This article covers Michel Foucault's Apparatus and Karen Barad's Apparatus.

Michel Foucault's Apparatus

A translation of Foucault's original French term, //dispositif// (literally, "device"), "apparatus" refers to an "analytic of power" (198) that describes the relationship between disparate entities.


1. "a thoroughly heterogeneous ensemble consisting of discourses, institutions, architectural forms, regulatory decisions, laws, administrative measures, scientific statements, philosophical, moral and philanthropic propositions -- in short, the said as much as the unsaid. Such are the elements of the apparatus. The apparatus itself is the system of relations that can be established between these elements."(194) (bold emphasis added)

In this, Foucault's conceptualization of an apparatus is similar to the concepts of assemblage (1, 2) and infrastructure (1, 2) because it is a relational system for unlike elements or nodes. As an "analytic of power", an apparatus is useful for analysing Power/Knowledge relationships.

2. "what I am trying to identify in this apparatus is precisely the nature of the connection that can exist between these heterogeneous elements. Thus, a particular discourse can figure at one time as the programme of an institution, and at another it can function as a means of justifying or masking a practice which itself remains silent, or as a secondary re-interpretation of this practice, opening out for it a new field of rationality. In short, between these elements, whether discursive or non-discursive, there is a sort of interplay of shifts of position and modifications of function which can also vary very widely." (194-5)

It is important to note that the relational connections in this apparatus are not static; they can shift depending on situation or purpose. Again, this holds similarities to the looseness of connections between nodes of an assemblage (1, 2) that allow for "play" or "wiggle room."

Definitional Structure of an Apparatus

3. "I understand by the term 'apparatus' a sort of -- shall we say -- formation which has as its major function at a given historical moment that of responding to an urgent need. The apparatus thus has a dominant strategic function." (195)

As noted by Wajeman, an apparatus is defined not only by its structuring of heterogeneous elements, but also by particulars of its creation -- its genesis.(195) Foucault identifies two important moments: first, the "prevalent influence of a strategic objective," and second, "it is the site of a double process." (195) These two functions are "functional overdetermination" and "strategic elaboration."

"Functional overdetermination" refers to the interactivity between effects ("positive or negative, intentional or unintentional") as they experience constructive or destructive interaction/interference. This can result in a need to adjust or rework the connections between elements. (195)

"Strategic elaboration" is a perpetual process whereby an unforeseen effect results from the re-utilisation of an apparatus. For example, the apparatus of imprisonment, besides coming to be seen as the best (efficient and rational) method for dealing with crime, also produced a "delinquent milieu" that was different from conceptualizations of illegal activities up until then. This happened because "[t]he prison operated as a process of filtering, concentrating, professionalising and circumscribing a criminal milieu." (196) According to Foucault, from the 1830s onward, the negative effect of imprisonment was able to be re-utilised to glean a positive benefit for political and economic ends, "such as the extraction of profit from pleasure through the organisation of prostitution." This was also called the "strategic completion" (remplissement) of the apparatus. (196)

Similarities to autopoiesis could be drawn if the apparatus is regarded as complex enough to exercise a level of autonomy in the production of unforeseen effects.

Contrast with "episteme":

The term "apparatus" is contrasted with "episteme" ("knowledge and discursive formations") by Foucault when he emphasizes that episteme is a "specifically discursive apparatus, whereas the apparatus in its general form is both discursive and non-discursive, its elements being much more heterogeneous." (197) Thus, an episteme is a type of apparatus, and the apparatus is the generalized category of an episteme.

Foucault had clarified the definition of episteme in light of his interview on "apparatus" (197) (bold emphasis added):

J.-A. Miller: With the introduction of 'apparatuses', you want to get beyond discourse. But these new ensembles, which articulate together so many different elements, remain nonetheless signifying ensembles. I can't quite see how you could be getting at a 'non-discursive' domain.

Foucault: In trying to identify an apparatus, I look for the elements which participate in a rationality, a given form of co-ordination, except that . . .
J.-A. Miller: One shouldn't say rationality, or we would be back with the episteme again.

Foucalt: If you like, I would define the episteme retrospectively as the strategic apparatus which permits of separating out from among all the statements which are possible those that will be acceptable within, I won't say a scientific theory, but a field of scientificity, and which it is possible to say are true or false. The episteme is the 'apparatus' which makes possible the separation, not of the true from the false, but of what may from what may not be characterised as scientific.

Although in the above exchange there is a focus on "discursive" versus "non-discursive" apparatuses, it does not appear to be Foucault's intention to create subcategories of apparatuses as such. It appears that Foucault intended to be inclusive in his definition of "apparatus" so as to not categorically exclude relational structures that were not explicitly language- or discourse-based (197) (bold emphasis added):

Le Gaufey: But going back to this question of the 'non-discursive', what is there in an apparatus, over and above the discursive utterances, except the 'institutions'?

Foucalt: The term 'institution' is generally applied to every kind of more-or-less constrained, learned behaviour. Everything which functions in a society as a system of constraint and which isn't an utterance, in short, all the field of the non-discursive social, is an institution.

J.-A. Miller: But clearly the institution is itself discursive.

Foucalt: Yes, if you like, but it doesn't much matter for my notion of the apparatus to be able to say that this is discursive and that isn't. If you take Gabriel's architectural plan for the Military School together with the actual construction of the School, how is one to say what is discursive and what institutional? That would only interest me if the building didn't conform with the plan. But I don't think it's very important to be able to make that distinction, given that my problem isn't a linguistic one.

Karen Barad's Apparatus

Informed by Niels Bohr's work in quantum physics and quantum entanglement, Karen Barad adapts Foucault's apparatus as a key element in the technoscientific production of knowledge and reality through agential realism.

Starting with several critiques of Foucault's discussion of apparatus, namely that Foucault did not better develop the ways that apparatuses and subject/objects mutually create and define each other nor clearly articulate the nature of the relationships between practices and phenomena, Barad moves to provide a more complete definition of apparatus in use:

"Foucault's insights concerning disciplinary practice and the 'microphysics of power' have profoundly altered the ways in which power and knowledge are currently theorized. However there are crucial features of power-knowledge practices that Foucault does not articulate, including the precise nature of the relationship between discursive practices and material phenomena; a dynamic and agential conception of materiality that takes account of the materialization of all bodies (nonhuman as well as human and that makes possible a genealogy of the practices through which these distinctions are made); and the ways in which contemporary technoscientific practices provide for much more intimate, pervasive, and profound reconfigurings of bodies, power, knowledge, and their linkage than anticipated by Foucault's notion of biopower (which might have been adequate to eighteenth-century practices, but not contemporary ones)." (200)

"Although Foucault insists that objects (subjects) of knowledge do not preexist but emerge only within discursive practices, he does not explicitly analyze the inseparability of apparatuses and the objects (subjects). In other words, Foucault does not propose an analogue to the notion of phenomenon or analyze its important (ontological as well as epistemological) consequences." (201)

In Barad's conceptualization, apparatuses are “not passive observing instruments” but are “productive of (and part of) phenomena.” (98) "Apparatuses are themselves material-discursive phenomena, materializing in intra-action with other material-discursive apparatuses." (102)

“the transducer does not allow us to peer innocently at the fetus, nor does it simply offer constraints on what we can see; rather, it helps produce and is 'part of' the body it images.” (101)

In other words, the technoscientific apparatus not only is proof of material transformation and utilization, but the phenomena it observes also get transformed by being observed through an apparatus.


Barad, Karen. Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and Meaning. Durham: Duke University Press, 2007.
Barad, Karen. “Getting Real: Technoscientific Practices and the Materialization of Reality” Differences, 1998: 10(2), 87-128.
Foucault, M. Ed. Gordon, C. "The Confession of the Flesh" in Power/Knowledge: Selected Interviews & Other Writings 1972-1977. New York: Pantheon Books, 1980.

Further Reading:

- Pløger, John. "Foucault's Dispositif and the City" Planning Theory. 2008(7): 51.