I was prompted to re-read these papers, when I was reading two books that

seem to sum up recent work on governmentality: Mitchell Dean's

Governmentality: Power and Rule in Modern Society (1999) and Nikolas Rose's

Powers of Freedom: Reframing Political Thought (1999). I was very impressed

by these books -- especially Dean's. It was a painful to re-read my own

papers in light of what I had learned from them. In particular, I was

embarrassed by the way that I had invoked an imaginary "we" and had

insisted despite myself on placing "us" in a context that could be fixed

historically. Since both of my papers were intended to disrupt the

assumptions that enable the standard understandings of who and what "we"

are -- in particular the ontology of sovereignty and the connected

epistemology of social science -- it is clear that they did only a small

part of what they were intended to do. This is mostly, but not wholly a

matter of my own limitations. Those limitations are interesting only to me.

But, there may be something more general is at stake, namely the difficulty

of speaking and writing (let alone acting otherwise) politically.

When I read Dean and Rose, I was reminded of the way that social scientists

(including so-called political scientists) are encouraged to write from

outside politics, and thus to deal with the political as a domain to be

described, analyzed, and -- ultimately -- explained away. I refer to this

way of writing in my second paper, in which I make a plea for a political

rather than a social science. I talk there about the naturalizing effects

of a social scientific approach, and appeal to an older, Aristotelian

tradition in which, as it were, one writes of the political from the

inside, as one actor/speaker among others. The conservatism of political

science (to which Hindess refers) is in part an effect of the effort to

speak from inside, as a citizen to fellow citizens. To speak/write

politically is always to imagine a community of some sort, with given

assumptions about its own nature and its own problems, a reading of its own

origins and possible destiny, and so on. So, to speak/write politically is

always to be constrained in particular ways, and always to be posed with

the question of which limits to disturb and which to take for granted. To

speak/write in a social scientific vein -- as I think most analysts of

governmentality do -- is to put oneself in a space where greater analytical

freedom seems possible. But, as the governmentality analysts themselves

might say, all freedoms come at a cost.

Persuaded as I am by Dean's and Rose's analyses, I am not persuaded by the

claim that an analytics of governmentality abandons social science (or

"sociology") in favour of something else. Quite the contrary. The rejection

of history in favour of genealogy and of the "why" in favour of the "how,"

do not make an account any less sociological. The practices described are

still at one remove from the analyst, and are still treated as "objects" of

understanding. Moreover, the imagined readerly community is one of social

scientists. To be sure, there is another community behind that community: a

political community (or communities) that might take up these analyses and

use them politically. But, what that political community is ­ or indeed

what politics is ­ remains unspecified. As Dean acknowledges (p. 198),

government is only one aspect of politics. The other aspect, he suggests in

his recent book, is "a struggle or competition between competing forces,

groups or individuals attempting to influence, appropriate or otherwise

control the exercise of authority" (ibid.). Hindess alludes to something

similar. So far, so good, but note: this is the standard sociological view,

the one that distinguished "government" and "politics" and gave

conventional political science its two main objects of study. What sort of

political science can we expect from that? What imagined community is the

ground from which one writes a sociology/genealogy of politics? What are

the politics of that particular imagining?

I continue to be intrigued by a line of thought that I identify with Hannah

Arendt (especially her famous essay on freedom, but also her speculation in

On Revolution) and Bernard Crick (In Defence of Politics). One might even

invoke Oakeshott's name in this context. What I am thinking of here is the

effort to identify a mode of action, domain of activity, or aspect of life

that is not reducible to the social, the economic, the cultural, or the

governmental. Arendt, famously, draws a line between the political on the

one hand and the "social" (which, for her would encompass the economic and

cultural) on the other. She also distinguishes politics from administration

(or what others would call "government"). However problematic her account

might be ­ and it is very problematic ­ it gestures at something that many

people have been trying to articulate: the stubborn presence of the

political (as both fact and aspiration) as something other than what we can

ever describe in social scientific terms.

If we have common language for describing politics, it is the language of

action, action in which we and other people are, have been, and will be

engaged. If politics has a distinctive domain, it runs through the

governmentalities, rhizomatically. And, it is "in here" as much as "out

there" when we do analytical or other work. Writing is always a political

activity that appeals to an implicit "we" and refers to a universe of

political discourse. So, our papers are all political actions that could be

analyzed as such.

One way of subjecting the governmentality literature to political analysis

would be to ask for whom it is written, and to what purpose. That might

raise some interesting questions about the role of sociology, the relations

between British, American, and continental European thought, the connection

between Western social science and the contemporary practices of Western

hegemony, and so on. I'll pursue another line of thought here. It is

suggested by Mitchell Dean's idea that the governmentalization of the state

is now being followed by the governmentalization of government. This is a

perceptive observation. What I would add is that the governmentalization of

sociology is also at stake: the re-definition of sociology's object and

method to bring it in line with the practices that it seeks to describe. A

certain reflexivity is involved here. This is neither good nor bad, but (to

pick up on the language) "dangerous". The dangers would be worth analyzing.

I take it that political analysis is ­ at least in part ­ an analysis of

dangers.

There's something about the trio, politics-danger-action, that demands

still further analysis. When we hold that trio before us, we get a very

different sense of what is at stake than we do when we focus on the duo of

politics-government. The governmentality literature brings the trio in (as

Foucault did) by talking about "strategies". So, typically, we have

strategies without strategists and wars without combattants. This is

puzzling, to say the least. Moreover, the militarism of the imagery

suggests rhizomatic connections that demand explanation. (Barbara

Cruikshank's distinction between strategic and tactical politics is

interesting in this respect. In insisting on a freeing, she moves us from

one sort of war to another. The imagery is not peculiar to her: We all use

it in one way or another. I find it hard to resist the inference that

something "lies behind" this way of thinking.) There is a relation between

the duo (the one that inclines us to think about politics in terms of

government, regulation, law, and hence justice) and the trio (the one that

inclines us to think of politics in terms of action in face of danger,

hence violence and war) that needs to be explored. Mitchell Dean moves in

that direction in his latest paper. But, there is another trio,

discussion-deliberation-decision, that must also be brought into some

relation with the other two. (Tully's work points us in that direction.)

The standard view in political philosophy seems to be that the central

political problem is to make this third trio govern the other two: to make

"politics,' in this sense, govern government and govern violence. So-called

political philosophy thus embodies a particular aspiration. How to make

sense of that aspiration is part of what is involved in giving an account

of politics. I suspect that the aspiration is not what it appears to be,

but I get little help from the governmentality literature in making sense

of it (or learning how to respond to it within myself).

One of the interesting features of the governmentality literature is the

way that it re-inscribes the ethicization of politics. The shift from

morality as a set of rules to ethics as a form of practice is presented as

a description, but also, implicitly, as a prescription for human activity.

"Ethics" is held out as the paradigmatic reflexive activity. Little noticed

is the way that this confirms the tendency -- especially evident among

philosophers -- to collapse politics into ethics, and hence to obscure the

ways in which ethics (in the plural) are deployed politically. (It is

interesting that neither Dean nor Rose refers to John Rawls, the most

influential "political philosopher" of the late 20th century. This, despite

the fact that Rawls offers a risk analysis of justice, and sets that out as

a normative foundation for politics.) Machiavelli's point about the

relationship between politics and ethics is still pertinent. Cruikshank

deepens our understanding of the matter.

One might well write a genealogy of the political, similar to the proffered

genealogies of the social and the economic. Such a genealogy would be

interesting, and it would help forestall tendency ­ perhaps evident in my

remarks here ­ to essentialize the political. Nonetheless, it would still

leave us puzzled about thinking politically. Part of the problem, I think,

is that genealogy, like the rest of social science, depends on an ontology

of sovereignty.

Warren Magnusson

28 July 2000