Theory & Science (2006)

ISSN: 1527-5558


Elke Weik
Dept. of Economics and Business Administration
Chair for Organization Studies
09107 Chemnitz, Germany
phone: +49/371/531-4150


The article discusses the various contributions to the structure-agency debate with regard to their metaphysical assumptions concerning structure and agency. It argues that dualism theories, for all their positive contributions, cannot overcome certain problems due to their Cartesian heritage, which causes them to conceptualize the relationship between structure and agency in certain – deficient – ways. The article then critically analyzes the duality theories of Giddens and Bourdieu, which claim to have overcome these problems. Finally, it discusses metaphysical alternatives in form of the non-Cartesian metaphysics of Heidegger and Whitehead.

”Everybody ought to be a nominalist at first, and continue in that opinion until he is driven out of it by the force majeure of irreconcilable facts.“ (C.S. Peirce)

The relationship between structure and agency has been the subject of countless publications. It has vexed, for example, organization scholars from very different schools and times such as March (Cyert & March, 1992; March & Olsen 1975), Weick (1969), Goffman (1959), Foucault (1977; 1978), Luhmann (1995), McKelvey (1997), or Giddens (1984): How do people act in structured environments like organizations? Do they act systematically differently? Can they act at all? Or is it rather “the system” and its characteristics, which propel an organization through the course of time? Similar questions have been asked, albeit with different stresses, in sociology, social theory, social psychology, history, philosophy and the many areas around and between them. A comprehensive discussion of them would, of course, fill several volumes. Hence, this article will center on the issue of structure and agency as it is discussed in social theory. Moreover, as the choice of “structure-agency” rather than “micro-macro” may already indicate (Ritzer, 1996), it will do so from a European perspective. My basic contention is that the reason why the aforementioned vexation continues can be found in an analysis of the metaphysical1 assumptions implicit in the different theories. In the first section, I will show that “dualism theories” (like structuralism or rational choice) still carry with them the metaphysical heritage of a Cartesian philosophy based on the distinction between mind vs. matter, and subject vs. object. This prevents them from overcoming either an “over-socialized” or an “over-individualized” position in the structure-agency debate. The next section will then be dedicated to the “duality theories” of Giddens and Bourdieu, which claim to have healed the woes besetting dualism. I will, however, argue that they, too, still employ a Cartesian metaphysics and hence suffer – albeit in a milder form – from the same problems. My final argument is directed towards explicitly non-Cartesian ontologies. In this category, I have decided on Heidegger and Whitehead for a brief exposure because they set out explicitly to attack and overcome the kind of thinking that is related to Cartesian philosophy.

Analyzing the assumptions of so many different theories, it seems only fair to explain the major presuppositions of this article. The first is that it makes sense to discuss metaphysics or ontology as different from epistemology or methodology. In this, I follow the very traditional line prominently expressed by Popper (1992) that there is a difference between what is (ontology) and what we say or know about it (epistemology), although we may never be entirely sure about the content of this difference. It is the line followed by such different traditions as realism, materialism, or idealism, but rejected by pragmatism or language-oriented theories in the Wittgenstein tradition. My second contention is that although it is perfectly possible to do social science without regard for metaphysical issues, there are times when such an analysis is helpful. One, as already Weber (1988:214) saw, is the time of paradigm shifts; another is the dissatisfaction with existing theories on a general plane, as in the case of the structure-agency debate. The idea behind it is that theories that differ in their explicit content may nevertheless share the same metaphysical commitments, which cause them to reproduce the same errors. This leads to my third presupposition, which is to compare and contrast different metaphysics rather than to argue for one “best way”. My concern is with their respective logical consistencies and limits, not with their correspondence to “reality”. It is an attempt to classify theories in order to deduce a priori some of their potentials and limits. This involves not only the practical concern of saving time in finding weak spots beforehand, but also judging theories on a different epistemic level. It means that there is a difference between stating that something does not work in a specific theory (which I will call episteme 1), stating why it does not work in this theory (episteme 2), and stating why it can never work in this kind of theory (episteme 3). This article is mainly concerned with epistemes 2 and 3. Finally, as empirical research cannot prove or disprove metaphysics, my argument strongly depends on logic. I consider the metaphysical assumptions as “premises” in the logical sense, on which rest “conclusions” forming the body of the respective social theory2.

The Metaphysics of Structure and Agency

Let us now have a closer look at the metaphysical background of the concepts of agency and structure, and their relationship. Basically, there are three ways of dealing with this relationship: regarding it as a dualism, as a duality, or avoiding any commitment in this respect. Reed (1997:23) characterizes the first two : “Participation within [dualism] necessarily entailed the formulation of a crucial analytical distinction between actors’ abilities to engage in forms of social conduct […] as opposed to the objective structural conditions or constraints which limit and regulate the innate transformative capacity of social action”, and on the other hand, “[Duality] approaches articulate a shared commitment to a social ontology that dispenses with the need to distinguish analytically between different levels or forms of social reality as represented in the ‘agency/structure’ distinction.” The third stance is characteristic for pragmatist approaches, postmodern approaches, as well as most approaches focusing on language, meaning, or sense-making by asking how people view or talk about structure and agency. They either refuse metaphysical inquiries for certain reasons (for example, postmodern or pragmatist approaches), or connect metaphysical and epistemological issues in a way that makes it impossible to inquire into metaphysics as different from epistemology (for example, constructionism). While not sharing Reed’s critique of these “flat ontologies”, I will nevertheless concentrate on dualism and duality approaches as only they make the kind of metaphysical assumptions this article is concerned with.

Let us start with the metaphysical assumptions underlying the pole termed “action”, “agency”, or “agent”. With regard to action, the metaphysical picture is most fuzzy. This is due to several reasons. First, a lot of sociological and social theories discussing action do not define it. Some of them they replace the term with communication and/or meaning – hence Campbell’s (1998:8) verdict: “Indeed, the judgment that the majority of ‘action theories’ do not, in practice, really address the phenomenon of ‘action’ is one which it is hard to avoid no matter which perspective is under examination.” (in a similar vein, Crespi, 1994). Others concentrate on agents’ endowments or characteristics (see below). Second, those which have defined the term often face the problems described by Luhmann (1982:366): “Perhaps actions are too well-known and too easy to observe, which means that conceptual precisions cannot go very far without failing.” Finally, the philosophy of action, while excellent in discussing conceptual problems, has mostly failed to provide positive conceptual support regarding central notions like intention, will, or decision (Fuchs, 2001). Cohen (1997:112) sums up: “For a variety of reasons, social action inherently defies all efforts to produce a broadly acceptable unifying theory. In this respect, action seems more like poetry than a natural phenomenon, presenting theorists with orderly rhythms, meanings, and forms, but no consolidating principle that brings all patterns together.” In consequence, I will limit my observations to the concepts of agent and agency. For most theories, the metaphysics behind the latter two is rather clear: when we refer to “agent”, we mean a human being who has a body and a mind and/or brain, and the major debate centers on the extent to which this agent is free in his/her choice of actions and to what extent he/she is determined by social forces. This represents, of course, the classical divide between “subjective” and “objective” theories as defined by Burrell & Morgan (1979), and the origin of the structure-agency debate. However, both3 sides agree on the image of a human being as “consisting of” body and mind, thus reproducing the classical Cartesian division4 as portrayed below. Agency is then defined as a capability of the agent (for example, Giddens, 1984), or – in contrast to behavior – as something involving individual intention or purpose.

Next, what are social structures? A large variety of definitions springs to mind, which I should like to present first in a loosely structured manner, before classifying them according to their metaphysical content in the preceding section. To locate the ideas in their respective theories, I shall also provide the name of one major proponent, which by no means implies that he is the only one to have put forward this idea. There are a number of what Smelser (1989) calls “generic features” of social structures. The most basic refers to the relation, mostly through interaction, of two or more people, as expressed in Elias’ (1999) definition of social structures as social relations. Within this relational characterization, for example, defining social structures as shared knowledge (Garfinkel, 1996) is a more actor-centered variety of the constitutive relation, while defining them as functions (Parsons & Bales, 1953) covers a more society-centered variety. Another generic feature is that structures are repetitive. The focus on repetition gives rise to definitions of social structures as generalizations, i.e. words or expressions coined by the observer, as in the second level analysis of systems theory (Fuchs, 2001), or of independently existing patterns within society and social life, as in the deep structures of Levi-Strauss (1968). Repetition is also at the heart of linking structures and practices, as, for example, in Weber’s (1979) routines. A third generic feature is the normative aspect of social structures. From there derive definitions of social structures as rules and resources (Giddens, 1984), or as constraints or tools for action (Barley & Tolbert, 1997). Defining social structures as reciprocal expectations (Luhmann, 1995) also carries a strong normative overtone. The power involved in the normative aspect can also lead to perceiving structures as forms of ordering or organizing a group of people, as, for example, in the disciplinary mechanisms of Foucault (1977). Some of these definitions overlap, some have been taken together to form several levels of structure embedded in one another (see for example Layder, 1981; Prendergast & Knottnerus, 1994; López & Scott, 2000), and some are, of course, contradictory.

Let us now consider the relation between structure and agency. Agency can be affected by structure through powers or needs traditionally located “inside” the agent. Thus, we find the idea that people know about structures and actively use them or avoid their sanctions, as, for example, the rule-following of Burns & Flam (1987). The repetitive character of social structures may relate to the fact that people want to repeat actions, either because it fulfils their ontological needs (Giddens, 1984), or because these actions are central to their identity, as, for example, in the individual preferences and goals of Rational Choice Theory (Elster 1986). In a less self-determined vein, social structures may induce actions due to habit (Weber, 1979), or because there exist subconscious motives (Lévi-Strauss, 1968). Actors may, of course, also simply be forced by others to act in a certain way, as, for example, by imposition (Scott, 1987), or lured by the creative power of structures that enable action, create sense etc. (Foucault, 1977).

Vice versa, the same authors would argue that the creation (and, respectively, the maintenance, modification, and abolition) of structures can be affected by agency through repetition, i.e. when actions repeated often and by many create or sustain structures, as, for example, in all accounts referring to habits, routines, and practices. In a diachronic perspective, repetition also occurs when structures are passed on to the next generation, as, for example, in all accounts featuring a historical, evolutionary or genetic perspective. A more causal approach is taken when the (re‑)production of structures is linked to knowledge or collective memory, i.e. when only structures that are remembered and taken into account by actors can survive, as, for example, in all accounts referring to knowledgeable actors. Finally, again, force and imposition act as sources for the creation of structures. For the sake of completeness, one group of authors that differs fundamentally from all the above approaches should also be mentioned , viz. those who argue that structures are not affected by (individual) actions because they precede them, as do most structuralist accounts.

Again, this enumeration contains elements that may be supplementary while others are contradictory. Nevertheless, I hope to have collected the most influential ideas. If we take these ideas as a kind of “corpus” on how we conceptualize social structure and its relation to agency, I believe, contrary to many complaints, that we already know a lot about how structures work at the point of action.

However, two problems endure. First, the location or the “where” of existence of social structures creates difficulties. However, any account of structure that omits this question lacks a basic definitional item. We simply cannot grasp an idea completely if we cannot imagine how or where it exists. The most prominent candidate for this localization is, of course, the individual mind, be it cognitive or emotive, conscious or subconscious, practical or discursive. A second, closely related location is the human brain. Other possible locations are the body, human “essence” or “nature”, the act itself, language, or some place independent of human interference. However, they all have their – metaphysical – problems, which impede their usefulness for social theory. I shall discuss them in a moment.

Second, the precise mode of social structure “influencing” or “affecting” agency is still unspecified. Is there causality involved, and if yes, what kind? If no, what else?

Both problems, which I shall call the “problem of existence” and the “problem of causality”, can be traced back to the beginnings of modern philosophy, generally associated with René Descartes. According to conventional wisdom 5, he conceptualized two most influential dichotomies: that between mind (which he called res cogitans) and matter (res extensa), and that between the perceiving subject and the perceived object. The former draws a strict line between the spiritual and the material sphere denying any form of exchange between the two. The latter forms the ground for epistemological questions regarding the observer-dependent or observer-independent characteristics of an object.

In the following paragraphs, I shall first discuss concepts of structure, which can be placed in the dualism of res cogitans vs. res extensa. I will then discuss non-dualist approaches.

Structure in the Dualist Conception � A Topography of Problems

Sorting the above collection of definitions for structure along the line of the res cogitans-res extensa distinction, which is, by definition, ontologically comprehensive and mutually exclusive (because every simple entity is either mind or matter, according to Descartes), we arrive at two groups. The first group considers structure a reflection or entity of the mind, and the other considers it an external entity. The former may formulate structure concepts in terms of generalization, perceived6 pattern, perceived function, rule, norm, expectation, mental entity, or human nature. The latter conceptualizes structure as order, external pattern, or external function. However, in order to relate these concepts to traditional metaphysical positions, we have to ask a few more questions (for an overview of the following see figure 1).

Do social structures exist only in the mind? Answering with a “no” establishes an ontological realism, which assumes that entities such as structures exist independently of the mind. With regard to the label “realism” a caveat is necessary: there exist more than thirty kinds of realism, which sometimes stand in stark contrast to one another. For example, the range goes from a “natural realism” stating that things are how we perceive them, via Kant’s “transcendental realism” to Peirce’s “semiotic realism” which assumes that only relations are real (Hoffmann, Halbfass, Trappe, Grünewald, & Abel, 1992). Thus, “realism” cannot be regarded a proper description of a position in itself. Most versions developed in the 20 th century focus on epistemology without even providing a precise metaphysics. Some of these would even have difficulties accepting the metaphysical propositions that I describe in the following. I shall thus label the kind of realism I use “realism ME” indicating that I refer to the medieval universalist realism. This realism ME states that universals, for example, social structures, have an existence independent of the human mind. Answering the initial question with a “yes” leads to either to idealist positions if we take “mind” in its literal sense, or to materialist positions if we equate “mind” with “brain” (Note that materialist positions are res extensa positions.)

Are there immaterial substances (in the mind)? A substance in the metaphysical sense is that which guarantees the sameness of an object in a change process. Whether an apple changes its color or a person from childhood to old age, there is supposed to be something, which continues and permits talk of “the same apple that now has a different color”. In order to distinguish idealism and materialism7, or respectively mind and brain as “carriers of ideas”, we have to ask whether substances can be immaterial.

Are there supra-individual immaterial substances? Saying “yes” to the former question gives rise to this question about whether we deal only with individual minds or with a supra-individual spiritual entity, be it social or divine in nature. Classically, theories that put society before the individual, like Durkheim’s, tend to occupy this position, which I will label “collective idealism”. Negating the question leads to an “individual idealism” discussing individual minds only.

Have brains got irreducible mental properties? Materialist positions that assume that there are only brains doing all the thinking and feeling may still be divided over the question of mental properties. Mental properties are the states that we ascribe to a mind in everyday language, like joy, pain, and thirst. They can be distinguished from physical states of the brain, like the activity of a certain brain sector or the impulse of a nerve. Some materialists, whom I shall call “property physicalists”, argue that mental properties can be reduced to physical properties, while others (“property dualists”) argue that they cannot.

As a result, we have five basic metaphysical positions concerning the nature of social structures when distinguished from action:

These five metaphysical positions meet – and always have met – with typical problems. Hence, the corresponding social theories, being logically dependent on their metaphysics (as conclusions are on premises), cannot avoid or solve these problems either. Nonetheless, each of them has specific explanatory advantages. They follow the (thumb-) rule of the reciprocal complexity of premises and arguments. On the one hand, presupposing many things will make the argument easy and elegant, but asks a lot in terms of accepting the premises. On the other hand, simple premises will be easily acceptable, but leave more to argue. Take, for example, the nature-nurture discussion: a newborn with a “tabula rasa mind” is easy to conceptualize, but to explain how such a child can acquire any knowledge is very difficult. On the other hand, presupposing innate knowledge opens much room for doubt and discussion, but makes the explanatory task much easier. The same principle applies in our case. Assuming an independent order (realism ME) or supra-individual entity (collective idealism) asks quite a lot of the audience – I shall call this the “problem of metaphysics” – but makes it easy to construe a convincing argument concerning how structures affect agency. The problem of metaphysics is further exacerbated by the problem of empirics, i.e. the failure of these positions to provide empirical proof for their presupposed entities. While this demand for proof is not entirely fair given the relationship between metaphysics and empirical research, I think it is nevertheless quite common among critics. On the other hand, assuming only individual minds or brains (individual idealism, materialism) is quite unproblematic, but complicates the argument in two respects. First, in the case of individual idealism and property dualism, we get the well-known problems of order and transcendence. The former refers to the question of how individual minds controlling individual action can account for societal order, the latter to the question of how this order can transcend individual actions and lives. Second, property physicalism suffers from a logical problem (Beckermann, 1999; Kim, 1996) having up to now failed to come up with a convincing argument of how mental states (e.g. love) can be reduced to physical ones (e.g. hormones).

With regard to our second enduring problem, that of causality or structures “affecting” agency, all positions have difficulties solving the classical mind-body problem which inheres the res cogitans-res extensa dualism. If the two res are thought to be completely different ontologically, then it is hard to see how they can interrelate, i.e. how matter can affect the mind and vice versa. However, fortunately, the structure-agency causality hardly ever corresponds to the mind-body causality. Instead of asking how the individual mind “transmits” its will, motives and desires to the acting body (a classical problem of the philosophy of action), social science is interested in a structure-mind causality, i.e. an explanation of how social structures affect the minds of agents. With regard to their stance on this causality, the positions are, again, divided. Both idealism and materialism locate structures where action is also supposed to start, viz. in the mind or brain. If we take the Aristotelian causa efficiens (i.e. physical causation), as the “normal” case of causation, then the only position without any problems is, of course, property physicalism because both mind and body are physical and can affect each other in that way. Property dualism suffers from the aforementioned mind-body problem, as some properties of the structure-agency relationship are mental, others physical, and it is not clear how they interrelate. Idealist positions can, of course, not refer to physical causation. In the case of collective idealism there has to be a mechanism translating structures in the social or divine “mind” to the individual actor. This, as a rule, is not too difficult, as some sort of spiritual access to this supra-individual “mind” is presupposed. For individual idealism, structures can only affect agency if they are conceptualized as something that can be known (not necessarily explicit knowledge). Finally, realism ME must also define some epistemic or metaphysical relation between (external) structures and the actor. This can be achieved in various ways ranging from sense perception to Platonic8 participation in eternal ideas. Table 1 gives a summary of the advantages and problems as well as of the position on causality.

To state it once again: these problems will resurface every time a researcher makes the respective metaphysical assumptions - they cannot be solved within this frame of reference. And while I still deem it legitimate to use these assumptions – for they also have their advantages – we must accept their limits to prevent ourselves from wasting time on arguments which can be shown to be logically inconsistent on a general plane.

Structure in Non-Dualist Conceptions � No More Problems?

For those who do not want to accept these limits, two roads remain open. One is basically to retain the familiar notions of human agent as mind/body-entity and of structure as portrayed above, but to eliminate those characteristics that make an analytical distinction necessary. This is the option of the duality theorists. The other is to reconceptualize agents and/or structure in a more thorough way establishing a completely new metaphysics. In this section, I will discuss both alternatives in turn, drawing upon the social theories of Giddens and Bourdieu for the duality option and upon the philosophies of Heidegger and Whitehead for the new metaphysics option. Selecting those four prominent authors is not to imply that they are the only ones in the respective fields nor can it provide an exhaustive argument in the sense of episteme 3. It may, however, serve to explore the possibilities and dangers that lie on these roads.

Giddens. For structuration theory (Giddens, 1984), the agent is a human being with body and mind whose major defining feature with respect to acting is his or her “knowledgeability”. Giddens defines it as a “cognitive skill” (1984:2). Agency, then, is the capability of doing things (1984:9), i.e. of actively changing the course of events by causal intervention. Action, finally, is defined as flow of events, or durée, constituted by a discursive moment of attention (1984:3). As to structure, Giddens (1984:16f) distinguishes between social structure as a generic term, structures, and structural properties. The first refers to a virtual, paradigmatic order of transformational rules. This structure does not exist independently but only in its space-time instantiations in social practices and memory traces. Structural properties, which belong to societies, make it possible for practices to span space and time. Structure, as he states explicitly, is neither a patterning of social relations nor external to human action. The relationship between structure and action is mediated by practices, one of the central concepts of Giddens’s theory. While his concept of structure, being instantiated in the human mind9, clearly tends to a res cogitans position10, the concept of practice is supposed to bridge the classical gap between structure and action by combining the two in a sort of “structured action” or “enacted structure”. A second stress lies on the explanatory role of unintended consequences as effects which couple structure and action while at the same time freeing action from (conscious, active) intention. Third, time and space are conceptualized as “locales”, i.e. as enduring features, which nevertheless shape local actions and perceptions by providing a specific meaning for limited situations. With these three concepts – practices, unintended consequences, and locales – Giddens aims to overcome the problems of order and transcendence connected with his otherwise individualist idealist stance.

Does it work? Not really. First, nothing (regarding our problem) is solved by substituting practices for actions. The concept of practice does not change the character of the individual action but only places it in a wider pattern of actions. This has been done before without using the word “practice”; think for example of generalizations or observed patterns. In other words, the individual action is still what it was: a routine or habit (which appears as item in our above enumeration) developed to ease ontological insecurities. Its structuring is not immanent but external, accomplished by the observer. No doubt, Giddens seeks to move away from classical concepts of agency by introducing non-intentional notions like practices and practical consciousness. However, the “tacit knowledge” displayed in these forms is still contained in the agent and seems thus just a special form of the classical knowledge concept. Here, as in the third case, Giddens is very much inspired by Heidegger (and Wittgenstein), but seeks to avoid, or rather neglect, the theoretical consequences of the original concepts11. We will come back later to this point. Secondly, the explanatory value of unintended consequences rests on them forming “acknowledged conditions of further action” (Giddens, 1984:14). Giddens continues: “To understand what is going on no explanatory variables are needed other than those which explain why individuals are motivated to engage in regularized social practices…” For this, structuration theory offers two explanations. First, ontological insecurity prompts individuals to adopt practices because others employ them as well and to repeat them in order to achieve personal continuity and predictability. Second, rational deliberation may tell people that these established structures provide one approved and sometimes even the only accepted solution to their problem. In both cases, there is an appeal to human nature involved in the argument, again something we can find in the above enumeration. However, this link between structure and action, while plausible in the explanation of individual conduct, cannot explain order on a societal level. For even if we assumed that everybody acted rationally (which would be quite an assumption), there is still nothing that could guarantee the social adjustments necessary to form a social unit. Adhering to the same rules is not enough if these rules only take the individual into account. Third, the concept of locale is based on the idea that time constitutes human action. Giddens (1987) here draws heavily upon Heidegger’s time conception which regards time and being as “equiprimordial” (see below). This enables Giddens to link two of the three “pure macrovariables” (Giddens, 1984:140), viz. time, space and number, to “micro-situations” of action and co-presence. While the concept is helpful in many respects and rightly places time at the center of social analysis, it is bought at the price of theoretical inconsistency. For taking Heidegger’s time conception is not possible without taking (most of) Heidegger’s metaphysics, especially his notion of Dasein. (We will return to this point in the section on Heidegger.) Giddens (1984), however, describes his agent in familiar sociological-psychological terms as a person with a body, consciousness etc. – the very ideas Heidegger set out to combat. Summing up, while Giddens certainly has been successful in rendering some aspects of the structure-agency relationship in a more precise and sometimes innovative way, his duality has not been able to overcome, at least in a consistent manner, the basic problem12 connected with his metaphysical position.

Bourdieu. In many of his writings, Bourdieu (1984; 1990; and Bourdieu, Chamboredon, & Passeron, 1991) attacks the culture-nature dichotomy as one resulting from the mind-matter dualism. His major point is that every difference is at the same time natural and cultural, or material and symbolic. Objects must not be separated from the perception of objects, as is the custom of the subject-object dualism. In all his central concepts – habitus, capital, social space – we find this ambivalence of social structure on the one hand and material characteristics (embodiment, accumulation, distance) on the other. Indeed, I think Bourdieu’s major claim is that both aspects are inseparable. His structuralist provenance shows in his conception of structures as permanent modes of generation and operation, which operate without a voluntaristic subject. He even explicitly stresses the primacy of relations over subjects (Bourdieu, Chamboredon, & Passeron 1991). Nevertheless, he transcends structuralism by introducing the possibility of strategic conduct within the field of structures. In my view, his theory is in many respects better suited to overcome the structure-agency dualism than Giddens’s structuration theory because he relies conceptually much less on the free, cognitive agent than does the latter. Still, his theory cannot completely overcome the problems connected with a structure-oriented approach. While he, in his empirical work (e.g. Bourdieu, 1984), tries to reduce the metaphysical problem of structure by arguing statistically (i.e. structure is what we perceive a significant number of people to do or be), Durkheim’s fait social still shines through on many occasions. In other words: if structure is only a statistical fact, what is the ontological status of that fact? Durkheim believed it an independent entity, which only became visible through statistics, while nowadays we mostly think of statistics as a human artifact. Bourdieu does not position himself clearly on this question13. Thus, again, the problem of metaphysics, albeit in many places convincingly circumvented, remains in the end unsolved.

In conclusion, both theories, for all their other merits, still suffer from the Cartesian heritage they claim to have overcome. It seems indeed that – as in many other cases – “being a bit pregnant” is not possible, at least not while at the same time maintaining theoretical consistency. With respect to the issues discussed in this article, duality is not an alternative of dualism, but just more of the same. Whether this is true only for the two authors discussed (in the sense of episteme 2), or on a general plane (episteme 3), should be a point of further study.

New ontologies. This leaves new (i.e. more recent than Descartes), explicitly non-Cartesian ontologies to be considered. Before I go on to examine the metaphysics of Heidegger and Whitehead, a caveat is necessary. Both theories, claiming nothing less than reforming our most basic and time-honored notions, are very complex and difficult in their arguments and terminology. It is, of course, impossible to render them adequately within the scope of this article. In particular, I cannot provide all the reasons and chains of arguments given for the presented propositions, and must ask the interested reader to look them up in the originals.

Quite interestingly – and coincidentally as far as my selection was concerned – both philosophies date from the 1920s, but it seems that the two authors took no notice of one another (Cooper, 1993). In the course of time, their theories have met a very different fate. Heidegger’s book “Being and Time” was praised as most influential even before it was published. It is still considered by many, among them Gadamer, Foucault and Derrida, one of the most important philosophical books of the century – while others, like Carnap (1931) and Adorno (1973) believe it utterly absurd hocus-pocus. Whitehead, on the other hand, somehow never made it to the limelight. Although he was the teacher of Bertrand Russell and a famous mathematician in his own right, whose later career as philosopher was crowned by being offered a professorship at Harvard at the age of 63 (!), there was no broad reception of his metaphysics. However, neither was he ever completely forgotten, probably because his philosophy stresses the primacy of relations and processes, an idea most welcome in the last two decades and the reason why I discuss him. Not surprisingly, some constructivists, for example Prigogine, take their departure from Whitehead (Rust, 1987).

Both authors have aimed their philosophies explicitly against traditional metaphysics, especially against the notion of substances as self-sufficient, ultimate building blocks of the universe. In traditional metaphysics – and in everyday language – it is those substances that we refer to when we say, for example, that John has grown up. Although we recognize that John as a man is very different from John as a boy in many respects, we would argue that he – the same he, i.e. John – has changed in his attributes (e.g. height, weight, intelligence), but has retained something which makes it possible for us to still think of him as the same John. This unchanging something, in classical metaphysics, is his substance. This substance makes John an entity, it guarantees his continued being through all the changes life may impose on him. This substance is, by definition, self-sufficient, which means that it is not derivative or dependent on something else for its (metaphysical) existence. Although it will, of course, need food etc. from outside to sustain its biological existence, we do not need another entity to conceive of it. An example for a derivative existence would be a proposition, which depends on words or concepts. With the triumph of natural science from the 17th century onwards, particles of matter came to be regarded the ultimate substances which everything else was composed of. This position, extended by the class of “spiritual” substances, also lies at the heart of Descartes’ dichotomy of res extensa vs res cogitans, where “res” is just another name for substance.

Whitehead and Heidegger have two major objections against this position: First, they do not accept the notion that the ultimate building blocks are static, unchanging, and ahistoric. Second, they do not accept their self-sufficiency as self-sufficiency leads to the idea of an isolated, context-free “thing”. For these reasons, their points of departure are notions that are temporal, historical, situated, relational, and – most difficult of all – not a thing or object in the traditional sense. Heidegger’s Dasein as well as Whitehead’s Actual Occasions, despite the fact that they are nouns, represent rather an activity or process than a solid piece of stuff.

For all these commonalities, it should not be forgotten that the two projects also differ considerably regarding some important criteria (Rapp, 1984). As to their ultimate purpose, Whitehead aimed at a cosmology. For him, events and processes in nature form the core of the argument, while human culture and history are derived from them. Heidegger, on the other hand, wanted to inquire into the existential situation of human being. Human being, for him, is categorically different from physical objects, and nature and history are only to be found as modes derived from human existence. With regard to method, Heidegger argues hermeneutically, while Whitehead seeks to apply a axiomatic-deductive method.

Heidegger’s Dasein. Within his fundamental analysis of being, Heidegger (1977; see also Dreyfus, 1993; Mulhall, 1996) counters “Cartesian” dualism with a conception of the human being as Dasein (literally “to-be-there”, German for everyday general human existence). Although Heidegger uses it as a noun, Dasein is best understood as an activity or focus for a series of activities. These activities spring from an ineradicable concern for the world in which it lives, and from its being-in-the-world, i.e. the fact that Dasein is always a part of the world that cannot be conceived in isolation. Activities always happen in a situation characterized by locations, tools, and aims. It is impossible to conceive, for example, of writing without a pen or keyboard. Moreover, activities have a temporal connotation, stretching from past knowledge of tools and situations to the projection of a desired result in the future. Hence, Dasein is always temporal in the sense of drawing on past, present and future, and always in the world and with others. Most of the time, Dasein even more or less dissolves in the world acting in the mode of “the One” (das Man), as used like in expressions like “one should do this” or “one goes there to be seen”, thus giving up any unique characteristics.

Social structures, too, enter Heidegger’s world through the sphere of activity. In its most basic everyday form, Dasein, without reflection, uses equipment in order to live its life. Equipment has its purposes. Those purposes form a system of “for-the-sake-of-whiches”. Now, this system is social in two respects. First, it includes other persons in the form of producers of equipment, suppliers or customers (without, of course, the economic connotation). In other words, we do not work for us alone and by us alone. Second, the purposes are not private, and a tool is a tool for anyone. Thus, the purposes have not been invented by a single Dasein, but are handed down and point to the existence of others in the world.

Whitehead’s Actual Occasions. Like Heidegger, Whitehead (1985; 1993, see also Irvine, 2002; Sherburne, 1988; Wenzel, 1990) sets out to criticize the Cartesian heritage concerning the duality of mind and matter, but comes to different conclusions14. For him, mind and matter are only abstractions developed out of our perceptions of concrete things. Taking them as the ultimate concrete building blocks, as classical metaphysics does, means committing the “fallacy of misplaced concreteness” (he is, in fact, the inventor of this expression). The concrete “things” in Whitehead’s philosophy are, however, not things at all but bundles of relations. Whitehead’s universe resembles a web of relations, and the nodes in the web are formed by Actual Occasions. An Actual Occasion consists of all the possible relations it has with other Actual Occasions. Hence it is not a substance in the classical sense but rather a process of becoming as the web of relations changes continually. From this, it is obvious that Actual Occasions are no isolated items but microcosms placed in a world whose structures they carry “inside” them. A second characteristic that distinguishes them from classical substances is the fact that they cannot be located in a simple space-time coordinate as this, again, would imply that they are closed, self-sufficient, isolated entities. On the contrary, Actual Occasions are also potentials, and as potentials they are in many places at once. As they relate to, or as Whitehead says “ingress into”, other Actual Occasions, their presence is not only limited to one place in time. Today’s booking transaction is tomorrow’s corporate governance scandal; a decision made at the White House seals fates in Iraq or Israel. Third, they are also temporal as present Actual Occasions in the course of their becoming relate to already finished Actual Occasions and use them for their own constitution, thus quite literally taking up the past. Moreover, Actual Occasions resemble15 organisms in that they have a future state of completion at which their becoming is aimed. They thus stretch out over past, present and future in a way more reminiscent of events, processes or activities than of things.

The four problems . Now, as far as our metaphysical problems are concerned, both Heidegger’s and Whitehead’s ontologies have a head start concerning the problems of order and transcendence since they do not assume an individual agent, who almost by definition stands against society as a whole. Both Dasein and the Actual Occasions are part and parcel of their respective worlds. In a sense, they are those worlds before they become something that can be distinguished from them. Hence, the problem of how to overcome the gap from individual mind to collective order or from individual life to collective duration does as such not arise.

The problem of empirics is one both authors have addressed explicitly and in a rather assertive manner by accusing traditional empirics of observational sloppiness. Heidegger applies the phenomenological method developed by Husserl, who claims that traditional science has not taken a close enough view at the objects of study but has allowed its assumptions and theories come between them and their observations. His motto “Zu den Sachen selbst!“ (to the things themselves) expresses his claim that only phenomenological observation is rigid enough to provide an objective empirical description of the objects as they are. Whitehead also bases his methodology on an attack of traditional empirical science. He claims that in its search for the simplest building blocks, empirical science has inductively formed abstractions from the collected data, but has then taken these abstractions to be the ultimate concrete stuff the universe is made of. This is the fallacy of misplaced concreteness. Whitehead’s solution is a sort of naïve realism, in which he more or less accepts the things as they present themselves to the observer. A very immediate consequence of this is that colors, tastes, and smells return to the micro level as real givens, not as something constructed by the observer’s mind. Whitehead’s world is, even on the micro level, not “a dull affair, soundless, scentless, colourless; merely the hurrying of material, endlessly, meaninglessly” (Whitehead, 1993:161). It is not made up of “irreducible brute matter [... which is] senseless, valueless, purposeless” (1993:143). On the contrary, each Actual Occasion is capable of feelings and purposes, however dim and vague they may be. This is another effect from being a microcosm that is mirroring the whole: the world affects the Actual Occasion in each moment, and vice versa, the Actual Occasion is attuned to the world. This is what Whitehead calls “feelings”. Heidegger expresses the very same idea with his notion of Stimmung (mood, attunement) stipulating that Dasein is attuned at every moment. In conclusion, both ontologies address empirical issues although not in the way traditional methodologies do.

Whitehead’s critique of the notion of “irreducible brute matter [... which is] senseless, valueless, purposeless” also leads directly to his position on causality (Whitehead, 1985). He follows Hume’s critique stating that we cannot infer causality from observing a temporal sequence of events (the famous fallacy of post hoc ergo propter hoc, or correlation implies causation): Although we may observe that one billiard ball touches the next, and this then starts to roll as well, we cannot observe that the latter rolls because of being touched by the former. Hume concluded that causality was only a habit of thinking brought about by watching the repeated co-occurrence of events. Whitehead, however, although sharing the critique does not share this conclusion. Instead, he argues that the ultimate building blocks have to be conceived as endowed with values and purposes in order to make real causality possible. Causality, for Whitehead, is brought about when a past Actual Occasion (as “the cause”) is taken up to be part of a new Actual Occasion (as “the effect”). This ingression, as Whitehead calls it, is not in the first place mechanical or biological, but a process that involves both physical and mental aspects. To put it somewhat crudely, the new Actual Occasion decides which aspects of the old Actual Occasion to include and how to include them. This decision is dependent on the new Actual Occasion’s subjective aim. It thus does not reproduce the past (as it would when taking up the old Actual Occasion as it is) but reinterpret it (i.e. take certain aspects from it). Hence, causality always includes both efficient and final-teleological moments, even if it relates to a hammer driving in a nail. This may be less difficult to understand if we remember that Whitehead does not see a hammer and a nail in the first place, but rather a sequence of events of hammer-driving-nail. These events, then, can be more easily conceived of as having a purpose and a valuation concerning other events as promoting or impeding the fulfillment of this purpose. While Whitehead spends a lot of time on the discussion of causality, Heidegger does hardly mention it. If his standpoint on causality is to be constructed, it has to be taken into account that Heidegger’s point of departure lies in the study of human existence, and that he considers nature, including efficient causality, to be derivative of this. In this sense, causality, like the whole of nature, exists because Dasein interprets certain events as causally related. This is, however, not an idealist stance declaring that nature does only exist as long as there is a human mind observing (and producing) it. Rather, Heidegger would argue that without human existence, the question of whether nature exists does not make sense at all (Blattner, 2004). The difference is subtle, but important. Idealism says nature does not exist without the mind, while Heidegger says it neither exists nor does not exist, because in order to ask the question (or give the answer), we need a concept of existence that requires human interpretation. In other words, if we take away Dasein, we take away the possibility of interpretation and thus the meaning of concepts, which rests on interpretation. Without meaningful concepts, however, propositions, be they affirmative or negative, do not make sense. From this, we can infer that causality, even in its efficient form, always carries some interpretation, some intentionality with it.

The problem of metaphysics, finally, is, of course, the central one. There can be not question that both authors address metaphysical issues in extenso. Still, what makes a metaphysical system convincing? As Collingwood (1940) points out, basic assumptions cannot be transported by arguments but must agree with the reader in order to be believed. While this is certainly true, there are fortunately also some criteria that may be open to a less subjective discussion. Whitehead himself (Whitehead, 1985:3) demands a metaphysical system should be coherent, logical, applicable to the issues that concern us and adequate in the sense of being applicable to every instance we choose. Within the confines they have chosen for their theories, and taking into account the complexity that by necessity accompanies such general systems of thought, I think both ontologies pass the test. In my opinion, which follows Heidegger’s supporters rather than his critics, he argues his case, with a few exceptions, stringently and consistently. He defines his unusual terminology. He makes few assumptions that I find hard to accept. However, “Being and Time” is certainly not a book to be thumbed through and understood at once. Furthermore, it needs elaboration by others on the issues concerning social theory. It may, as a result, have other problems, but it can solve the ones I have discussed. As with Heidegger, Whitehead’s metaphysics certainly is nothing for a ten-minute lecture (or a couple of pages in a paper) but it is well worth looking into and discussing under the premises I laid out in this article. Whitehead’s style, with his definition of axioms and categories from which he infers and explains the whole system, betrays his mathematical background. His reasoning is complex, but always to the point, and his applications are convincing. As he is mainly concerned with natural processes, the social science side, again, certainly needs some more exploration into. Still, I hope to have shown that it is worth the endeavor.

Isolated ideas . What happened to these philosophies in the social sciences? Being such a prominent author, Heidegger’s influence has also extended to social theory, sociology and organization studies, where, however, only isolated concepts from his philosophy have been imported. Most important, none of the more influential authors ever adopted his metaphysics. Derrida avoids or deconstructs metaphysical claims; Foucault constitutes his very special (and, for our concern, still rather structuralist) subject. Phenomenological sociology was ever closer to Schütz than to Heidegger. And Giddens, perhaps the most prominent author indebted to Heidegger, as we have seen, borrows heavily in many places, but stops short of accepting Heidegger’s metaphysics – even at the cost of theoretical inconsistency. In consequence, it is fairer to say that some of Heidegger’s ideas have been modified to fit the old Cartesian metaphysics than to revolutionize the foundations of the social sciences – as they were once supposed to do.

In Whitehead’s case, too, we meet the procedure of taking the ideas without taking the metaphysics from which they have arisen. Keynes, himself a student of Whitehead’s, develops some of his logical-mathematical ideas. Parsons, who worked with a close friend of Whitehead’s at Harvard, adopts his methodological concept of “analytical realism” in the beginning but progressively abandons it . Constructionist theory, perhaps currently the major recipient of Whiteheadian ideas, places the stress on processes, relations and becoming, and takes act-supplement-relations as “moving production sites” for reality construction (Hosking, 2002) – very close to Whitehead’s actual entities – but explicitly denies a sharp distinction between epistemology and metaphysics. “It is on the basis of epistemological processes that individual and social phenomena obtain ontology, that is, are interpreted as real or as having a particular meaning” (Dachler & Hosking, 1995:1). This is not at all what Whitehead, coming from a natural sciences tradition, had in mind. For him (Whitehead, 1967), the “stuff” the world is made of differs from the Cartesian res cogitans and res extensa, but it is still ontological “stuff” which exists before all human interpretation. Persons and societies differ in degree, but not in principle from stones and atoms.

One can only speculate as to the reasons for this strategic neglect. One reason which springs to mind immediately is, of course, the inconvenience associated with reconceptualizing so many well-loved ideas and ways of thinking that would come with a new metaphysics. Many theorists, from Luhmann to Derrida, have found language less than cooperative in formulating new ideas. However, I believe that there is a second reason much closer home in the social sciences. It lies in the importance of conducting empirical research. Empirical research in its classic form is a child born of the Cartesian distinction, and it cannot be expected to produce meaningful results if applied to another metaphysics (see also Bohman, 1994; Gergen, 1999). Bearing this in mind, it becomes clearer why, for example, Giddens (1984) risks inconsistency proposing the (discursive) investigation to (non-discursive) practical consciousness: he has to because otherwise he would have to give up empirical research in the traditional way.


Adorno, T. W. 1973. The Jargon of Authenticity. Evanston: Northwester University Press.

Archer, M . 1995. Realist Social Theory: the Morphogenetic Approach . Cambridge: Cambridge University Press.

Barley, S. R. & Tolbert, P. S. 1997. Institutionalization and Structuration: Studying the Links between Action and Institution. Organization Studies , 18(1): 93-117.

Beckermann, A . 1999. Analytische Einführung in die Philosophie des Geistes. Berlin: de Gruyter.

Blattner, W. 2004. Heidegger's Kantian Idealism Revisited. Inquiry, 47: 321-337.

Bohman, J. 1994. New Philosophy of Social Science. Cambridge: Polity Press.

Bourdieu, P. 1984. Distinction. Cambridge/Mass.: Harvard University Press.

Bourdieu, P . 1990. The Logic of Practice. Stanford/Cal.: University Press.

Bourdieu, P., Chamboredon, J.-C., & Passeron, J.-C . 1991. The Craft of Sociology . Berlin: de Gruyter.

Brook, A., & Stainton, R. J. 2001. Knowledge and Mind. Cambridge/Mass.: MIT Press.

Burns, T. & Flam, H. 1987. The Shaping of Social Organization. London: Sage.

Burrell, G., & Morgan, G. 1979. Sociological Paradigms and Organizational Analysis. London: Sage.

Campbell, C. 1998. The Myth of Social Action. Cambridge: Cambridge University Press.

Carnap, R . 1931. Überwindung der Metaphysik durch logische Analyse der Sprache. Erkenntnis , 2.

Cohen, I. J. 1997. Theories of action and praxis. In B. Turner (Ed.), The Blackwell Companion to Social Theory: 111-142. Oxford: Blackwell.

Collingwood, R. G. 1940. An Essay on Metaphysics. In R. G. Collingwood (Ed.), Philosophical Essays, Vol. 2. Oxford: Clarendon Press.

Cooper, R. 1993. Heidegger and Whitehead: A Phenomenological Examination into the Intelligibility of Experience. Athens: Ohio University Press.

Crespi, F. 1994. Hermeneutics and the theory of social action. In Sztompka, P. (Ed.), Agency and Structure. Langhorne: Gordon & Breach.

Cyert, R. M., & March, J. G. 1992. Behavioral Theory of the Firm. Cambridge/Mass.: Blackwell.

Dachler, H. P., & Hosking, D. M. 1995. The primacy of Relations in socially constructing organizational realities. In: D. M. Hosking, H. P. Dachler, & K. J. Gergen (Eds.). Management and Organisation: Relational Perspectives . Avebury: Ashgate.

Dreyfus, H . 1993. Being-in-the-World: A Commentary on Heidegger’s „Being and Time“, Division I . Cambridge ( Mass.): MIT Press.

Elias, N. 1999. Die Gesellschaft der Individuen. In N. Elias (Ed), Die Gesellschaft der Individuen: 17-98. Frankfurt/Main: Suhrkamp.

Elster, J. 1986. Rational Choice. New York: New York University Press.

Foucault, M. 1977. Discipline and Punish. New York: Pantheon.

Foucault, M. 1978. The History of Sexuality, Vol. 1: An Introduction. New York: Pantheon.

Fuchs, S. 2001. Beyond Agency. Sociological Theory, 19(1): 24-40.

Garfinkel, H. 1996. Studies in Ethnomethodology. Cambridge: Polity Press.

Gergen, K. J . 1994. Toward Transformation in Social Knowledge . London: Sage.

Gergen, K. J. 1999. An Invitation to Social Construction. London: Sage.

Giddens, A . 1979. Central Problems in Social Theory. Houndsmills: MacMillan.

Giddens, A . 1984. The Constitution of Society. Cambridge: Polity Press.

Giddens, A . 1987. Time and Social Organization. In A. Giddens (Ed), Social Theory and Modern Sociology: 140-165. Cambridge: Polity Press.

Giddens, A. 1992. Modernity and Self-Identity. Cambridge: Polity Press.

Goffman, E. 1959. The Presentation of Self in Everyday Life. Garden City/N.Y.: Doubleday.

Grawe, C. 1972. Doppelnatur. In J. Ritter, & K. Gründer (Eds), Historisches Wörterbuch der Philosophie: 282-285. Darmstadt: Wissenschaftliche Buchgesellschaft.

Harré, R. 1997. Fitting the Body to the Mind. American Behavioral Scientist, 40(6): 798-812.

Heidegger, M . 1977. Sein und Zeit. Tübingen: Max Niemeyer.

Hoffmann, F., Trappe, T., Halbfass, W., Grünewald, M., & Abel, G. . 1992. Realismus. In J. Ritter, & K. Gründer (Eds), Historisches Wörterbuch der Philosophie: 148-178. Darmstadt: Wissenschaftliche Buchgesellschaft.

Hosking, D.-M. 2002. Constructing Changes. dian_marie_hosking/ oratie.html: 14.11.02.

Hosking, D.-M., Dachler, H.-P., & Gergen, K. . 1995 . Management and Organisation: Relational Alternatives to Individualism. Aldershot: Avebury.

Irvine , A.D. 2002 . Alfred North Whitehead. Stanford Encyclopedia of Philosophy, 02.06.02.

Kim, Jaegwon . 1996. Philosophy of Mind. Boulder , CO: Westview Press.

Knebel, S.K. 1998. Subekt/Objekt. In J. Ritter, & K. Gründer (Eds), Historisches Wörterbuch der Philosophie : 401-407. Darmstadt: Wissenschaftliche Buchgesellschaft.

Layder, D . 1981. Structure, Interaction and Social Theory. London: Routledge & Kegan.

Lévi-Strauss, C. 1968. Structural Anthropology . Allen Lane: Penguin.

López, J., & Scott, J . 2000. Social Structure. Buckingham: Open University Press.

Luhmann, N. 1982. Autopoiesis, Handlung und kommunikative Verständigung. Zeitschrift für Soziologie, 11(4): 366-379.

Luhmann, N. 1995. Social Systems. Palo Alto: Stanford University Press.

March, J.G. & Olsen, J.P. 1975. The Uncertainty of the Past: Organizational Learning under Ambiguity. European Journal of Political Research, 3: 147-171.

McKelvey, B. 1997. Quasi-natural Organization Science. Organization Science, 8(4): 352-380.

Moravia , S. 1995. The Enigma of the Mind. Cambridge: Cambridge University Press.

Mulhall, S . 1996. Heidegger and "Being and Time". London: Routledge.

Parsons, T., & Bales, R. F. 1953. The Dimensions of Action Space. In T. Parsons, R.F. Bales, & E.A. Shils (Eds), Working Papers in the Theory of Action: 63-109. New York: Free Press.

Popper, K . 1992. The Logic of Scientific Discovery. Routledge: London.

Prendergast, C., & Knottnerus, J. D . 1994. Recent Developments in the Theory of Social Structure. Current Perspectives in Social Theory, Suppl. 1: 1-26.

Quine, W. V.O. 1969. Natural Kinds. In W.V.O. Quine, Ontological Relativity and Other Essays: 114-138. New York: Columbia University Press.

Rapp, F. 1984. Zur Geschlossenheit metaphysischer Konzeptionen: Whitehead, Hegel und Heidegger. In H. Holz, & E. Wolf-Gazo (Eds.), Whitehead und der Prozeßbegriff: 424-443. Freiburg/München: Karl Alber.

Reed, M . 1997. In Praise of Duality and Dualism. Organization Studies, 18(1): 21-42.

Ritzer, G. 1996. Modern Sociological Theory. New York: McGraw-Hill.

Rust, A . 1987. Die organismische Kosmologie von Alfred N. Whitehead. Frankfurt/Main: Athenäum.

Schatzki, T. 1997. Practices and Actions. Philosophy of the Social Sciences, 27(3): 283-308.

Scott, R. 1987. Institutions and Organizations. Thousand Oaks: Sage.

Sherburne, D . 1988. A Key to Whitehead's Process and Reality. Chicago: University of Chicago Press.

Shilling, C. 1997. The Undersocialised Conception of the Embodied Agent in Modern Sociology. Sociology, 31(4): 737-754.

Shilling, C. 1999. Towards an Embodied Understanding of the Structure/Agency Relationship. British Journal of Sociology, 50(1): 563-562.

Smelser, N. 1989. Social Structure. In N. Smelser (Ed.), The Handbook of Sociology: 103-129. Newbury Park: Sage.

Specht, R . 1986. Descartes und Spinoza. Unpublished Lecture, University of Mannheim, Mannheim.

Weber, M. 1979. Economy and Society. Berkeley: University of California Press.

Weber, M. 1988. Die „Objektivität“ sozialwissenschaftlicher und sozialpolitischer Erkenntnis. In M. Weber (Ed.), Gesammelte Aufsätze zur Wissenschaftslehre. Tübingen: Mohr: 146-214.

Weick, K. E. 1969. The Social Psychology of Organizing . Reading/Mass.: Addison-Wesley.

Wenzel, H . 1990. Die Ordnung des Handelns. Frankfurt/Main: Suhrkamp.

Whitehead, A.N. 1959. Symbolism. New York: Capricorn.

Whitehead, A.N. 1967. Adventures of Ideas. New York: Free Press.

Whitehead, A. N . 1985. Process and Reality . New York: Free Press.

Whitehead, A. N . 1993. Science and the Modern World. In M. Adler (Ed), Great Books of the Western World: 135-234. Chicago: Encyclopedia Britannica.


1 In this article, I use “ontology” as a sub-concept to “metaphysics”. Although the former is a far more popular term in the social sciences, its subject matter is traditionally limited to the study of the most general attributes of being – something I do not inquire into. For the kind of study I present here, I think the term “metaphysics” is more appropriate.

2 This does not imply that I consider the discipline of philosophy prior to or more fundamental than the social sciences. I agree with Quine (1969) that both must co-evolve and share their insights. However, just as the current laws of physics limit what can be said in chemistry, so do ontology, epistemology, or logic for the social (and natural) sciences in general.

3 This is important because it shows that the Cartesian division is not overcome by stating that the agent is socialized, constrained by society or even a Foucaultian product of it. To put it differently, a res cogitans position is only concerned with the “stuff” the agent is made of, not with how this stuff is formed. Hence it implies neither that the agent is free and self-determined nor that he/she is not.

4 Dachler and Hosking (1995) characterize this as a “possessive individualism” prevalent in the management and organization literature. Harré (1997) seems to suggest a similar state for the social psychology literature before the “second cognitive revolution”. Within sociology, both definitions of agent and agency suffer from a very unequal treatment of body and mind resulting in the “disembodiment” of agent and agency or at least an undersocialized conception of the embodied agent (Shilling, 1997 and 1999).

5 Let me do justice to poor René, whose “Cartesian dualism” has been blamed for almost everything that went wrong in modern organization theory, from rationalism to representationalism: he simply did not do it. Although he made it popular, the division between res cogitans and res extensa is much older and goes back to Late Antiquity (Grawe, 1972; Specht, 1986), while the division between subject and object in contemporary usage was developed by as yet unidentified authors after Descartes. Descartes himself uses “subject” still in its pre-modern, medieval sense (Knebel, 1998), which inverts the modern meaning (or rather, vice versa). “Subiectum” here is that which underlies all perception, hence the (modern) object of an inquiry (still traceable in the English “subject” of a conversation).

6 The term “perceived” here indicates that it is merely an appearance, and that the perceiver does not claim to have discovered patterns external to him or her. The antonym, in this case, is “external”, which means independent of perception.

7 The logically correct classification would be “substance dualism” and “substance physicalism” (see Beckermann, 1999) instead of idealism and materialism, but I have opted for the more popular terms as the definitions do not differ significantly for our purposes. The underlying question of whether the mind is material and thus subject to physical laws is also discussed, albeit with different labels, in Brook & Stainton (2001) or Moravia (1995).

8 Contrary to intuition, Plato must be considered a realist ME as he assumed that ideas had a real existence independent of the human mind.

9 Archer (1995) and Shilling (1997) even criticize structuration theory for overemphasizing the actor’s cognitive skills.

10 Although he locates structure once (1984:25) as “out of time and space”, he does not elaborate this statement further. Instead he focuses entirely on the instantiations.

11 For a more detailed account of the inconsistencies regarding practice and practical consciousness in Giddens’s and Bourdieu’s theories see Schatzki (1997).

12 Interestingly, Giddens explicitly (1984:xxxvii, 1987:153) neglects what he calls “Parsons’ problem”, i.e. the problem of order.

13 His arguments on the relationship between theory and praxis (Bourdieu, 1990), while insightful in other ways, are rather shadowy as far as this problem is concerned.

14 It is, however, also interesting to see the parallels between the two authors (cf. Cooper, 1993).

15 Whitehead, in fact, calls them organisms, but does not refer to biological organisms. Instead, he aims at including notions of systemness (like order, dependence, emergence).