william sewell, logics of history

(6-8, 10): time as 'fateful -- because events occur once and only once in a given temporal sequence, there is something irreducibly contingent about events. historians see enormous events having the same sort of fatefulness and contingency as smaller events

(10): here the claim is that social life is fundamentally consituted by culture (in the widest possible sense). social relations can be transformed, when cultural systems are transformed.

(14): 'underlying causal structures themselves undergo mutation' [again, a priori this can't be conceded -- there has to be some account of how and why this takes place. Brenner's argument about the transformation of fedual rules of reproduction into capitalist rules of reproduction is an excellent example]

(16): against naturalism

(16): 'societies are culturally distinct at some deep level'

(84): 'eventful temporality' (against 'teleological' and 'experimental' temporality)

(84): a telelogical explanation as one which attributes a cause to 'abstract transhistorical processes' [this is a mischaracterization, if it's meant to apply to Marx -- especially once we've purged him of his Smithianism]

(86-87): rejoinder to Wallerstein -- can't understand the part, from the whole, as he tries to do.

(90): the objection to Tilly [in the concrete, we see the pay-off -- which is nothing, really. all you have done is said that Tilly has mis-specified, and that a different causal argument makes better sense of his findings]

(93-94): similarly, the objections to Skocpol (that she doesn't have the full gamut of empirical cases she would need to make her argument solid and that some empirical examples demonstrate that 'revolutions' may happen in the absence of military reversals and fiscal crises; her cases are not 'equivalent' and 'independent'

(100): 'events as the relatively rare subclass of happenings that significantly transforms structures'

(102): proving Marxism as 'teleological' by pointing to how it thinks we 'can't go backwards'.

(103-105): on the Paris Commune [again, the argument is a reasonable, concrete amendment to a causal claim. instead of social class --> attitude, we get social class --> organization --> attitude. the fact that there may be contingencies here doesn't change the fact that there is a causal argument being offered]

(109): in the Longshoremen example, he's suggesting that 'strategy''s importance demonstrates the iportance of contingency. [nonsense! strategy can matter, but it is not an ontological statement about the way things work. sometimes it will, sometimes it won't. you can only make the point in the concrete]

(114-123): response to the "scaling-up" objection (that contingency matters for small things, but not for big things) -- and here, telling, that Michael Mann is his ticket out [again, it's not clear that he really proves what he wants to prove, by it]

(125): 'structure' arguments lose the importance of agency

(139): important reminder that ideas are thoroughly inconsistent -- a theory of ideology can't purport to explain everything

(143): to be an agent means exerting control over the social relations in which one is enmeshed

(146): 'depth' and 'power' (how structures vary, in his argument)

(148-149): his brief statemtents on the State and Capitalism are almost intentionally opaque -- and entirely unhelpful

(236): 'came to be understood' (the Bastille came to be understood as a revolution in which the people rose up) [but who's doing the understanding, here? his argument is a chronicle of the delegates to the National Assembly. this proves a world-historic transformation in the understanding of people to power???]

(220): watch, as contingency is stretched to its most absurd limits: had they slaughtered the soldiers (and not just de Launay), the modern conception of revolution might never have been born!!! [no comment]

(257): the 'tissue' of privilege destroyed, a new social order [here the Revisionists would be helpful; he's factually mistaken]

(268): a profound emotional experience matters causally [the trouble here, as with much of his argument about culture, etc., is that there is no way of proving that this is unique (let alone the difficulty of proving (a) that it happened; (b) that it had the effect he is suggesting it had]

- - - -

[1] temporal heterogeneity -- this is unapologetically violent to the tasks of social science ('explanation')! you can argue that events occur once and only once -- which is of course true, formally; it was the grounds of Hume's skepticism -- but social science is the attempt to abstract from that which is irreducibly particular to arrive at general conclusions across contexts.

not unlike Chakrabarty, the claim that events are always 'unique' (or the claim that a given act is affected fundamentally by the nature of the social world in which it happens, such that you can have 'different logics'; that societies are 'culturally distinct' at some deep level) is very unstable. if that's the case, then what are the grounds to speak across any contexts whatsoever? and what's to stop 'contexts' from becoming infinitesmal, without some larger account of why general conclusions are possible. of course, all good social scientists can admit that things can have different "logics," but because they are distinct instantiations of a common logic which prevails at a higher level of abstraction.

part of the problem might also rest in the definition of 'logic' (I am understanding it as 'laws of motion,' give or take -- what does he mean? something like the 'laws of gravity' don't apply here?)

[2] 'global contingency' -- either this means something basically trivial (that people have to be attentive to context, that one's causal argument might be rendered invalid when transposing it), or it means something completely ludicrous. in the former case, it's a fine objection to make -- but it is impossible to make in the abstract. his concrete example (the longshoremen) are interesting, but are nothing more than highly specific causal sequences. in the latter case, he is making this a statement of abstract principles. in which case social science becomes impossible. you and Niall Ferguson share this commitment, if that's any consolation.

this isn't entirely a question of scale. you can have contingencies that matter in big ways; but there is a presumption that as you zoom out, contingency becomes less and less relevant as an analytical tool. in the passage where he cites the Mann, though, one gets the sense that he wants to resist even this.

[3] 'cultural transformation' -- this is fundamentally unhelpful, though. at one level, there is no account of causal priority. when he qualifies this, Sewell will want to say that this can be true within limits. that culture matters, but not absolutely. but if the importance of 'cultural transformation' is true within limits, then how do we specify those limits in the abstract? his formulation has the effect of leaving it entirely 'open'. moreover, doesn't there need to be some account of why cultural systems might be transformed? otherwise 'changes in meaning systems' is totally opaque, as a causal argument -- unless you have some account of how and why they change, how do you resist the argument that they're always changing. there's, further, a problem of measurement -- to which Skocpol calls attention, in her rejoinder to the Bastille argument -- the ability to speak of a unified 'cultural system' (and, by extension, uniform changes in that unified cultural system) demands enormous proof. one would have to show that this works its way through a number of social actors that find themselves in various places in the social hierarchy, etc. quite inconceivable, really -- as it should be, given a materialist understanding of how culture operates.

[4] 'societies as culturally distinct at some deep level, meaning that putative 'social laws' can only be valid locally': i'm sorry, but aside from the fact that this depends on the arbitrary assertion of 'unitary societies' (so India can be a whole, perhaps?), which raises the whole question of instability again -- this is utter Orientalist gibberish (the 'Hindu' logic). that he has abandoned their normative project is no credit to him. you can valorize it whichever way you like (nativist, orientalist, or ambivalent); but this can tell you nothing about India, China, etc.

now, you can make a case -- like Wolf did, or Marx did -- about how different rules of reproduction might prevail in different, owing to the existence of different socio-property relations, whatever. but in this case Society A and Society B are distinct elements of the same Set (so, they are related at a higher level of abstraction--both are societies in which humans enter into definite relations of production in order reproduce themselves). there is no statement of essential difference, here.

[5] rejoinder to Wallerstein -- KNAVE! Wallerstein is making a concrete argument. it might very well be the case that you can't read the part, from the whole -- this is the thrust of Brenner's objection. that what's happening in England is internal to its socio-property relations. but you can't make this objection as an abstract declaration of faith. it needs to be made concretely; show that the specific causal claim Wallerstein is making is invalid.

[6] objection to Skocpol -- he purports to demonstrate a 'crisis of method', but all he shows are empirical weaknesses, and concrete difficulties. the claim that Skocpol's cases aren't 'equivalent' and 'independent' misunderstands the fact that this is an exercise in approximation -- the claim has to be that these are 'equivalent'/'independent' enough to be compared in this fashion. Burawoy makes a decent claim that they are not, but it's a concrete evaluation. it does not follow, from this, that there is no such thing as 'equivalence' and 'independence'. say, for example, you want to compare welfare state retrenchment? what does his assertion mean for that research project? hogwash, that's what.

[7] events transform structures -- whatever we think, surely this claim has to meet two challenges: (1) logical, you have to demonstrate that there's a mechanism. what is the mechanism, in his argument? how do events transform structures?; (2) empirical, you have to show us that 'events' reliably did this. the Bastille example, for all the reasons Skocpol enumerates, is unsatisfying.

[8] Marxism as telelogical -- but there's a logical reason for that, as Erik Olin Wright and others argued, in the Theory of History debates. it will not be in any social actor's interest to oversee the degradation of the productive forces. whatever you think of the claim, it ceases to be 'telelogical'. he has to do a hitjob on the theory in order to make it seem mystical.

[9] Structure and agency -- no, it's just that agency gets incorporated into a 'structural' argument. action can matter, but it matters within limits that are set by forces beyond any give agent's (or collective agent's) control. there are infinte examples of this, of course: take the theory of the capitalist State. mobilization can matter, but it matters differently, and within limits (until the final showdown!).

incidentally, I have never understood the temptation to stress the importance of agency (as he puts it on p. 143, "to be an agent means to be capable of exerting some degre of control over the social relations in which one is enmeshed"), despite everything. it becomes emphatically reactionary. are people poor because they failed to pull themselves up by their bootstraps? this is just silly.

[10] his rejoinder to Bourdieu, where he raises the question of 'existing social conditions' and ideas' relation to them -- in this he is correct to stress the heterogeneity of ideas, but he throws out the baby with the bathwater. theories of ideology cannot, he's right, hope to explain everything that people believe. it has to be a story of propagation of popular, salient ideas, in the aggregate