of the retreat of the classic institutions: state, class, nuclear
family, ethnic group. The roles that reproduced linear agents and systems
in the first modernity are progressively erased. Yet the result is not the
disappearance of the subject, or a general irrationality. The subject relating
to today's fragmented institutions instead has moved from a position of
reflection to one of being reflexive. Yet this subject is so constantly in motion
and so involved in the world that it makes little sense to talk about a subject-
position. The subject is still with us and so is knowledge. Only knowledge
itself is of uncertainty. What happens now is not non-knowledge or anti-
reason. Indeed, in reflexive modernity we are better educated, more know-
ledgeable than ever. Instead, the type of knowledge at stake changes. It is
itself precarious as distinct from certain, and what that knowledge is about
is also uncertain ­ probabilistic, at best; more likely `possibilistic'.4
Beck has written extensively about globalization, about cosmo-
politanism, in the years since the publication of Risk Society (1992) and
Reflexive Modernization (Beck et al., 1994). Indeed, cosmopolitanism itself
presumes a certain movement of strategic locus both extensively and inten-
sively away from the nation-state. This intensive shift will have implications
for the family. What happens more generally is, to a certain extent, a
generalized outsourcing. In Beck's Risk Society or what John Urry and I
(1987) at about the same time called `disorganized capitalism', there is a
generalized outsourcing of functions, of operations. The hierarchical
economic organization begins regularly to make decisions, not to `make' but
to `buy'. A whole host of functions of the firm are outsourced in this age of
vertical disintegration. The welfare state begins to outsource functions to
private and charitable sector organizations. There is, it seems, also an
outsourcing of the family. This is not just a disorganization and destruction
of the family. It is, on the one hand, a distantiation: hence the increasing
gift of mobile phones from parents to pre-teen and teenage children. This
is so parents can keep track of children at a distance. Children, for their
part, tend to use these same phones for other objectives. Consider the
fragmentation of the family through divorce or separation, or even the long-
distance couples connecting London and Los Angeles, Paris and New York.
Here divorce and separation can mean that children have, not no family but
two or three families, to whom they are connected by long-distance tele-
phony, transport and the Internet. Thus there is, not so much the destruc-
tion, as an outsourcing of the family. At stake then is first an anomic
disorganization but then a new normalization, which again is a normaliza-
tion that institutionalizes abnormality, institutionalizes not the normal but
the state of exception (as Carl Schmitt might have it). It is, in a sense, a
routinization of Weberian charisma that winds up not as bureaucracy but
instead as somewhat more regularizable charisma.
What is happening, however, is not just an outsourcing but also an
52 Theory, Culture & Society 20(2)
insourcing. Giddens (1993) of course has written extensively about this. A
number of properties, functions and activities previously attributable to the
nation-state, the welfare state, the hierarchical firm, the family, the central-
ized trade union have been otherwise located. Some of them have been
extensively displaced onto global instances, while others have been inten-
sively displaced to more private instances. The shift of activities onto small
firms has been such an intensification. Today's start-ups ­ not so much the
dot.coms, but patent-generating technology firms and copyright-generating
new media firms ­ have very private, personal and intense characteristics.
They are not, so to speak, paternalistic as they were in the bygone days, not
least because women now run a number of them. It has to do with the fact
that so many of the employees are freelancers and subcontractors. It has to
do with the eminently non-paternalist charisma of firm leaders. These are
not, so to speak, `leaders of men', but may instead be risk takers.
So there is in our times an outsourcing of governance functions of the
state, of national rights to become global rights, of accountancy organiz-
ations, of economic functions onto supra-national economic bodies, and
supra-national cultural instances like the world-wide spread of biennales.
There is just as much an offloading of functions onto private instances.
These are the two constitutive features of the second modernity. In the first
modernity the (reflective) subject was constituted in consonance with a set
of roles in a variety of institutions. Now these institutions are in crisis, and
functions which were once taking place at the interface of institution and
role are now taking place much more intensively and closer to the subject.
What has happened is that there has been a de-normalization of roles. The
subject has become, and Beck too uses this word, `nomadic'. There has been
a move toward complexity, indeed towards `chaos'. But it is somehow a regu-
larizable chaos. The `roles' of the first modernity depended very much on
what Kant called determinate judgement, on prescription, on determinate
rules. Now the subject must be much more the rule-finder him- or herself.
Determinate judgement is replaced by `reflective judgement'. Reflective
judgement is not reflection because there is now no universal to subsume
the particular. In reflective judgement it is we who must find the rule.
Reflective judgement is always a question of uncertainty, of risk: it also
leaves the door open much more to innovation. Thus Ulrich Beck and
Elisabeth Beck-Gernsheim's Das ganz normale Chaos der Liebe (1990) has
appeared perhaps misleadingly in English as The Chaos of Love. The
German title translates literally as The Totally Normal Chaos of Love. Love
here becomes dissociated from roles and hence chaotic, while this chaos
becomes totally normal, becomes regularized in a fashion. Becomes, even,
more or less predictable. Yet at the same time it remains nonetheless chaos.
Chaotic love, regularizable chaotic love is non-linear love and part and
parcel of non-linear reflexivity.
This all leads to the question of institutions. Today's debates around
globalization and cosmopolitanism have led to a considerable literature on
the extensive outsourcing of, in this case, sovereignty. The pursuing, for
Lash ­ Reflexivity as Non-linearity 53
example, of governance as discussed by David Held, of economic functions,
as discussed by Saskia Sassen, of rights (Homi Bhabha). But what about
institutions that regulate the above-mentioned insourcing of functions?
What kind of institutions can regulate what Beck since the early 1980s has
understood under the heading of `a life of one's own' (Eigenes Leben) (Beck
et al., 1997)? What kind of institutions can regulate subjects whose differ-
entia specifica is precisely not to be determined by the rules of institutions?
What institutions can enable us to be reflexive in the sense of being rule-
finders? At issue here are subjects that are not so much anomic as auto-
nomic. This may be the pivotal contradiction as it were of reflexive
modernization. Governance of second-modernity flows is always going to be
very different from governance of first-modernity structures. Perhaps at
stake is a question of institutions so different that for us they are almost
unrecognizable as institutions. It may make sense in this context to think
in terms of two types of institutions: of institutions that proffer us two types
of rules. Let us go back to the distinction between constitutive and regula-
tive rules. Constitutive rules are those that let us play the game, much akin
to constitutional law. Without them there is no playing field. Regulative
rules are more prescriptive. What kind of constitutive rules, we may ask,
are consistent with a set of regulative rules that do not regulate? What type
of constitutive rule is consistent with a set of rule-finding, as distinct from
rule-determining, activities? Perhaps new second-modernity institutions
must be comprised primarily of, not regulative but constitutive rules. And
in this sense they may not be recognizable to us as institutions.
It has become commonplace to say that in the global information age,
in the second modernity, power and inequality operate less through exploi-
tation than exclusion. Beck is very cognizant of this with his notion of
`Brazilianization'. The point I am trying to make here is that exploitation
takes place through regulative rules, while exclusion take place through
constitutive rules. This is consistent with Hardt and Negri's (2000) argument
in Empire that the transition from one mode of production to another is
governed less by class struggle than by `lines of flight'. Here the subordi-
nate class escapes literally as flight, flow or flux away from the dominant
class and its institutions. So perhaps the key institutions at stake in the
second modernity are those that govern exclusion. And here is where I have
my strongest quibble with Beck's work. I think that a great number of these
characteristically second-modernity institutions, if that is the word for them,
are now not solely social, but socio-technical. Now this is completely
consistent with the technologization of science thesis that has been so
prominent in Beck's work. Pivotal for me among the socio-technical (consti-
tutive more than regulatory) institutions that govern contemporary power
relations are the likes of platforms, operating systems, communications
protocols, standards, intellectual property and the like. There is a certain
awareness of such socio-technical institutions also in Elisabeth Beck-
Gernsheim's (Beck and Beck-Gernsheim, 2002) work on genetic databases.
But I do not think that the technological dimension is sufficiently taken on
54 Theory, Culture & Society 20(2)
by (Ulrich) Beck. Nor are the dimension and extent to which social relations
are mediated through the (now interactive) mass and non-mass media of
communications. Reflexive modernization, the Becks argue, is a question of
`place-polygamy'. The point is that such place-polygamy is always necess-
arily technologically mediated, by cheaper air flights, by mobile phones, by
microprocessors in various smart boxes, by protocols and channels enabling
communication at a distance between individuals.
This brings me back to the centrality of non-linearity in Beck's notion
of reflexivity. The first modernity was linear, the second non-linear. The first
modernity a question of determinate judgement and rule-following, the
second a matter of rule-finding and reflective judgement. There is, as I
argued above, a very loose and implicit notion of complex (open in both
senses) and non-linear systems in Beck's work. But these are always social
systems. The point I want to make is that there has been a shift here, which
again is implicitly addressed in Beck's work. In the first modernity we were
faced with relatively mutually exclusive and exhaustive systems: of (Parson-
sian) social systems on the one hand, and engineering-like technical
systems on the other. The second modernity's totally normal chaos is regu-
lated by non-linear systems. It is also regulated by an extraordinarily
powerful interlacing of social and technical systems: by, precisely, socio-
technical systems. It is at the interface of the social and the technical that
we find the second modernity's reflexivity. It is at this interface that we take
on the precarious freedom of a `life of our own'; that we `invent the political',
that we take on ecological responsibility. Reflexivity in the second
modernity is profoundly socio-technical.
This very `Latourian' point leads me back to considerations of Bruno
Latour's contribution to this Theory, Culture & Society `round table' on reflex-
ive modernization. Very often Beck and Latour have been juxtaposed to each
other as realism versus constructionism, with Beck the realist and Latour
the constructionist. Latour's otherwise compelling contribution to this issue
does have some constructionist echoes. But I think this is more or less just
an inconsistency that emerges from time to time in his work. This is because,
for me, the distinctive feature of the `turn' in science studies from the Strong
Programme takes place indeed with Latour's break with constructionism. The
Strong Programme has been a thoroughly constructivist sociology of know-
ledge, in which knowledge is constituted through what is effectively a
positing (individual or collective) subject. This was a welcome departure
from the nave subject­object realism that was previously dominant in soci-
ology. It was a departure, of course, from previous (Mertonian and indeed
Popperian) positivism in a hermeneutic or phenomenological direction. The
problem however was that constructionism tended to repeat the
subject­object dualism of realism in another register, in fact giving more
power to the now-constituting subject than even the previous Cartesian
assumptions. Latour's departure and the subsequent paradigm-shift in the
sociology of science were in his powerful `objectualism', which is not objec-
tivism. In this objectualism previously dualist objects and subjects become
Lash ­ Reflexivity as Non-linearity 55
quasi-objects and quasi-subjects in a monist plane of actor-networks. Latour
is at his worst when constructionist. He is at his best when his immanen-
tism breaks with dualisms of both constructionism and realism.
Beck, for his part, oscillates between realism and immanentism.
Beck's non-linearity comes from implicit Luhmannian assumptions of self-
reproducing, environment-scanning systems. Latour's actor-networks are
situated in a problématique of flows and `stoppages', in Duchamp's sense, of
`lines of flight' and `plateaus'. He doesn't use the language of systems, but
surely actor-networks can be more or less systemic. Latour is also not unin-
fluenced by information theory, in his case coming not through Luhmann
but instead Michel Serres. In Another Modernity, A Different Rationality
(1999), I criticized Latour for assumptions of instrumental rationality. This
was wide of the mark. The driving logic of actor-network immanence is of
course to break with dualisms of instrumental and substantive rationality,
of exchange-value and use-value. Let me revise this again. In his moments
of backsliding towards constructionism, Latour's actor-networks do take on
instrumentally rational colours. Equally, when Beck backslides into realism
he once again moves back into the first modernity's problématique of reflec-
tion. All this said, the main thrust of Beck's and Latour's work does seem
to converge in this pox-on-houses of both realism and constructionism. For
both Beck and Latour are situated in registers of such monism and non-
linearity. And this is what the second, reflexive modernity is all about.
Notes
1. To be fair there is a dimension of this idea of reflexivity in both Beck's work up
until the mid-1990s and indeed in my section of Reflexive Modernization (Beck et
al., 1994). That is the wonderful thing about the notion of reflexivity and reflexive
modernization. I develop the notion again beginning from Garfinkel in the chapter
on `Technological Phenomenology' in Critique of Information (Lash, 2002). What
that is leading to, and is perfectly consistent with, is the idea of non-linear reflex-
ivity addressed here.
2. This article draws on a number of long discussions with Jakob Arnoldi and June
Hee Jung. I am grateful to them for a number of points here. The judgements,
however, are my own, and they should not be held responsible for them.
3. To be fair to Giddens, his notion of agency has important dimensions of non-
linearity, especially with the centrality of unintended consequences and the indi-
vidual as experiment in his work.
4. This position is to my knowledge most fully developed by Jakob Arnoldi in his
PhD thesis on uncertain knowledge, University of London, 2002. I am indebted to
Arnoldi for these ideas.
References
Beck, U. (1974) Objectivität und Normativität ­ Die Theorie-Praxis-Debatte in der
modernen deutschen und amerikanischen Soziologie. Reinbek: Rowohlt.
Beck, U. (1992) Risk Society. London: Sage.
Beck, U. and E. Beck-Gernsheim (1990) Das ganz normale Chaos der Liebe. Frank-
furt: Suhrkamp.
56 Theory, Culture & Society 20(2)
Beck, U. and E. Beck-Gernsheim (2002) Individualization. London: Sage.
Beck, U., A. Giddens and S. Lash (1994) Reflexive Modernization. Cambridge:
Polity.
Beck, U. et al. (1997) Eigenes Leben. Munich: C.H. Beck.
Castells, M. (1989) The Informational City. Oxford: Blackwell.
Giddens, A. (1993) The Transformation of Intimacy. Cambridge: Polity.
Hardt, M. and A. Negri (2000) Empire. Cambridge, MA: Harvard University Press.
Hitzler, R. (1988) Sinnwelten: Ein Beitrag zum Verstehen von Kultur. Wiesbaden:
Westdeutscher Verlag.
Lash, S. (1999) Another Modernity, A Different Rationality. Oxford: Blackwell.
Lash, S. (2002) Critique of Information. London: Sage.
Lash, S. and J. Urry (1987) The End of Organized Capitalism. Cambridge: Polity.
Latour, B. (1993) We Have Never Been Modern. London: Longman.
Scott Lash is Professor of Sociology and Director of the Centre for Cultural
Studies at Goldsmiths College, University of London. His recent books
include Another Modernity, A Different Rationality (1999), Critique of
Information (2002) and Recognition and Difference (co-ed. with Mike
Featherstone).
Lash ­ Reflexivity as Non-linearity 57