Explanatory Networks and Political Reason
There has recently been an interesting (and somewhat turbulent) discussion regarding Latour’s Actor-Network Theory (ANT) and the Object-Oriented Ontologies (OOO) that are influenced by it, in relation to the kind of politics these theories can support.
There is obviously Nick Srnicek’s very interesting piece from the Militant Dysphoria conference (available here), which tries to show how ANT provides some useful resources for reconsidering the nature of political action, and his recent additional commentary on it (here), which situates this piece in relation to the notion of folk-politics (something I myself have talked about here, but with a slightly different twist).
Then there is the more fiery (though now thankfully cooled) exchange between Reid Kane at Planomenology (here and here) and Levi at Larval Subjects (here, here, here and here) over whether either Latour’s ANT or OOO has neo-liberal political implications. This obviously got out of hand, but it strikes me that the real intuition behind the argument that Reid was making (and that others have also been making), was never made fully explicit. Without wishing to blow on the embers, I feel that it would be helpful for this intuition (as I see it) to be properly formulated. This also gives me an opportunity to work out some other thoughts about Latour’s position which have been haunting me.
The proviso here is that I am neither an expert on Latour or on OOO, although I will admit to having read more of the former than the latter. So, it is possible that my reading, and the implications I draw from it, will be faulty. As ever, I am happy to be corrected. That being said, I will proceed anyway, while the point is fresh in my mind.
Edit: It is of course also important to note that there are different variants of OOO, and not all will endorse or take up the Latourian positions I’m trying to analyze here, at least in the way they are found here.
Firstly, it is important to reiterate a point that Levi made, and which I entirely agree: ontology is in no way beholden to politics.
That an ontology has unsettling political implications (for instance, by undermining the theistic justifications of the monarchical state) is simply something one has to deal with. The truth of Being is something that must be dealt with on its own terms, and all of those who espouse any of the variant realist positions popping up in the loosely continental tradition should be able to agree upon this. This is not to deny that the activity of doing ontology, and the ontologists who carry it out, are situated within some concrete political situation, and that their theoretical activity often has both political motivations and political implications. Rather, this is a properly normative claim, to the effect that we should endeavour to assess the truth of ontological claims and arguments independently of these motivations and implications. In practice, we might fail to live up to this norm, but we should act (and think) otherwise.
2. Latour and the Structure of Explanation
Now that this is out of the way, it seems that there are two crucial features of Latour’s work with relevance to the current issue. I will address these as features of Latour’s work as it is not always clear to me to what extent these are directly adopted or transmuted in their uptake by OOO. These features are as follows:-
1) The collapse of the distinction between might and right. We might also call this the reduction of normative force to causal force, although this characterisation might later be problematised.
2) The non-modernist elision of the distinction between nature and culture. This is meant to be opposed to modernist separation between these two domains.
I would venture that what motivates both moves is a certain concern with explanation, and in particular with the kinds of explanation that we should deploy within the social sciences, and most obviously, in the sociology of science in particular.
This is most obvious in the second move, as it is a response to the inability of modernist discourses to take into account the causal efficacy of natural entities in the explanation of social phenomena. For Latour, this is particularly evident in science, where he attempts to substitute in a subtle picture of the many ways that both the social structures of scientific research programs, myriad scientific instruments and the very things being studied all interact to produce competing scientific theories and the disputes between them.
However, it is also operative in the first move, insofar as Latour is suspicious of the introduction of the kinds of loosely platonic entities he takes would be necessary to ground normative force. There seems to be a (not entirely illegitimate) suspicion that if one were to allow for normative force, then the kinds of entities one would have to admit would thereby corrupt the structure of explanation. Such corrupt forms of explanation are to be found in Latour’s main target – the broadly modernist approach to culture and history. One might pick out Hegel and his inheritors as classic examples here, whose grand narratives about the dialectical development of the norms immanent within given societal structures and historical epochs have a tendency to efface the nitty gritty details of how it is that concrete states of affairs are brought about.
It seems that what Latour is aiming for with both of these principles is to enable a kind of explanation in the social sciences that is both sensitive to the specific, local factors making up any given situation (the virtue which Nick’s paper locates in ANT), but which also avoids falling into the trap of trying to explain cultural phenomena either entirely in terms of some internal cultural logic, or in terms of some set of classically ‘natural’ entities that excludes the cultural.
We could provide a simplified caricture of certain debates in historiography as an example here, namely, the debate over the role that ‘ideas’ play in historical development. On the one hand we have those who understand ideas as the driving force of History, and on the other those who denigrate the importance of ideas and posit other brutely materialistic forces as driving historical development. A Latourian response would be to show how ideas are one of many different interacting factors that determine historical development to differing extents in different situations, in tandem with the sort of brute material matters that they are held apart from in the above caricature. The question then becomes a local one of what role any given idea plays within the network it is situated within.
Having laid out these foundational Latourian positions, I will now proceed to examine some potential problems that I (and others) see with them.
3. The Force of Reasons
We will first turn ourselves to the first Latourian position we identified above, the reduction of normative force to (loosely) causal force. Tom at Grundlegung has already performed a very good analysis of this and its problems in relation to rationality (here and here), but I will endeavour to add something to what he has said already.
Firstly, it is important to recognise that there are different forms of normativity that this Latourian move undercuts. There is the fundamental normative structure of rationality as such, which provides the form of reasoning (both theoretical and practical) independently of the content of one’s premises (e.g., the law of non-contradiction) and then there are the constitutive norms which make up that content, which license certain inferences on their basis, and demand others. It is these constitutive norms which demand that we acknowledge certain inferences, and they thus provide the so called force of reasons. If you are committed to the fact that ‘It is Sunday today’ then you must admit that ‘It is Monday tomorrow’, not in the sense that you are being held at gun point, but because that is simply a result of the norms governing the use of those temporal concepts.
There is then an additional level of normativity, that of practical norms governing action. These can come in varying different forms, from personal maxims, through explicit societal laws, and implicit norms of social conduct, all the way to supposed fundamental ethical principles which govern how we should act in all circumstances. When we are engaged in debate over the ethical and political implications that a given ontological theory might have, or indeed over the resources it might provide for pursuing some ethical/political dictum, we are in the domain of practical norms of action.
Here then is the problem that the first Latourian move produces: it undermines both the force of the fundamental and constitutive norms required for reasoning as such, and it undermines the force of the practical norms that constitute the ethical/political domain. Let us take these two implications one at a time.
Firstly, there are two dimensions to reasoning: public and private, or dialogical and monological, respectively. As some have already pointed out, the Latourian move undermines the distinction between reasoned argument and persuasive rhetoric in the public sphere, because all that matters is what produces (or causes) one to come to the position one occupies, not anything to do with the normatively assessable correctness of this process. This makes all public communication into what Habermas would call a strategic matter. It is about how one makes one’s position come out on top, rather than how one justifies it in accordance with some independent set of norms. In accordance with Latour’s wider metaphysics, an argument is just a struggle for dominance between actors in a network, just as takes place in every network whatsoever. There is no special kind of dominance here.
Now, the importance of drawing the public/private distinction is that this strategic notion of rationality that is motivated by Latour’s metaphysics can only function if an individual is capable of some form of private practical reasoning. One must both have ends and have ways of analysing the various means of achieving them in order to engage in anything strategically, let alone communication. We are not of course implying that non-human actors engage in practical reasoning of this strategic kind. The networked power struggles that take place between various non-human objects are not reason involving in any way whatsoever, nor should they be. The crucial point is that if we are to honestly believe in the truth of the Latourian position, then the only sincere (and self-conscious) form of reasoning that is open to us is private practical reasoning.
Now, the second implication we worked out above is that practical norms of action have no force on the Latourian account. This means that if we are sincere Latourians, then such norms need play no role in our private practical reasoning – they cannot count as reasons for action. This means that the only motivating factors we are left with in our practical reasoning are our own desires or preferences. This doesn’t mean that we can’t take others’ desires and preferences into account in our reasoning, or even that we can’t effectively reason in accordance with practical norms, it is simply the case that the binding force of these must be shored up by our own desire or preference to so take them into account. This is the standard neo-Humean account of practical reasoning.
However, the implication goes further, because insofar as private practical reasoning is all that shores up communication and reasoning as such, then the theoretical dimension of reasoning, and the fundamental norms governing it, must be grounded in this neo-Humean account of practical reasoning. This is neither paradoxical nor unachieveable, and there is a prominent theory of rationality which achieves just this, by conceiving rationality as a matter of preference maximization and the formal structures of reasoning as limit-cases of this preference maximization structure. This theory is rational choice theory, and it seems that sincere Latourians have no place else to turn.
4. Rational Individualism and its Discontents
We are now in a position to articulate the intuition lying behind the connection between Latourian philosophy and neo-liberalism. I must stress in advance here that I am not claiming that any Latourians are card carrying neo-liberals, or that they display worrying submerged neo-liberal tendencies. The claim is one about the implications of the ontology they endorse, and as I’ve noted before, such implications can outrun our grasp of them, and thus, the commitments we acknowledge and what we are actually committed to may pull apart. Again, then, these implications are not necessarily a reason to abandon Latour’s claims (insofar as these are ontological claims), they might just be a disquieting consequence of the nature of the real.
Moving on, we have found that if Latour’s first move is correct, and we sincerely take it to be correct, then we must not only endorse a neo-Humean approach to practical reasoning, but a more specific variant which can properly ground reasoning as such in its strategic element, such as rational choice theory. It is my contention that this is the point at which the political implications of Latour’s philosophy (and OOO, if it takes up this aspect of his philosophy) are to be found.
This is what I take to be the underlying intuition behind Reid and other’s association of Latourian philosophy and neo-liberalism. Neo-liberalism, more than any other political philosophy (even Marxism), is driven by a certain economic model. This economic model is fundamentally based upon the individualistic instrumentalist conception of reason propagated by rational choice theory. The idea is thus that neo-liberalism, and its ideal of politics as a matter of purely self-interested co-operation, is motivated by precisely the kind of individualistic conception of reasoning that Latour’s approach forces us into. Neo-liberalism is the best historical example of a principled political philosophy which eschews all appeal to any independent practical norms, be they social mores or ethical principles, and indeed systematically undermines the possibility of any such appeal.
Much of what is so frustrating about radical neo-liberalism (and it must be admitted that neo-liberalism comes in various more or less self-conscious flavours), is that it does not even admit of the possibility of a genuine (i.e. normatively compelling) process of collective reasoning which works toward the positing of collective ends and the calculation of the most appropriate means to achieve them. It is in precisely this fact that the latently anti-democratic thrust of neo-liberalism consists. This tendency is most obviously manifest in the undemocratic institutions which make up the superstructure of the contemporary global economic system (the WTO, IMF, G20, etc), and the disproportionate influence they hold over democratically elected governments.
Once one has redefined reason itself so as to exclude the very collectively emancipatory potential that formed a cornerstone of the enlightenment project, then all vestige of genuine politics is gone. We are left with ‘office politics’, ‘back room politicking’, and ‘power struggles’.
Now, does this mean that Latour’s position implies the exact content of contemporary neo-liberalism? Not exactly. It might be possible for Latourians to provide a case for a political arrangement that differs from the current one, maybe even a post-capitalist one. However, the real problem is that they have very skant resources for doing so. They can take one of two paths:-
1) They can try neo-liberalism’s own route, but try to show how it leads to a different conclusion. This involves using the resources of something like rational choice theory itself to show how some alternative societal model produces the kind of optimal maximization of preference we require.
2) They can simply accept that they share a desire/preference which grounds some general ethical or political norm which they cannot further justify (e.g., ‘I just want us to live in a Marxist Utopia’), and proceed on the basis that this norm is then a given.
Neither possibility is a very convincing.
The point is thus not that all Latourians and OOO enthusiasts are closet neo-liberals, but just that if they persist in collapsing the distinction between might and right, then the political implications of this move are ones that only neo-liberals have so far successfully swallowed. I would also add that they are not alone here, insofar as they are not the only ones to have collapsed this distinction to some extent. I would say that there is good reason to say that Nietzsche, Spinoza and even Deleuze have also made similar moves (although I would suggest with slightly different motives and in a slightly more nuanced fashion).
There is also a further point to make here, which is that all of this depends upon a certain ideal of sincerity and self-consciousness. However, this is itself a normative notion. There is nothing which demands such sincerity other than its effectiveness in allowing us to maximise our preferences, and in some cases it can thus be suspended, i.e., when we prefer to delude ourselves and others (and can get away with it). Even more alarmingly perhaps, the very critical constraint that we began with – that ontology is not beholden to politics – is just the kind of general norm of reasoning that seems to dissolve on contact with Latour’s position.
5. Hybridization Vs. Naturalization
Now, we’ve worked out the somewhat disquieting implications of the initial Latourian move, but these are not enough on their own to undermine the Latourian approach. However, I think that there is an additional, and related problem to be found in the second Latourian move: the non-modernist elision of the nature/culture divide. This problem undermines the very explanatory motives for adopting the Latourian project in the first place.
A good way to get a handle on this is to contrast the Latourian move with what it is not – the naturalization of culture (which I have written about and argued for here). This is the attempt to either reduce or situate the cultural within nature. Now, it is easy to see why Latour’s (famously irreductionist) position is opposed to a crude reduction of the cultural to the natural. Such a crude reductionism would in effect deny that there is anything about cultural entities that is not describable in the terms of some favoured natural science (anything from physics to evolutionary biology and neuroscience). However, less naive naturalist approaches attempt to situate the cultural phenomena within nature, rather than straightforwardly reducing them to the terms of some favoured science. It is less obvious how Latour’s position differs from this more nuanced naturalism (which Ray Brassier has called methodological naturalism).
I think the best way of characterising this difference is by describing how it relates to our ordinary grasp of the cultural domain. We have a very characteristic understanding of culture as it is from the inside. We might call this our internal understanding of culture. On this understanding, the cultural realm is furnished with all kinds of entities which play important roles in our intellectual and social life. There are obviously people who occupy various social roles, as well as artifacts, fetishes, and works of art, which have obvious ties to the natural realm, but there are other less obviously embodied entities, like ‘ideas’, ‘values’, ‘arguments’, ‘reasons’, ‘norms’, ‘laws’, and so on.
It seems as if the Latourian strategy gives these entities their independence, such that they can play roles within explanatory networks, thereby interacting with all manner of non-cultural and non-human actors in determining both cultural and natural states of affairs. However, as we have already seen, it denies these actors anything other than a sort of generalised causal force (again, this point is mildly problematic in relation to OOO, but we’ll get to it). It strips the furniture of the cultural domain of its normative force, of its apparently value-laden character. This is not to say that it thereby denies the existence of ‘values’, but instead that these become other actors in networks of interaction.
A good example of this stripping of force is descriptions of the values of cultures other than our own. For instance, it is possible for us as amateur sociologists to try and give a description of the norms governing a given society or culture’s taboos, such as the Jewish kosher laws, without thereby endorsing those norms. We present the norms in a conditionalized manner, almost as if they were candidate norms that we could adopt and thus take up the force of (e.g., if one is to prepare food in a kosher way, then one must…). The same is true of much of the internal furniture of other cultures. We can describe the roles the various cultural entities play, describe the values which they are bound up with, without thereby endorsing or adopting those values ourselves. The first Latourian move seems like a generalisation of this detached sociological perspective, whereby we treat all cultures, even our own, as at an arms length.
However, there is a genuine question as to whether this detached perspective is really an external perspective. An external perspective would be able to assess the extent to which the terms in which a culture talks about cultural phenomena, and the ways these are brought about, are in fact adequate to those phenomena. It would as such allow for the possibility of a disconnect between our internalistic conception of the furniture of the cultural domain and the real furniture of that domain. The move of ‘detaching’ from our own culture so as to enable ANT style analyses of how various cultural phenomena are produced is not necessarily adequate to this, because all it does is suspend the force of the values and other normatively constituted entities littering the social landscape, while nonetheless leaving the internal configuration of the cultural domain as it is. The danger is thus that such internalistic conceptions of culture maintain a residual normative structure, even when normative force has been suspended.
So, if we return to the example of the Jewish kosher laws, we can examine them in a conditionalised and detached way, and we can even then situate them in a network of causes through which they were produced and have other societal effects, and often the explanations which are produced will pass muster. The problem is that there really isn’t anything resembling a ‘law’ at any point in the real causal networks which determine the social phenomena we are thereby examining. Yes, there is a fairly stable behavioural regularity within a certain population. Yes, this regularity is a self-sustaining regularity, like a self-regulating process instantiated in a network of interacting human agents. Yes, the internalistic discourse that through which these agents have to interpret and implement the law is part of the mechanisms through which this regularity maintains itself. However, the regularity fluctuates and changes – ways of interpretation and implementation practices evolve. Moreover, the populations’ practice is always heterogeneous, irreducible to any fixed authority which would be the repository of the law.
One can sometimes present simplified accounts of phenomena, that use certain internalistic cultural concepts as shorthand in explanation, such as appeals to the effects of a ‘law’. However, if one looks deeper, what one discovers is a complex causal process that it would be disingenuous to identify as a ‘law’ even if it is correct to claim that this is where the law is manifest. Simplified explanations which appeal to the internal furniture of the cultural domain are perfectly acceptable, especially when we have yet to develop a more detailed conception of the real processes at work, but these should not distract us from the goal of reaching these real processes so that we can hope to achieve more than a simplified explanation of cultural phenomena.
My contention is thus that Latour’s elision of the distinction between nature and culture is a conjoining of the internal view of the cultural domain and the scientific view of the natural domain, so as to allow for hybrid networks through which local analyses of the production of cultural phenomena can be explained. This process of hybridization is to be contrasted to the naturalization of culture, through which we develop an external perspective on the cultural domain, which, although informed by our internal understanding, is not bound by it.
6. Explanation and Causation
The problem with the Latourian approach as I see it is that it hamstrings the very explanatory motives that provide its impetus. The reason for this is that it seems to take the authority over what counts as a genuine causal interaction away from the natural sciences. This has to be phrased carefully, so that it is not interpreted as a relapse into the crude reductionism mentioned above. The point is that Latour sometimes seems to be trying to set up an explanatory perspective which supplants that of the natural sciences. I don’t think this is actually Latour’s intention. It seems that he rather sees himself as aggregating the results of the natural sciences and the social sciences in order to produce insightful networked explanations of cultural phenomena.
However, if one is to get a genuinely causal explanation out of this, then one needs to have a concept of causation which situates all of the different kinds of causal interactions that the different sciences discuss in relation to one another. If we are going to have these flat hybridized networks, then what we need is a way of understanding what it is for any of these potential actors to have a causal affect on one another. For example, we must have some idea about how interactions between economic institutions map on to interactions between human beings, if we are going to posit explanatory networks in which interesting features of certain humans produce novel localised economic effects, or vice versa. The suspicion is that Latour doesn’t have an adequate account of how the kinds of causation between cultural actors map on to the kinds of causation between natural actors.
Latour leaves us with the idea of a sort of generalised causal force (to which normative force is reduced) which all actors exert upon each other. The question is whether we treat this as a lacuna in the approach, that must be filled in with an account of causality that situates all other forms of causation in relation to those examined in the natural science, or whether we treat this as a positive supplement to the approach of the natural sciences. The latter approach seems to be the direction taken by OOO.
Now, I have at several points noted that the idea that this is a strictly causal matter is somewhat problematic, and the reason for this is the insistence by OOO practioners that causality is only one way in which objects are related to one another. In a certain sense, this is fine. That I am taller than my girlfriend is a relation I have to her, which is not a matter of any kind of causal interaction between us. There are however some suggestions that there are non-causal interactions between some objects (I talked a little bit about the notion of causation in relation to Levi’s project here). I’m not sure what this could mean, or how these could play any properly explanatory role at all. However, I believe this is the point where I have to go away and read more about OOO itself. These remarks here are still somewhat tentative.
Ultimately, it seems like what the Latourian approach demands is a complete reworking of the notion of explanation. On the one hand, as has been seen, it undermines the rational norms which are meant to structure the process of explanation itself. On the other hand, it seems to demand a reworking of the notion of causality which potentially throws into question the forms of explanation deployed in the natural sciences. However, it is questionable as to whether these demands are in fact met, or indeed, whether they are even meetable. The two-pronged Latourian approach seems to undermine that which motivates it.
7. Conclusion: Everybody loves Networks
Now, I’ve gone into some detail about my own conception of how the relation between the normative and the causal and the cultural and the natural are to be conceived elsehwere (most notably here, here and here). However, I think it is quite interesting to briefly compare the approach I espouse with Latour’s.
I think both myself and Latour reject a certain kind of approach, namely, what one might call naive modernism. This position takes nature and culture to be radically separate, and gives the cultural domain a kind of ontological status independent of the natural world. A good example of this kind of position would be Heidegger, who castigates Kant for taking Being to only to extend to nature, which he grounds in an account of the Being of the object, and then tries to provide an account of the Being of the subject (Dasein) which grounds the geisteswissenschaften as Kant grounded the natural sciences. Heidegger’s whole approach isolates the cultural domain, and prevents the kind of detailed, local analyses that require some account of the interface between the putatively cultural and the natural.
However, my response to naive modernism is quite different from Latour’s. Whereas Latour purges our internal conception of culture of normative force, and then conjoins this with nature. My approach is to separate our internal conception of culture from the task of giving a genuinely external account of culture. This is to say that I advocate the naturalisation of culture. This is not a matter of reducing the cultural to the terms of some other science, but providing a general methodology for naturalist sciences of culture, which necessarily posit their own entities and kinds of causal interactions. It is simply a matter of allowing those sciences to develop conceptions of the real furniture of the cultural domain, independenly of our naive internalistic conceptions of it.
The obvious benefit of this position is that it retains a space for normative force, albeit one which refuses the normative the kind of explanatory role that Latour fears could corrupt explanation of social phenomena. This allows for both the general norms of rationality, but also for the possibility of ethical and political normativity. (Brandom’s alternative to the neo-Humean account of practical reason espoused by rational choice theory (Articulating Reasons, ch. 2) provides very good resources for undermining neo-liberalism in my opinion.)
However, we shouldn’t think that in rejecting Latour’s position that we must thereby abandon the kind of detailed local analyses of social phenomena that it is meant to enable. I am largely in agreement with Nick’s points about the necessity for performing local analyses of the political situation in order to find points of leverage (not least because they please my inner Foucauldian). I think that most of the methodology of ANT can be made consistent with naturalism. We simply need to recognise that we are not dealing with ‘hybrid’ actors or networks in performing these analyses, but are dealing with thoroughly natural phenomena. As the title of this conclusion says, everybody loves networks, and who am I to refuse them?
Ultimately, the potential to develop radically different conceptions of ourselves both as individuals and as groups that a real commitment to naturalism entails can be a genuinely emancipatory potential. True, it always holds out the possibility of new and more fearsome technologies of control, but it also holds out the possibility of new and more powerful techniques of resistance.