Thursday, 24 October 2013

A concept of power without potentia

In a recent post I reflected on the significant limitations Latour's Deweyan political mode imparts upon his political philosophy.  I concluded by arguing that politics can only make sense given a concept of power.  Latour rejects power for largely philosophical reasons (and arguably sound reasons as far as they go).  Therefore, we need a concept of power that is compatible with the best of his thought but doesn't submit to the unnecessary limitations that he has placed upon himself.  The following constitutes an initial, schematic, impressionistic move to that end.  A first stream of thought that is hopefully headed in the right direction but has not yet located its target with any precision.

The problematic of power without substance.  To abstain from a concept that is essential for all political thinking because of philosophical quibbles is simply petty.  Power qua the potentia of a substance may be conceptually inadequate for our purposes but etymology is not an impassable barrier to conceptualisation.

Power as placement within an assemblage, a network with hierarchical or non-symmetrical properties.  P has power not because of any innate or bodily potential (his body plays a part, of course, but scarcely more than his aortic valve) but because of his placement within a vast assemblage of corridors, telephone wires, aides, secret service personnel, flags, lawns, bureaucrats, journalists, tanks, guns, bombs, etc.  All these things bear a relationship to P that is asymmetrical.  They are there for him in ways that he is not there for them (and vice versa, in fact).  His face is on practically every flat surface; a house of mirrors with him in the sweet spot. Those who buzz and flit around him are practically invisible by comparison.  He issues orders, they follow them with degrees or gradients of resistance, elements of disobedience and betrayal, yes, but far weaker, shallower gradients, far lesser betrayals than orders issued in the opposite direction!  (This particular Circle is massively lopsided.)

It isn’t difficult to think of power within networks (nor is it especially original).  Power is a property of actors arrayed in a highly particular fashion, a property of network structure – and structures only endure because of processes keeping them that way, yes, but given those processes power is real enough to be referred to by name.  A figuration, a synecdoche, yes – but that is true of all proper nouns.

Gradients of resistance.  The master issues orders and it is as though his words roll downhill, effortlessly.  The servant follows them dutifully without any real objection; his only immediate concern: trying to avoid being crushed under their seemingly massive weight!  The servant can make requests but only at great expense, with difficulty and only on special occasions.  It is as if he has to tenderly, gingerly roll his request up a hill and the master can disinterestedly swat it away, commanding, as he always does, the high ground.  With the slightest of disinclining gestures the request rolls back down again.  The servant is worse off than even Sisyphus, he cannot keep rolling his stone up again and again!  Next time the master’s response may not be as benign as a disinterested swat…

‘Ah, so the master ‘has power’ and the servant ‘does not.’’  Indeed, this is how it appears and it appears this way because this is how the network is structured.  And as long as it remains structured in this fashion then the dialectic keeps a’rollin’.  However, all networks are unstable – and structures are the least stable network-configuration of all!  All networks require institutions to keep them solid.  And this is especially true of tightly structured networks that must, if they are to remain self-respecting, appear to remain unchanged over decades, even centuries.

However, if they do hold (an empirical question) then the master really does ‘have power.’  The amount of ‘work’ he has to do in order to rebuff his servant’s meekest of requests is inconsiderable; the servant, contrariwise, has to work day and night just to meet his master’s approval and even then his words are so weak that they barely leave his mouth.  ‘Yes, sir.  Thankyou, sir.’  These are not his words even if they issue from his lips.  To speak his own words in his master’s presence is extremely risky – that is unless his master has granted him that privilege: ‘You, there! Say what’s on your mind, lad!’  A small foothold on the mountain.

The differential gradients are as real as anything – as long as the structure holds.  We can and should interrogate these structures to see how they work, however we needn’t and, indeed, cannot pick apart every structure every time.  Black boxes, habit – these are not the pudding-skin of false consciousness.  A habituated, instituted structure is a real structure with real effects.  It is churlish and unnecessary, therefore, to deny the existence of power.  Yes, it is a network-effect, yes it is unstable – but what isn’t?  You can’t tackle any explanandum without some explanans and sometimes that means taking network-effects for granted (they grant us this privilege when they stay stable, for whatever reason).  Indeed, this is the only way anything can ever happen.

If every network fell apart all at once there would be nothing because there is nothing beneath networks; no Nature, no God.  Nothing to pick up the pieces.  If the servant stopped to consider his situation, perturbed from the trajectories of obedience for whatever reason, he may come to rebel.  Habit keeps the relationship stable.  The very taken-for-grantedness is one thing that keeps the structure intact.  So, might we not want to untie those bonds for that very reason, to help unburden the poor servant?  Yes, we may very well do.  But (a) one cannot unburden every agent all at once (and not all want saving), (b) habit is only one thing keeping the structure stable, causing the servant to rebel could be very bad for him – he could be fired, thrown out by his unsolidarist former colleagues, blacklisted, even beaten or worse and with impunity because of the master has the ear of the local sheriff, etc. –  and (c) this is only an example, an archetypical (the archetypical) instantiation of a power relationship.  The point is to show that power gradients are real and that they can legitimately function as explanans if they relate to steadily structured network relations, the configuration of which establishes the differential work-gradients that we recognise as ‘power.’

The fact that these gradients depend on more than themselves only matters insofar as (1) we are morally or politically obligated to unpick these configurations in particular (in which case we particulate them and them alone) or (2) these structures aren’t sufficiently stable for our purposes and our explanation risks falling apart when they yield.  In any case, political power and networks are no more incompatible than electricity and the National Grid.

If kept stable the power is indeed there ‘in’ your wall socket.  If patriarchy remains intact then the man sat at the head of the table really 'has' power over his children and womenfolk.  If P succeeds in getting re-elected, avoiding impeachment and assassination, keeps his donors 'onside,' avoids being made a 'lame duck' (where he becomes P in all but name and is then confined to activities that belong purely to his office, i.e. bombing somewhere with a desert) then, yes, he is the most powerful man in the world.

Etymology be damned, a concept of power without potentia is entirely possible – and it is quite simply a requirement for any kind of politics (including [pol]) to make any sense whatsoever.

And before wise people object that none of this is new: I know.  That's my point.  Latour has abstained from 'power' for many, many years and his political philosophy is horrifically malnourished as a result.  My point is that there is simply no need for this.  The problematic is more a stream than a raging river and with some fairly light conceptual work the stream can be bridged.

Power is not a utility in short supply.  There is no danger of blackouts if we tap into the grid.


  1. if we just stick to "thick" descriptions of who is doing what to whom (and in what context of course) do we really also need a concept of power? not sure what advantage this gives us, what additional work it might enable?

  2. Because not every question can be answered with thick description. It's not a panacea. Sometimes thought (and research) has to move fast and deal with vast, high velocity, widely dispersed phenomena.

    Thick description is a luxury. Were we always able to simply describe every actor in a situation, how it gets strong, how it is able to do what it does that'd be great but that simply isn't a method that's applicable in all cases (or even many - or any if it's taken to an extreme).

    I agree that we need to avoid slipping back into the idea of having 'Theories' that explain cases. In this positivist mode of research, cases become nothing more than data to be mushed up and inputted into the machine. The theory becomes everything and the case nothing. I don't want that.

    Again, I have to refer to my interest in international relations. 'Following the actors themselves' simply isn't an option for most of the situations IR deals with. As well as the degree of complexity involved, a large proportion of the phenomena under consideration are strictly secret. You can't, as a political scientist, wander around the White House watching people work. Different modes of inference are necessary; different kinds of things have to be taken as the explanans. Thick description has its explanans, too (description and explanation are not opposed - a thick enough description *is* an explanation). If it couldn't take certain entities for granted it would never be able to describe anything. If it had to unpick every black box and trace every relation nobody would ever be able to write up their field notes. They'd never be able to stop taking field notes! Their study would end up looking like something off one of those 'hoarder' shows! Notebooks up to the ceiling!

    But also this isn't just to do with research and methodology. It's also to do with how we think about politics - as citizens, as activists, as human beings (and not just as researchers or scientists). Taking 6 months off work and doing an in depth ethnographic analysis isn't really an appropriate way to respond to seeing a news story on TV for most people.

    Thick description is an admirable and underrated method - in many ways its the ideal - but it cannot suffice for all circumstances. Or, perhaps, what counts as thick description changes radically depending on what you're describing. Perhaps 'power' is, in some circumstances, simply one of those things that thick description has to describe thickly in order to describe a wider range of phenomena.

  3. well one can be as accurate as possible in understanding what is at hand but I'm still not sure what work power as a concept adds to descriptions of what has happened, as for responding to the news (and all complexity of events/circumstances) that raises some serious questions about the possibilities/potentials of actually being an 'informed' voter/citizen (an idea[l] that I find increasingly dubious), and just as a point of information (I never studied the field) how do political scientists check their results/assumptions/methods, against what? thanks, dmf

  4. According to the ethos of thick description (as I understand it) one should never assert any kind of 'potency' to any actor but simply describe what actors do in practice in concrete situations and describe how they are able to do so by reference to the 'context' or the networks in which they are situated. Therefore, if any actor appears to have any essential ‘potency’ then that is because the description is not thick enough yet. The analysis is complete if and when ‘potency’ *disappears*. Thick description is great if you can afford to move that slowly and that freely. But that is rarely the case.

    With regard to the possibility of an ‘informed electorate,’ in fact Latour’s whole political mode is based upon that question as posed by Lippmann in The Phantom Public and Dewey in The Public and its Problems. Their idea is that since the public cannot possibly be informed of all the complexities of the contemporary world (and they were writing in the 1920s!) then democracy has to operate on the basis of publics that form around particular issues that are of concern to them and elite decision makers must, in the end, appeal to the support of these publics to decide undecidable issues. I don’t think that Dewey and Lippmann are quite as profound a pair of thinkers as Latour seems to (he seems to think that they’ve resolved the problem, at the level of theory anyway) but they definitely address this question. The way I’m trying to address it is approaching it from a completely different direction, really.

    As for political science, there are no universal methods but the positivist philosophy of science very much predominates in North America, where the emphasis is on abstract models (methods often imported from economics who, in turn, appropriated or perhaps misappropriated them from physics). One constructs such models, which are often but not always mathematical, and feeds them data, usually quantitative, and thereby ‘tests’ the model. The ‘best’ model is the one that explains the most cases with the least number of assumptions. It’s a bullshit philosophy of science but it’s popular. But then, of course, there are poststructuralist discourse analytic methods, historical materialist methods, more traditional empirical-historical methods and so on. All kinds. And then there are ethnographers who, for example, try to understand the experiences of people in war zones or the memorialisation practices of military families, etc.