Tag Archive: Harman


I have been glad to see that both Levi and Bogost have replied to my post about Object-Oriented Empiricism. Both Levi and Bogost have correctly raised some concerns about using the word empiricism. I admit that I share some of their concerns and there are certain problems about using the label empiricism. Maybe it would have been better to classify Object-Oriented Empiricism as Object-Oriented Philosophy in Action, which would represent the praxis dimension of Object-Oriented Philosophy. However, there is something that I find invaluable about empiricism, which I will attempt to express in this post.

The caricature of the empiricist is the researcher that goes out to observe facts from the external world, and uses their senses within experience to observe the world. The empiricist is then a dry observer, passively gazing upon the world. Another problem of the empiricist is they tend to focus on the human-world gap, making (traditional) empiricism a form of correlationism. The result is that the empiricist concentrates on how ideas come from our impressions of the world, which are derived from our human experiences. For example, I have an idea of a cake because I previously smelt, tasted, touched, and seen the cake. Therefore, for the empiricist, we never experience the world directly.

However, there is something invaluable about empiricism that I would claim is connected to Object-Oriented Philosophy. A lot of Object-Oriented Philosophy is indebted to correlationist philosophy, which puts the correlate of human and world at the centre of their concerns. For example, Shaviro’s Whiteheadian/Deleuzian realism is indebted to Kant, and Harman’s  OOP is indebted to Heidegger, and it seems unproblematic to claim that both Kant and Heidegger are correlationists. The interesting move that both Shaviro and Harman make is that they radicalize Kant and Heidegger, suggesting what they claim about the human-world correlate is also relevant for the relations between objects. I would also like to propose that it is possible to radicalize empiricism, and suggest what (traditional) empiricism says about the human-world correlate is also relevant for conceptualising object-object relations.

At one level, empiricism is critical of representational realism, suggesting that we experience secondary qualities (i.e. sound, taste, smell…) and not the actual object. In addition, the secondary qualities experienced do not resemble the object. For example, the sweetness of a cake is a secondary quality that I would experience through my senses, and not a quality of the cake in-itself. However, empiricism makes a key proposition here that is relevant for object-oriented philosophy. Empiricism suggests that objects have the capacity to affect humans, even if the objects remain a mystery. In other words, the encounter with objects is a productive encounter with difference that produces the subject. The result is that the ‘outside’ world has objects that have the capacity to affect humans, even if humans do not represent those experiences.

Yet, at this level empiricism remains ‘human, all too human’ and needs to be radicalized in order for empiricism to be relevant for OOP. What if we suggest that what empiricism claims about the human-world gap is also ‘true’ about other relations, which do not need to include the human-world relation. If empiricism claims that we need to account for how a human is formed from experience, then OOP needs to account how objects are formed from experience. Therefore, we cannot think of experience as experience of some subject and need to consider the inhuman, impersonal, or anonymous plane of experience. In addition, empiricism can inform us that encounters of experience are productive and non-representational. For example, the sunlight that encounters the plant has the capacity to affect the plant, but it would seem strange to suggest the plant is representing the sunlight. Instead, it would be more accurate to claim that different objects have the capacity to affect other objects, and the multiple relations between objects are productive, which have the capacity to produce emergent properties.

Hopefully, OOP has some potential to be empiricist, and can illustrate how experience is not an exclusive plane reducible to human experience, which can help to produce a non-anthropocentric empiricism. Therefore, there is a rich field of experience, where objects act upon other objects, and humans are only one node within this plane.

As I have noted, my recent attention has been concerned with the relevance of Object-Oriented Philosophy (herein after OOP) for political analysis (here). It recently struck me that OOP needs to put forward a program of what I will term Object-Oriented Empiricism (herein after OOE), which is  the actual practice of Object-Oriented Studies in action.

At one level I want to differentiate between the theory (or philosophy) of OOP and the praxis of OOP, which will be designated as OOE. The former (OOP) will primarily be engaged in the philosophical discussion and theoretical debates of an object-orient approach, and the main role of OOP will be to produce Object-Oriented Ontologies. The latter (OOE) will primarily be concerned with illustrating the benefits (and limitations) of Object-Oriented Ontologies for the analysis of the experiences of the ‘real’ world, aiming to research particular objects(or events) and how these objects act and relate to other objects. In other words, the Object-Oriented Empiricist will use (or steal) the ontologies produced in OOP and design  their research projects in accordance with what object-oriented ontology they adopt.

I feel there seems a need for OOP to move towards the stage of OOE. The rich work of Byrant, Harman, and Shaviro has reached a stage where there is enough theoretical discussion to move towards the empirical analysis of objects. Of course, OOE will have to be familiar with the different ontologies of OOP, the debates within OOP, and the consequences of adopting one ontology of OOP over another. For example, a debate within OOP is if an Object-Oriented Ontology needs a virtual dimension, or if Object-Oriented Ontology is purely actual. However, it is time for OOP to develop into OOE, which can show the praxis of OOP.

At present, I envision that OOE will adopt two principles from OOP (and speculative realism)

  1. Critical of Correlationism
  2. Against the Hegemonic Fallacy (and here)

In terms of the first principle, OOE will have to illustrate why research methods that concentrate on the correlate (or relation) between the humans and the world is in itself a limitation. The second principle will critique any research that argues that one difference makes all the difference. The result of the second principle is that OOE research will be unable to claim that analysis is reducible to one phenomenon (e.g. language or atoms). Therefore, OOE will argue against all forms of reductionism that are evident in a lot of other paradigmatic approaches. I envision that it is the second principle where OOE will have to empirically challenge a lot of other engrained approaches. The (hopeful) result of OOE will be that the research it produces will demonstrate that the ‘real’ is not reducible to social norms (social constructivism), pursuits of self-interest (political realism), discursive fields (discourse analysis), autonomous individuals (liberal individualism), and so on. The benefit is that OOE will be able to expose the limitations of other approaches from illustrating how they pertain to the hegemonic fallacy and reduce the ‘real’ to one phenomenon.

*[Update, Levi has written a recent post that is relevant for this post, discussing how objects act and the actual and virtual dimensions of objects, read here]

I have recently been thinking about what political praxis object-oriented philosophy (herein after OOP) can offer political analysis. In conjunction, I have been thinking about the differences between how an object is defined from reading Levi Byrant, Graham Harman, and Steven Shaviro. In the following post I want to focus on Shaviro’s reading of Whitehead’s “eternal objects’ in his book Without Criteria and his post (here). The aim is to defend an OOP that argues objects are both actual and virtual.

One of the main debates within OOP is between actualists (Harman) and virtualists (Byrant & Shaviro). My reading of the debate concerns if an object-oriented ontology requires a virtual dimension. Criticism of OOP arguing for a virtual dimension seems to put forward the argument that virtualists limit an object to what is possible. However, it seems that this confuses the difference between possible and potential. For an object-oriented philosopher of the virtual, the virtual is a crucial (and real) ontological dimension that enables one to contemplate how an object is not reducible or exhausted by its actual existence. In other words, any ontology that includes a virtual dimension attempts to account for potentiality.

Following Deleuze, Shaviro asserts that we should not differentiate between the virtual and the real, ‘the virtual…is altogether real in its own right…it is just that this reality is not actual. The virtual is like a field of energies that have not yet been expended, or a reservoir of potentialities that have not yet been tapped’ (Without Criteria, p35). As a reservoir of potentials, there are processes of actualisation, when certain potentials of the virtual are actualised. However, actualisations of the virtual do not exhaust the potential of the virtual. Shaviro is aware that Whitehead never uses the word virtual, but does argue that Whitehead’s distinction between the actual and the potential resembles the Deleuzian distinction between the actual and the virtual (Without Criteria, p37). Therefore, for Whitehead the potential is a real (virtual) ontological dimension of the world that needs to be acknowledged. We cannot only focus on actual identities (although actualists will contend this point).

We may then ask what Whitehead claim composes the virtual/potential aspect of his ontology. One answer is that the virtual is a composition of eternal objects. As Shaviro notes, ‘alongside events or actual entities, Whitehead also posits what he calls “eternal objects.” These are “Pure Potentials.” Shaviro defines Whitehead’s eternal objects in the following:

“’any entity whose conceptual recognition does not involve a necessary reference to any definite actual entities of the temporal world is a called an ‘eternal object’ (Whitehead, Process and Reality). This means that eternal objects include sensory qualities, like colors (blueness or greenness) and tactile sensations (softness or roughness), conceptual abstractions like shapes (a helix, or a dodecahedron)…eternal objects are ideal abstractions that nevertheless (in contrast to Platonic forms) can only be encountered within experience, when they are ‘selected’ and ‘felt’ by particular actual occasions” (Without Criteria, p39)

At least two things are noteworthy about eternal objects. First eternal objects are virtual, but nevertheless real. Second, eternal objects are actualised in particular occasions, but are not reducible to these particular local actualisations. To understand why eternal objects are virtual and not reducible to their local manifestations in particular point in time, it is necessary to regard eternal objects as neutral and indifferent:

“’Eternal objects ‘involve in their own natures indecision’ and ‘indetermination’ (Whitehead, Process and Reality); they always imply alternatives, contingencies, situations that could have been otherwise. This patch of wall is yellow, but it might have been blue. This means that their role is essentially passive…You might say that yellowness  ‘in itself’, understood as a pure potentiality, is utterly indifferent to the actual yellow color on this particular patch of wall…Eternal objects, like Deleuze’s quasi-causes, are neutral, sterile, and inefficacious, as powerless as they are indifferent” (Without Criteria, p41)

There seems an important argument that eternal objects need to be thought of as neutral. At one level, the neutrality ought to mean that the virtuality of eternal objects is not reduced to local or particular manifestations. In other words, we need to be careful about equating the virtual with the actual. In terms of thinking ‘yellowness’ as both an eternal object with particular manifestations it seems unproblematic. Therefore, the book that sits in front of me with a yellow cover does not exhaust the potentiality of ‘yellowness’ as an eternal object. However, it is perhaps worth considering if the neutrality of eternal objects has any significance for an OOP political analysis. We can maybe draw the political significance of eternal objects from a (re)reading of Foucault’s analysis of the Panopticon in Discipline and Punish. [I have also selected this example to provocatively suggest Foucault has the potential to be read as a type of object-oriented philosopher].

It might seem odd to (re)read Foucault’s analysis of the Panopticon as an example of discussing an ‘eternal object.’ At one level, Foucault is examining the physical attributes and design of a particular building. Foucault’s fascination with Bentham’s Panopticon is that the design of the building allows an observer a fixed central position that makes it possible to gaze upon those people confined within the Panopticon. The result is that Foucault is able to consider how the Panopticon functions as a form of disciplinary power that produces the subjectivity of those confined. One of Foucault’s main findings is that as a form of imprisonment, the Panopticon differs from a dungeon as ‘full lighting and the eye of a supervisor capture better than darkness, which ultimately protected. Visibility is the trap’ (Foucault, Discipline and Punish, p200). However, Foucault is aware that merely focusing on the Panopticon as a means of imprisoning delinquents is in itself a limitation. We could suggest that this type of analysis of analysis is too empirical and remains at the level of actual objects. In other words, it fails to conceive of the Panopticon as a (Deleuzian) virtual idea or (Whiteheadian) eternal object. Indeed, Foucault is aware that the potentiality of the Panopticon is more than particular local manifestations of a Panopticon. Foucault writes that the Panopticon ‘is polyvalent in its applications; it serves to reform prisoners, but also to treat patients, to instruct schoolchildren, to confine the insane, to supervise workers, to put beggars into to work (Discipline and Punish, p205). In other words, the Panopticon as an eternal object is not reducible to those experiences that actualise a manifestation of panopticism, and Foucault warns that the Panopticon ‘is in fact a figure of political technology that may and must be detached from any specific use’ (Discipline and Punish, p205). In order to detach the Panopticon from any specific manifestation there requires the need to conceptualise as an ‘eternal object’ or what Deleuze and Guattari would refer to as an abstract machine. The effect is that a Panopticon is both virtual and actual.

One other issue needs addressed. Whitehead and Shaviro both understand eternal objects as ‘neutral’ and ‘indifferent’ to their manifestations in particular occasions, and it seems strange to claim that Panopticon – as an eternal object – is ‘neutral’ and ‘indifferent’. Indeed the Panopticon seems overtly political and not neutral. Yet such a criticism I would suggest would be too concerned with the particular and actual uses of panopticism in our actual experience. As an eternal object the Panopticon is indeed ‘neutral’ and ‘indifferent’, we need only think that particular manifestations of the Panopticon could be used from different and antagonistic political ideologies (Fascism, Liberalism, Socialism). Therefore, as an eternal object, panopticism is indifferent to particular manifestations and is not exhausted from these previous manifestations.

Speculative Heresy have posted a collection of 13 papers/articles from Graham Harman. One of the key thinkers in the emergent speculative realism paradigm. This collection looks an exciting read and I look forward to understanding Harman’s Object-Oriented philosophy.

The papers/articles can be accessed from here.

Harman on DeLanda

Here is a link to a recording of Graham Harman discussing Manuel DeLanda’s Assemblage Theory.

it is from a talk Graham gave at LSE comparing assemblage theory with Bruno Latour’s actor-network-theory.

Thanks to the anthem-group for posting the recording.