The current issue and full text archive of this journal is available at

www.emeraldinsight.com/0368-492X.htm

Second-order
The biological metaphor of a observer
second-order observer and the
sociological discourse
531
Loet Leydesdorff
School of Economics (HEC), Université de Lausanne, Lausanne, Switzerland
Amsterdam School of Communications Research (ASCoR),
University of Amsterdam, Amsterdam, The Netherlands

Abstract
Purpose – In the tradition of Spencer Brown’s Laws of Form, observation was defined in Luhmann’s
social systems theory as the designation of a distinction. In the sociological design, however, the
designation specifies only a category for the observation. The distinction between observation and
expectation enables the sociologist to appreciate the processing of meaning in social systems. Seeks to
address this issue.
Design/methodology/approach – The specification of “the observer” in the tradition of systems
theory is analyzed in historical detail. Inconsistencies and differences in perspectives are explicated,
and the specificity of human language is further specified. The processing of meaning in social
systems adds another layer to the communication.
Findings – Reflexivity about the different perspectives of participant observers and an external
observer is fundamental to the sociological discourse. The ranges of possible observations from
different perspectives can be considered as second-order observations or, equivalently, as the
specification of an uncertainty in the observations. This specification of an uncertainty provides an
expectation. The expectation can be provided with (one or more) values by observations. The
significance of observations can be tested when the expectations are properly specified.
Originality/value – The expectations (second-order observations) are structured and therefore
systemic attributes to the discourse. However, the metaphor of a (meta-)biological observer has
disturbed the translation of social systems theory into sociological discourse. Different discourses
specify other expectations about possible observations. By specifying second-order observations as
expectations, social systems theory and sociocybernetics can combine the constructivist with an
empirical approach.
Keywords Sociocybernetics, Social systems, Communication, Expectancy theory
Paper type Conceptual paper

1. Introduction
Sociocybernetics and systems theory have often been dismissed by sociologists with
the argument that the prevailing metaphor in these traditions has remained
meta-biological (Giddens, 1984; Grathoff, 1978; Habermas, 1987; Habermas and
Luhmann, 1971). In an important collection of articles on the subject entitled
Sociocybernetic Paradoxes and edited by Felix Geyer and Hans van der Zouwen,
Luhmann (1986, p. 172) addressed this problem when he formulated that “because it is
tied to life as a mode of self-reproduction of autopoietic systems, the theory of
autopoiesis does not really attain the level of general systems theory which includes Kybernetes
Vol. 35 No. 3/4, 2006
brains and machines, psychic systems and social systems, societies and short pp. 531-546
interactions.” He proposed considering the processing of meaning as another, that is, q Emerald Group Publishing Limited
0368-492X
non-biological, form of autopoietic organization, which he formulated as follows: DOI 10.1108/03684920610653791
K The concept of autopoietic closure itself requires this theoretical decision, and leads to a sharp
distinction between meaning and life as different forms of autopoietic organization; and
35,3/4 meaning-using systems again have to be distinguished according to whether they use
consciousness or communication as modes of meaning-based reproduction. (. . .) The general
theory (of autopoiesis), however, is meaningful only if this implementation succeeds, because
otherwise we would be unable to determine which kinds of attributes are really general
(Luhmann, 1986, p. 173).
532
It will be argued in this contribution that Luhmann was successful in achieving this
“implementation.” He elaborated how meaning-processing systems are different from
living systems. However, in using “observation” without sufficient reflection on the
status of this concept in sociological discourse, Luhmann’s theory of social systems
remained vulnerable to the epistemological critique of using a meta-biological
metaphor (Habermas, 1987; Leydesdorff, 2000).
The clarification of this issue is urgent because the confusion has been used as an
argument against sociocybernetic approaches and in favor of action theory and its
derivatives in sociology (Habermas, 1981; Giddens, 1984; Münch, 1982/1988; Beck et al.,
2003). While observation as an operation can be considered as action, the reflexive
entertaining of expectations is based on perceptions and experience. The introduction
of a “second-order observer” at the level of social systems has not solved the issue, but
led to philosophical discussions that distract from an empirical orientation in research
(Luhmann et al., 1990; Baecker, 1999; Fuchs, 2004). The systems-theoretical approach
remains confusing in sociological discourse when “observation” is not provided with a
sufficiently precise meaning.

2. Epistemological considerations
“Observation” can – for analytical reasons – be attributed to an “observing system”
(Von Foerster, 1982). In an article entitled “Cybernetics of Cybernetics,” Von Foerster
(1979) proposed distinguishing first-order cybernetics as the cybernetics of observed
systems (e.g. rockets) from second-order cybernetics as the cybernetics of observing
systems (Baecker, 1996; Glanville, 2002). This formulation can be considered as the
culmination of Von Foerster’s work in the Biological Computer Laboratory at the
University of Illinois, of which he had been the director for many years (Von Foerster,
1975; Varela and Goguen, 1978).
Shortly before this date, Maturana (1978) had proposed defining an observer in the
context of developing autopoietic systems theory. Note that Maturana explained the
generation of an observer operationally as a consequence of the biological process of
autopoiesis in living systems and not – like Von Foerster – as an epistemological
assumption underlying observation as an analytical category. Thus, this turn leaves
Kant’s (1956) epistemological concept of the observer as a transcendental subject more
definitively behind because the observer is now grounded in the biological domain.
Consequently, the focus is on the biological constraints and dynamics of an observer
(Edelman, 1989). Maturana (1978, p. 49) formulated his definition of an observer as
follows:
The magnitude of this recursive ontogenic structural coupling in any particular organism
depends both on the degree of structural plasticity of its nervous system and on the degree to
which the actual structure of its nervous system at any instant permits the occurrence of
distinct relations of relative neuronal activity that operate as internal structural
 perturbations. When this takes place, even in the slightest manner, within the confines of a Second-order
consensual domain, so that the relations of neuronal activity generated under consensual
behavior become perturbations and components for further consensual behavior, an observer observer
is operationally generated.
In other words: two variations can disturb each other as shown in Figure 1. When the
disturbance is structural and thus repeated over time, the interface can develop into a
third system (e.g. a synapse). In other words, the uncertainty in the overlap can first be 533
considered as a transmission between two variations. This co-variation can develop
into a co-evolution over time. This third dimension was considered by Varela (1975,
p. 7) as the emergence of an autonomous state which can be distinguished from the two
states that went into the process.
The mechanism of how noise can become a signal in a selection environment can be
specified by using Arthur’s (1989, 1994) model for “lock-in.” A threshold can be passed
in a random walk by chance and thereafter the process is self-reinforcing if the
inscription of the signal leaves traces that can be further developed (Leydesdorff,
2001a). This process of lock-in and inscription occurs on both sides of the co-variation,
a co-evolution can be developed as in a process of “mutual shaping” (McLuhan, 1964).
This evolutionary perspective, however, is not addressed in Spencer Brown’s (1969)
Laws of Form. Varela (1975) elaborates on Spencer Brown’s focus on two states: the
“marked” and the “unmarked” ones. In a much later discussion, Spencer Brown (1994,
p. 51) added the possibility of an evolutionary process, but he formulated this
mathematically as only an oscillation between the states:
Similarly, when we get eventually to the creation of time, time is what there would be if there
could be an oscillation between states.
The conditional is used by the author because one needs the additional assumption of,
for example, evolution theory that an oscillation between two subsystems can be
expected over time. The recursive operation (i.e. the clock of the system), however, is
created on grounds external to the Logic of Forms (Spencer Brown, 1969), notably on
the basis of substantive theorizing (Varela, 1975, p. 20). The evolutionary theorist is not
interested in the momentary appearance of an observer, but in its existence over time in
terms of specific operations (Günther, 1967; Varela and Goguen, 1978).
Maturana (1978) noted that the behavior of an “observer” over time cannot be
distinguished from the establishment of a semantic domain. He formulated this
conclusion as follows:

Figure 1.
Two variations disturbing
each other generate the
possibility of selection of a
signal by “lock-in” when
the disturbance is repeated
over time
K In still other words, if an organism is observed in its operation within a second-order
consensual domain, it appears to the observer as if its nervous system interacted with internal
35,3/4 representations of the circumstances of its interactions, and as if the changes of state of the
organism were determined by the semantic value of these representations. Yet all that takes
place in the operation of the nervous system is the structure-determined dynamics of
changing relations of relative neuronal activity proper to a closed neuronal network
(Maturana, 1978, p. 49).
534
Thus, Maturana’s observer remains completely endogenous to the system(s) under
observation. The behavior of this system can be described by an external
super-observer in terms of an observer emerging within the system. Alternatively,
one can describe this same effect as a semantic domain. The two descriptions are
identical in the sense that they cannot be operationally distinguished at the level of the
system itself. The emergence of a semantic domain is shown in Figure 2. The overlap
can generate a third domain within the system when the interaction is sustained by the
two interacting systems over the time axis. The variations select upon each other
mutually and thus are able to select certain variants for stabilization over time from the
previous selections.
The mutually selected variants can be considered as providing a second-order
variation. I have added to this picture in Figure 2, the next step that Maturana also
took, namely the development of a language as the second-order operation of the
semantic domain. This selection on the second-order variation can also be considered
as a second-order selection or feedback. According to Maturana, “denotation” is the
necessary requirement for the linguistic exchange. Denotation arises “only in a
metadomain as an a posteriori commentary made by the observer about the
consequences of operation of the interacting system” (Maturana, 1978, p. 50).
In other words, language can be considered as an operation within the semantic
domain that enables a first-order observer to enter into symbolic communication (e.g.
by using words). The denoted entities can then be “discussed” regardless of whether
these entities still exist in the domain from which they originated. While the

Figure 2.
The structural coupling of
the disturbance terms
develops a semantic
domain, that is, a
first-order observer
coevolution adds time as a third degree of freedom to the two interacting variations, Second-order
this next-order globalization requires one more degree of freedom in a system. A observer
self-organizing system, therefore, can be modeled in a four-dimensional (hyper)space
(Leydesdorff, 1994, 2001b).
One should keep in mind that “language” is here defined in terms of a system’s
operation and is not yet identified with “human language”. Maturana’s “biological
language” is used, for example, among molecules that are structurally coupled in an 535
autopoietic process of self-reproduction. Maturana carefully distinguished this
“language” from the language used by biologists as supra-observers who entertain a
biological discourse. In the biological case these two levels can clearly be distinguished.
Sociology, however, has a more difficult task in this respect.
Maturana (1978, p. 49) drew a radically constructivist conclusion from this
specification of the biological discourse as a meta-language by stating that “one cannot
speak about whatever one cannot talk.” The human cognitive domain and its
languaging are operationally closed (Maturana, 2000). However, it follows from this
generalized definition of language as an operation that the domain of human language
can only be generated within one semantic domain or another. The linguistic domain
adds a reflexive layer of meaning processing to the semantic layer. The latter layer
processes information as noise insofar as this information is not yet reflected and
provided with meaning by its second-order domain.
Let me anticipate my conclusion that human language can be considered as specific
in achieving one degree of freedom more than biological languages such as exchanges
among molecules or animals. This degree of freedom enables us to develop a structural
coupling between the (Shannon-type) information exchanges and the meaning
exchanges. For example, some information can be provided with meaning, while the
remainder is considered as noise. The same events can be provided with different
meanings, and these meanings can again be communicated. The communication can
again be noisy.
In the case of biological languages, the linguistic domain cannot further be
developed as a communication system that feeds back as another operation on its own
semantic domain. The “meanings” of the biological communications are provided in a
linguistic domain, but these meanings changes among, for example, the cells according
to the biology of the system. The life-cycles of the biological systems thus fix the
meanings at each moment of time. These exchange processes are then changed by
wear and tear. One would need a super-observer to reconstruct the biological domain
as in molecular biology or biotechnology. However, the super-observer uses an angle of
reflection different from that of the biological system under study. The neuronal
system operates naturally within its own domain and generates a language that
remains embedded (since naturalistically given).
In other words, a biological system has to adapt – at the risk of extinction –
because the selection pressures are “natural.” Selection environments, however, may
already vary among ecological niches. Biological exchange relations can therefore
experience additional selection pressure. However, the two types of selection pressures
cannot be distinguished by the agents living in these niches unless these carrying
agents would have the reflexive capacity to entertain concepts for mapping their
situation. When these concepts can be communicated as denotations, a linguistic
domain is generated operationally on top of the semantic domain. The system of
K communications can then become self-organizing and autopoietic in the sense of
35,3/4 increasingly controlling and fine-tuning its own reproduction. This happens in
biological cultures like insect populations.
Human language extends the biological concept of a linguistic domain because
order is not constructed in a biological environment and then stabilized, but remains
flexible and under construction as an expectation of order by using language and
536 next-order (e.g. symbolically generalized) media of communication. The constructed
order can be changed by a next-order system or at a next moment in time, i.e. by adding
reflexively a new dimension to the system. Note that “reproduction” in this case no
longer means reproduction in the biological sense, but rather the further development
of the communication through the system’s previously organized retention
mechanisms. The lower-level systems upon which the next-level systems build can
be innovatively reconstructed given the additional degree of freedom provided by
linguistic exchanges. The represented and the representing systems may then begin to
co-evolve. This co-evolution can be sustained by codes at the symbolic level.

3. The double hermeneutics in inter-human communication

In human interactions people are able to act as participant-observers and/or as external
observers (Giddens, 1976). These two roles span other domains and roles for observers.
However, the distinction between these two roles is analytical. In inter-human
communication, this degree of freedom can be handled reflexively. The codification of
this reflexivity spans the domain of sociological discourse (Giddens, 1979, 1984).
In other words, the “emic” perspective can be distinguished from the “etic”
perspective by using a reflexive discourse (Geertz, 1973). One might even say that the
sociological analysis emerges from reflexivity about this distinction. Under what
conditions and precisely how can an analyst add knowledge to the knowledge of the
participants? Can this knowledge be codified in a discourse which is different from and
reflexive to the discourse of the subjects under study?
By focusing mainly on the meaning processing of the social system from an “etic”
perspective, Luhmann (1984, 1997a) could make a number of important steps forward.
For example:
.
The possibility of functional differentiation of the communication. Since, the
communication system is considered as a system different from the carrying
action systems (which Luhmann prefered to call “consciousness systems” in
order to stress the priority of meaning processing; cf. Wasser, 2001), the
differentiation can be expected to carry functions for the social system that differ
from those functional for action. Thus, the four-function paradigm developed by
Parsons (1937, 1951) for social action systems could be abandoned. The social
system may use another alphabet of functions for the communication (Simon,
1969).
.
These functions are carried by the symbolically generalized media of
communication. These media operate by codifying the communication with
potentially different meanings. Thus, using natural languages can be considered
as a “primitive” (since natural) form of communication, notably, one that is not
yet codified by using symbolic media. Prices, for example, enable us to speed up
the economic process with an order of magnitude on compared with bargaining
at a market.
 .
Consequently, communication and language can be further distinguished. Second-order
Within scientific discourses, for example, communication is governed by code observer
which determines whether or not the communication can be considered as “true.”
The code of communication closes the specificity of a scientific paradigm and
thus a next-order jargon can be developed. This distinction between codified
communication and language enables us to explain, for example, why and when
a difference between “restricted discourse” and “elaborate discourse” – as 537
distinguished in discourse analysis (Coser, 1975; Bernstein, 1971) – can be
expected (Blauwhof, 1995; Leydesdorff, 2002).
.
In addition to the functional differentiation, Luhmann (1975, 2000) suggested a
systemic differentiation in social communication between the levels of
interaction, organization, and society. While the first two levels can be
considered as specific formats for the integration, the level of society can evolve
into a dynamics of functional differentiation and the self-organization of
meaning processing. The dynamics between the functional and the institutional
layer drive the system potentially in the higher gear of a knowledge-based
system (Leydesdorff, 2001b, 2003).

Let us now return to the question of the status of the observer and his/her observation.
Luhmann (1990, 1993) followed Von Foerster in generalizing the concept of
“observation” to the operation of a second-order system. As noted, the attribution of
the observation to an observer was made by Von Foerster for the purpose of extending
evolutionary theorizing in biology to meta-biological systems. However, Luhmann
attempted to generalize this attribution also in terms of mathematics by claiming
support from Spencer Brown’s (1969) Logic of Forms.
Spencer Brown formulated on the last page of this study as follows:
An observer, since he distinguishes the spaces he occupies, is also a mark (. . .) In this
conception a distinction drawn in any space is a mark distinguishing the space. Equally and
conversely, any mark in a space draws a distinction.
We see now that the first distinction, the mark, and the observer are not only interchangeable,
but, in the form, identical (Spencer Brown, 1969, p. 76).
In an extensive note Spencer Brown (1969, p. 84) explained that the observer in this
case does not have to be a human being. It can also be an animal. Thus, Spencer Brown
was not yet specifying at the level of social systems theory, and not even at the level of
biological evolution theory. The observer remained unspecified in terms of its
operation because a “distinction” is drawn at a specific moment in time.
As noted, Varela (1975) extended Spencer Brown’s logic to a calculus for
self-referential systems by specifying the notion of identification of the distinction from
a biological perspective. He speculated (p. 22) that this solution at the biological level
might provide a sound basis for a general theory of self-referential systems. Luhmann
(1993), however, defined the observer a priori at the level of general systems theory.
This has led to confusion between the definitions in terms of general systems theory
and the specific distinctions and designations possible at the level of social systems.
Varela (1975; cf. Goguen and Varela, 1979; Varela and Goguen, 1978) elaborated on
Spencer Brown’s definitions in order to develop a calculus of self-reference. These
authors argued that an “indication” adds to a “distinction” by marking one of the two
K distinguished states as primary; for example: “this,” “I,” “us,”, etc. The identification
35,3/4 can be considered as the very purpose of the distinction. However, the (mathematical)
distinction provides a necessary, but not a sufficient condition for an indication
(Spencer Brown, 1969, p. 1)[1]. Indication requires the translation of the mathematical
generality (of a distinction) into the contingent domain of substantive theorizing (for
the identification). The identification completes the generation of an observer by using
538 a second dimension. This accords with Maturana’s (1978) definition of the
(momentary!) generation of an observer. As noted by Varela (1975), the extension of
the system with a third (temporal) dimension makes an autonomous system possible.
On the basis of this consensus Maturana and Varela (1980) could fully develop the
biological theory of autopoiesis into an epistemology (Luhmann, 1990, p. 81). Spencer
Brown’s observer is considered only a distinction, i.e. an instance (“mark”) at one
specific moment in time. It remains part of a logic or an arithmetic. An arithmetic was
defined by Brown as “a calculus in which the constants operated on all have specific
values” (Varela and Goguen, 1978, p. 296). Maturana and Varela’s observer, however, is
part of an evolutionary process, i.e. it can be expressed as an algorithmic operation.
The value of the variables can then change over time, while the variables remain
identifiable (dx/dt). Because of the addition of selection over time, the variation is no
longer subjected to a single selection mechanism operating at each specific moment in
time. A relative stabilization of two selection mechanisms operating upon each other as
mutual disturbances becomes possible. It has been argued above that some selections
can then be selected for stabilization as an observer.
Note that Maturana and Varela’s observers are defined as first-order observers
because they are entrained in a life-cycle. The “observing systems” of Von Foerster,
however, could be second-order observers which are not necessarily alife. (As noted
above, Maturana distinguished additionally between a first-order semantic and a
second-order linguistic domain). Luhmann (1993) denied this distinction at the
epistemological level because he wished to ground social systems theory as a
substantive theory like biology in “observation” as an abstract operation defined at the
level of general systems theory:
The second-order observer, mind you, is a first-order observer as well, for he must distinguish
and designate the observer he intends to observe (Luhmann, 1993, p. 20).
This deliberate blurring of the levels is fully consistent with Luhmann’s proposal to
consider the “observer” as a universal category that would provide a fundament to
both second-order systems theory and the theory of social systems. For example,
Luhmann (1993) could therefore formulate that “on the level of second-order
observation it is then possible to see and state that . . . ” However, communication
systems do not have eyes for seeing and mouths for stating. Luhmann’s texts tend to
oscillate between first-order and second-order observers and this confusion generates
what he indicates as a paradox[2].
It is argued here that first-order and second-order observing systems operate
differently, and therefore can be expected to entertain substantively different
perspectives. “Observing systems” do not have to be identified as “observers” in a
biological (or psychological) sense. The only requirement is that observing systems
have to be able to carry “distinctions and designations.” Systems other than
identifiable “observers” may be able to do this. Social systems cannot be considered as
biological observers because they remain distributed. Thus, they cannot be expected to Second-order
perform bodily operations like seeing and speaking. observer
Luhmann himself sometimes used the word “dividuum” in order to emphasize the
difference between a social system and individual observers. However, he defined
“observation” as the general operation of drawing a distinction and making an
identification. Thus, he extended the biological definitions of Maturana and Varela to a
general systems notion that would be applicable to social systems as well, because he 539
erroneously believed that the general definition had been provided by Spencer Brown
in mathematical terms. However, this definition was a static one, whereas one needed
Varela’s biological extension for the evolutionary concept. The generalization of the
notion of an observer – however, abstract – into the sociological domain led
eventually to a confrontation with Maturana. Maturana (1990) insisted on the physical
existence of an observer. Von Foerster (1999) similarly expressed his reservations
about Luhmann’s generalizations.

3.1 Can a second-order observer be considered a meta-observer?

By putting the “observer” in social systems theory on a par with the abstract definition
at the level of general systems theory a number of problems was generated. According
to Luhmann’s own theory of social systems, the social system would have to be an
observer that is qualitatively different from (since orthogonal to) human observers.
However, by defining this second-order observer in terms of the biological metaphor of
a first-order observer, the second-order observer could also be considered as a
supra-observer and thus a meta-biological metaphor was invoked. The second-order
observer could be endowed with capacities (like “seeing”) that tend to
anthropomorphize this abstract notion. The Greek gods seemed to be back on stage.
In the Luhmannian tradition, the consequent confusion is attributed to the limitations
of language and discussed in terms of paradoxes (Fuchs, 2004). However, the
paradoxes are not a result of linguistic limitations, like a tendency to reification in
the German language, but the consequence of a category mistake at the level of the
definitions which can be hidden by using linguistic metaphors.
This confusion of general (mathematical) and specific (biological) models has led to
a stagnation in the research program of social systems theory during the 1990s.
Gumbrecht (2003), for example, argued in a critical appraisal of Luhmann’s legacy that
the focus on the epistemology of paradoxes led to discussions during the 1990s that
failed to add new insights to the well-known problem of the hermeneutic circle in the
philosophy of language. This circle was not broken by distinguishing, for example,
between a semantic (first-order) and a linguistic (second-order) domain in terms of
different operations.
A distinction by a distributed system remains distributed and therefore cannot be
identified other than as an uncertainty. An identification, however, would assume an
identifiable observer like the analyst him-/herself. A distribution can only designate
such distinctions in a subsymbolical mode. The uncertainty remains and cannot be
resolved as in the case of an observation. The notion of “observation” as a concept of
general systems theory – including trans-individual observations – and the
attribution of the observation to a conscious observer would generate confusion at this
level because the operation of the social system can lead to changes in the semantic
fields among human beings beyond human will or intention. Such changes in the
K communication can be accessed by a reflexive observer, but they occur behind their
35,3/4 backs as the non-linear results of their and others’ interactions (Bhaskar, 1998; Marx,
1974, p. 176).

3.2 Expectations, observational reports, and observations

The specification of an uncertainty by making a distinction and an identification can
540 be considered not yet as an observation in sociological discourse, but as the
specification of an expectation. The distinction and designation specify a category, but
not yet the value to be given to this category in the research process. One needs
additionally a (scientific) observation for the specification of the latter. Thus, Luhmann
used an epistemological notion of observations as empirical givens (from biology) that
contradicts the concept of observation in empirical research in the social sciences.
For example, if one correlates two variables in a social science design, one is able to
calculate whether the observations are statistically significant by using, for example,
chi-square statistics. The values for the margin totals, that is, the nominal categories
serve us for the specification of expected values, but empirical measurement provides
the observed values. This difference between observed and expected is fundamental to
the sociological design also in qualitative case studies. The cases that have occurred
could have been different because of chance factors. In the qualitative case one confines
the measurement to only the nominal level – that is, the description of the events in
words – but the sociological design is thoroughly analytical and should avoid
historicism (Weber, 1917; Popper, 1967). Case studies can only inform analytical
theorizing because they provide the discourse with observational reports.
Sociological discourse processes observational reports rather than observations
(Pask, 1975; De Zeeuw, 1993; Matsuno, 2003). While the observations can be attributed
to observers or more generally to agency[3], only observational reports can function in
the communication. In scientific communications, for example, the observational
reports can be validated, i.e. provided with meaning by using the code of the scientific
communication system. Are the observations reported also robust? Are they valid and
reliable? A set of criteria can be applied and further developed within the scientific
communication system used for the evaluation of knowledge claims in articles
submitted for publication. The knowledge claims expressed in a linguistic domain
provide the variation to a next-order codification in a symbolic domain.
In summary, the use of the word “observation” for the operation is not sufficiently
reflexive in the case of the social system because this system operates in terms of
expectations that can be informed by observational reports on the basis of observations
by individuals. Observations update the expectations at the level of agency, but the
communication requires a reflexive turn. I elaborated this above in the case of scientific
communications, but similarly, in the economy, prices can be considered as
expectations of the values of goods. Like other expectations entertained at the level of
the social system, prices can be updated by observations if the reports about the
observations can be appreciated at the level of the social system. The social system
provides value to observations by codifying some observational reports more than
others.
On the other side of the divide of the structural coupling between social systems and
agencies value is provided to expectations by measurement because human beings can
perceive a (natural or social) state of affairs by using sense data. Thus, the mechanisms
for generating value are different in social systems and in consciousness systems. This Second-order
distinction is obscured when “observation” is used as a general denominator for both observer
expectation and observation, for the generation of value in terms of observations and in
terms of coding (e.g. true/false), and for the generation of an observer.
If we distinguish these categories, it becomes clear that the meta-observer fails to
exist at the level of the social system because this system remains distributed. The
social system should not be anthropomorphized by using a biological metaphor 541
(Hollak, 1963). This system has no organs for making an observation, but it is able to
entertain substantive distinctions and designations in a distributed mode and over
time. Luhmann (1984) himself emphasized these three dimensions of the
communication: a communication is at the same time substantive, social, and
temporal. The complex system which emerges from the non-linear dynamics of three
interacting subdynamics, is no longer necessarily grounded, that is, stabilized. In this
case, the distinctions and designations function as expectations at the various levels of
the system’s organization as in interactions, organizations, and at the (potentially
globalized) level of society (Luhmann, 1997b).

4. Conclusions
I have argued that a biological flaw in social systems theory has blocked its heuristic
value in empirical research. The notion of identifiable observation was not sufficiently
distinguished from that of a reflexive expectation. While the Luhmannian perspective
has offered an elaboration of the etic-perspective in anthropology and sociology –
because it abstracts from the intentions and the attributions of individual agency – it
has remained tied to the biological origins of systems theory because of the
generalization of the biological metaphor of an observer.
This generalization was justified in terms of Spencer Brown’s (1969) Logic of Forms.
However, one needs Varela’s (1975) biological addition of “identification” as a
necessary condition for “observation” for carrying the inference. Spencer Brown (1969)
discussed “distinction” as a necessary condition for “identification,” but only the
former can be considered as a mathematical operation. Identification is substantive: it
provides meaning to a distinction. Varela’s evolutionary perspective, for example,
could provide this substantive reflection. At the level of the social system, however,
observations are not relevant for the system without further reflection in an
observational report. These reports provide a distribution of observations that can be
compared (and sometimes tested for their significance). The distribution is relevant for
the social system because this system can operate only in terms of distributions.
Distributions at each moment in time can be expected to contain information. A
redistribution of the distributed substance over time communicates an information or
– in other words – generates a probabilistic entropy (Shannon, 1948). Thus, the
perspective of social systems theory can be combined with the tools available in the
mathematical theory of communications (Leydesdorff, 1995). However, this requires
that what are considered as “observations” in biological systems theory be recognized
reflexively and redefined as “expectations” for epistemological reasons.
Furthermore, this change in perspective from observations to expectations enables
us to bridge the gap with the study of meaning processing in symbolic interactionism.
Meaning is processed reflexively, that is, in terms of expectations (Knorr-Cetina and
Cicourel, 1981). However, these expectations are structured and therefore systemic
K (Glaser, 1992; Lazarsfeld, 1995). In this tradition because of its focus on interactions, the
35,3/4 systemic dimensions are indicated only as referentials. The deontologization of social
systems theory intended by Luhmann (1984, p. 243) accords with this perspective
(Gibson, 2000)[4], but the meta-biological metaphor has disturbed the translation of
the heuristic value of social systems theory in sociological discourse (Giddens, 1984,
p. xxxvii; Leydesdorff, 1993).
542
5. Implications
Biological systems theoreticians are in a position different from that of social systems
theoreticians because the biological systems are usually observable. Biological
systems can be considered as “natural,” and therefore the biologist is inclined to begin
with the specification of an observable variation rather than the uncertainty of an
expectation. As Maturana and Varela (1980, p. 90) formulated it emphatically:
Notions such as coding and transmission of information do not enter in the realization of a
concrete autopoietic system because they do not refer to actual processes in it.
While these authors insisted on the biological realization of “actual processes,”
Shannon’s co-author Weaver (1949, p. 116f) noted the problem of defining “meaning”
from a mathematical perspective. The epistemological challenge of sociology is to
abstract from observers as biological systems and to study systems of communication
that are able to communicate about meanings and expectations in addition to
observations (Leydesdorff, 2003).
Discourse does not require the realization of an observer. An identifiable observer
would be only one among its possible realizations. The self-organization of meaning,
however, can be considered as different from the self-organization of living (Luhmann,
1986). The systems-theoretical concept of “distinction and designation” can be used in
sociological discourse only if the result of this operation is accorded the epistemological
status of an “expectation.” Given this reflection, the proof of the pudding remains the
explanatory power of the specific perspectives of sociocybernetics and social systems
theory among other sociological discourses (Leydesdorff, 1997). By specifying
second-order observations as expectations, social systems theory and sociocybernetics
can combine the constructivist with an empirical approach.

Notes
1. The reasoning is analogous to the information theoretical statement that a distribution
contains an uncertainty. The specification of this uncertainty, for example, in terms of bits of
information is yet content-free. The specification of a system of reference for the information
can provide this information with meaning (Leydesdorff, 2003).
2. See Kauffman’s (2001) argument that mathematical reasoning about observers necessarily
leads to paradoxes.
3. Agency may also include the “institutional agency” of a research group as an aggregate of
agents or “principal agency” when underlying agents are represented.
4. “The point from which all further investigations in systems theory must begin is therefore
not identity but difference.
This leads to a radical de-ontologizing of objects as such – a discovery that corresponds to
the analyses of complexity, meaning, the pressure to select, and double contingency”
(Luhmann, 1995, p. 177).
References Second-order
Arthur, W.B. (1989), “Competing technologies, increasing returns, and lock-in by historical observer
events”, Economic Journal, Vol. 99, pp. 116-31.
Arthur, W.B. (1994), Increasing Returns and Path Dependence in the Economy, University of
Michigan Press, Ann Arbor, MI.
Baecker, D. (1996), “A note on composition”, Systems Research, Vol. 13 No. 3, pp. 195-203.
Baecker, D. (Ed.) (1999), Problems of Form, Stanford University Press, Stanford, CA.
543
Beck, U., Bonns, W. and Lau, C. (2003), “The theory of reflexive modernization: problematic,
hypotheses and research program”, Theory, Culture & Society, Vol. 20 No. 2, pp. 1-33.
Bernstein, B. (1971), Class, Codes and Control, Vol. 1: Theoretical Studies in the Sociology of
Language, Routledge & Kegan Paul, London.
Bhaskar, R. (1998), “Societies”, in Archer, M., Bhaskar, R., Collier, A., Lawson, T. and Norrie, A.
(Eds), Critical Realism: Essential Readings, Routledge, London, pp. 206-57.
Blauwhof, G. (1995), “The non-linear dynamics of technological developments: an exploration of
telecommunications technology”, unpublished PhD thesis, University of Amsterdam,
Amsterdam.
Coser, R.L. (1975), “The complexity of roles as a seedbed of individual autonomy”, in Coser, L.A.
(Ed.), The Idea of Social Structure. Papers in Honor of Robert K. Merton, Harcourt Brace
Jovanovich, New York, NY, pp. 237-64.
De Zeeuw, G. (1993), “Improvement and research: a Paskian evolution”, Systems Research, Vol. 10
No. 3, pp. 193-203.
Edelman, G.M. (1989), The Remembered Present: A Biological Theory of Consciousness, Basic
Books, New York, NY.
Fuchs, P. (2004), Der Sinn der Beobachtung: Begriffliche Untersuchungen, Weilerswist, Velbrück.
Geertz, C. (1973), The Interpretation of Cultures, Basic Books, New York, NY.
Gibson, B. (2000), “Practical applications of Luhmann’s work: observations from a grounded
theory perspective”, Luhmann On-Line, Vol. 1 No. 4, available at: www.sociocyberforum.
org/newsletter/lol_nwsltr_1n4.pdf
Giddens, A. (1976), New Rules of Sociological Method, Hutchinson, London.
Giddens, A. (1979), Central Problems in Social Theory, Macmillan, London.
Giddens, A. (1984), The Constitution of Society, Polity Press, Cambridge.
Glanville, R. (2002), “Heinz von Foerster”, Soziale Systeme, Vol. 8 No. 2, pp. 155-8.
Glaser, B.G. (1992), Emergence vs Forcing: Basics of Grounded Theory Analysis, Sociology Press,
Mill Valley, CA.
Goguen, J.A. and Varela, F.J. (1979), “Systems and distinctions: duality and complementarity”,
International Journal of General Systems, Vol. 5, pp. 31-43.
Grathoff, R. (Ed.) (1978), The Theory of Social Action. The Correspondence of Alfred Schutz and
Talcott Parsons, Indiana University Press, Bloomington, IN.
Gumbrecht, H.U. (2003), “What will remain of Niklas Luhmann’s philosophy? A daring (and
loving) speculation”, paper presented at the Opening of Systems Theory, Copenhagen,
23-25 May.
Günther, G. (1967), “Time, timeless logic, and self-referential systems”, Ann. N.Y. Acad. Sci,
Vol. 138, p. 396.
Habermas, J. (1981), Theorie des Kommunikativen Handelns, Suhrkamp, Frankfurt.
K Habermas, J. (1987), “Excursus on Luhmann’s appropriation of the philosophy of the subject
through systems theory”, The Philosophical Discourse of Modernity: Twelve Lectures, MIT
35,3/4 Press, Cambridge, MA, pp. 368-85.
Habermas, J. and Luhmann, N. (1971), Theorie der Gesellschaft oder Sozialtechnologie, Suhrkamp,
Frankfurt.
Hollak, J. (1963), “Hegel, Marx en de Cybernetica”, Tijdschrift voor Philosophie, Vol. 25, pp. 270-94.
544 Kant, I. (1956), Kritik der reinen Vernunft, Felix Meiner, Hamburg.
Kauffman, L.H. (2001), “The mathematics of Charles Sanders Pierce”, Cybernetics & Human
Knowing, Vol. 8 Nos 1-2, pp. 79-110.
Knorr-Cetina, K.D. and Cicourel, A.V. (Eds) (1981), Advances in Social Theory and Methodology.
Toward an Integration of Micro- and Macro-Sociologies, Routledge & Kegan Paul, London.
Lazarsfeld, P.F. (1995), “Analyzing the relations between variables”, in Barnes, B.G. (Ed.),
Grounded Theory 1984-1994, Sociology Press, Mill Valley, CA, pp. 83-102.
Leydesdorff, L. (1993), “‘Structure’/‘Action’ contingencies and the model of parallel processing”,
Journal for the Theory of Social Behaviour, Vol. 23 No. 1, pp. 47-77.
Leydesdorff, L. (1994), “The evolution of communication systems”, Int. J. Systems Research and
Information Science, Vol. 6, pp. 219-30.
Leydesdorff, L. (1995), The Challenge of Scientometrics: The Development, Measurement, and
Self-Organization of Scientific Communications, DSWO Press, Leiden University, Leiden,
available at: www.universal-publishers.com/book.php?method ¼ ISBN&book ¼
1581126816
Leydesdorff, L. (1997), “The non-linear dynamics of sociological reflections”, International
Sociology, Vol. 12, pp. 25-45.
Leydesdorff, L. (2000), “Luhmann, Habermas, and the theory of communication”, Systems
Research and Behavioral Science, Vol. 17, pp. 273-88.
Leydesdorff, L. (2001a), “Technology and culture: the dissemination and the potential ‘Lock-in’ of
new technologies”, Journal of Artificial Societies and Social Simulation, Vol. 4 No. 3, Paper
5, available at: http://jasss.soc.surrey.ac.uk/4/3/5.html
Leydesdorff, L. (2001b), A Sociological Theory of Communication: The Self-Organization of the
Knowledge-Based Society, Universal Publishers, Parkland, FL, available at: www.
upublish.com/books/leydesdorff.htm
Leydesdorff, L. (2002), “The communication turn in the theory of social systems”, Systems
Research and Behavioral Science, Vol. 19 No. 2, pp. 129-36.
Leydesdorff, L. (2003), “The construction and globalization of the knowledge base in inter-human
communication systems”, Canadian Journal of Communication, Vol. 28 No. 3, pp. 267-89.
Luhmann, N. (1975), “Interaktion, organisation, Gesellschaft”, Soziologische Aufklärung 2,
Westdeutscher Verlag, Opladen, pp. 9-20.
Luhmann, N. (1984), Soziale Systeme. Grundriß einer allgemeinen Theorie, Suhrkamp, Frankfurt.
Luhmann, N. (1986), “The autopoiesis of social systems”, in Geyer, F. and Zouwen, J.v.d. (Eds),
Sociocybernetic Paradoxes, Sage, London, pp. 172-92.
Luhmann, N. (1990), Die Wissenschaft der Gesellschaft, Suhrkamp, Frankfurt.
Luhmann, N. (1993), “Die paradoxie der form”, in Baecker (Hrsg.), D. (Ed.), Kalkül der Form,
Suhrkamp, Frankfurt, pp. 197-212.
Luhmann, N. (1995), Social Systems, Stanford University Press, Stanford, CA.
Luhmann, N. (1997a), Die Gesellschaft der Gesellschaft, Suhrkamp, Frankfurt.
Luhmann, N. (1997b), “Globalization of world society: how to conceive of modern society?”, Second-order
International Review of Sociology, Vol. 7 No. 1, pp. 67-79.
observer
Luhmann, N. (2000), Organization und Entscheidung, Westdeutscher Verlag, Opladen.
Luhmann, N., Maturana, H., Namiki, M., Redder, V. and Varela, F. (1990), Beobachter:
Konvergenz der Erkenntnistheorien?, Wilhelm Fink Verlag, München.
McLuhan, M. (1964), Understanding Media: The Extension of Man, McGraw-Hill, New York, NY.
Marx, K. (1974), Grundrisse der Kritik der politischen Oekonomie [1857-1858], Dietz Verlag,
545
Berlin.
Matsuno, K. (2003), “Quantum mechanics in first, second and third person descriptions”,
BioSystems, Vol. 68, pp. 107-18.
Maturana, H.R. (1978), “Biology of language: the epistemology of reality”, in Miller, G.A. and
Lenneberg, E. (Eds), Psychology and Biology of Language and Thought. Essays in Honor of
Eric Lenneberg, Academic Press, New York, NY, pp. 27-63.
Maturana, H.R. (1990), “The biological foundations of self consciousness and the physical
domain of existence”, in Luhmann, N., Maturana, H., Namiki, M., Redder, V. and Varela, F.
(Eds), Beobachter: Konvergenz der Erkenntnistheorien?, Wilhelm Fink Verlag, München.
Maturana, H.R. (2000), “The nature of the laws of nature”, Systems Research and Behavioural
Science, Vol. 17, pp. 459-68.
Maturana, H.R. and Varela, F.J. (1980), Autopoiesis and Cognition: The Realization of the Living,
Reidel, Dordrecht.
Münch, R. (1982/1988), Theorie des Handelns, Suhrkamp, Frankfurt, Understanding Modernity
Routledge, London.
Parsons, T.S. (1937), The Structure of Social Action, The Free Press, Glencoe, IL.
Parsons, T.S. (1951), The Social System, The Free Press, New York, NY.
Pask, G. (1975), Conversation, Cognition and Learning, Elsevier, Amsterdam.
Popper, K.R. (1967), The Poverty of Historicism, Routledge & Kegan Paul, London.
Shannon, C.E. (1948), “A mathematical theory of communication”, Bell System Technical Journal,
Vol. 27, pp. 379-423 and 623-356.
Simon, H.A. (1969), The Sciences of the Artificial, MIT Press, Cambridge, MA.
Spencer Brown, G. (1969), Laws of Form, George Allen and Unwin, London.
Spencer Brown, G. (1994), “Self-reference, distinctions and time”, Teoria Sociologica, Vol. 2-3
No. 1, pp. 47-53.
Varela, F. (1975), “A calculus for self-reference”, International Journal of General Systems, Vol. 2
No. 1, pp. 5-24.
Varela, F.J. and Goguen, J.A. (1978), “The arithmetic of closure”, Journal of Cybernetics, Vol. 8,
pp. 291-324.
Von Foerster, H. (1975), “Notes on an epistemology for living things”, L’unité de l’homme, Plon,
Paris.
Von Foerster, H. (1979), “Cybernetics of cybernetics”, in Krippendorff, K. (Ed.), Communication
and Control in Society, Gordon and Breach, New York, NY, pp. 5-8.
Von Foerster, H. (1982), Observing Systems, Intersystems Publications, Seaside, CA.
Von Foerster, H. (1999), “An Niklas Luhmann”, in Bardmann, T.M. and Baecker, D. (Eds), Gibt es
Eigentlich den Berliner Zoo noch? Erinnerungen an Niklas Luhmann, Universitätsverlag
Konstanz, Konstanz, pp. 13-15.
K Wasser, H. (2001), “Luhmanns theorie psychischer systeme und das freudsche unbewusste: zur
beobachtung strukturfunktionaler latenz”, Vielbeck Wissenschaft, Vol. 1, available at:
35,3/4 www.velbrueck-wissenschaft.de/pdfs/haraldwasser.pdf
Weaver, W. (1949), “Some recent contributions to the mathematical theory of communication”, in
Shannon, C.E. and Weaver, W. (Eds), The Mathematical Theory of Communication,
University of Illinois Press, Urbana, IL, pp. 93-117.
546 Weber, M. (1917), “Der Sinn der ‘Wertfreiheit’ der soziologischen und ökonomischen
wissenschaften”, Gesammelte Aufsätze zur Wissenschaftslehre, Mohr, Tübingen,
pp. 489-540, 31968.

Further reading
Luhmann, N. (1990), “Sthenographie”, Beobachter: Konvergenz der Erkenntnistheorien?, Wilhelm
Fink Verlag, München, pp. 119-37.

Corresponding author
Loet Leydesdorff can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints