• Ingen resultater fundet

Chapter 10 concludes in terms of the main research question of the dissertation

2 Identity configuration and conflict: Discourse as structure, agency and interaction

2.1 Identity as discursive double structure: void and narrative narrative

2.1.4 The structure of temporality in policy narratives

2.1.4.1 The first aspect of time: the future visible from the present

This subsection develops the first aspect of time – the future – to be a horizon of expectation articulated to be springing out of the present. A discussion of the roles of death, of mythical fulfillment, and of the demand for closure – all constructing definitive points of arrival for time in narratives – leads to a theoretical preference for the more indefinite 'horizon of expectation'. The proceeding subsections show how the relative undecidability of a horizon may be reduced by present articulations of the past as a space of experience.

First: Death. "Heidegger subscribes to the ancient maxim that 'to philosophise is to learn how to die'. Mortality is that in relation to which we shape and fashion our selfhood." (Critchley 2009a) Therefore, according to Heidegger,

the primary phenomenon of time is the future that is revealed to me in my being-towards-death. Heidegger makes play of the link between the future (Zukunft) and to come towards (zukommen). ... The human ... always projects towards the future.

(Critchley 2009b)

The conclusion is, in the words of Heidegger, that "[temporality temporalizes itself primordially out of the future] Zeitlichkeit zeitigt sich ursprünglich aus der Zukunft"

(Heidegger 1984:327 quoted in Critchley 2008:147).

It is true that the human projects towards the future – and policy narratives do too.

Heidegger's account, nevertheless, puts too much emphasis on the future – and especially one element of the future: death. It might be so that philosophy is learning how to die. But politics seems to me much more to be concerned with how to live on.

Notably, as we will return to in a couple of pages: how to live on.

Part of the reason might be that politics involve collective identity, and collective identities do not face death in the same way as individuals. Death, according to Heidegger, is "unbezüglich, gewiβ, unbestimt and unüberholbar: non-relational, certain, indefinite and not to be outstripped" (Critchley 2008:143 paraphrasing Heidegger 1984). As we shall see below, Critchley takes issue with the element of non-relationality. Nevertheless, 'death' for a collective identity might be indefinite and not to be outstripped –– but death does not appear certain for a collective identity in the same way as it does for a thoughtful individual. A series of scholars of nationalism and religion even suggest that the function of the narratives which constitute these phenomena is to imbue the mortal individual with a sense of immortality (Anderson 1991:ch.2; Smith 1991:161ff). The shift of weight away from certainty towards indefinity alone makes a difference for the temporal structure of policy narratives.

Contrarily, for a collective identity death may be overcome by narration. And it may be overcome exactly by continued narration. As Campbell puts it in relation to one

type of collective identity: "For a state to end its practices of representation would be ... death." (1998:12) Perhaps death is the vanishing point of narration – but in different ways, when it concerns individual and collective identity: "Silence is the impossibility of meaning and the possibility of the termination of inter-subjective contiguity. ... Silence=Death ... However, this silence is not the end of the story; on the contrary it is the beginning. Silence is a necessary condition of the act of narrative production" (McQuillan 2000b:27; italics added). Authentic individuals might narrate-towards-certain-death – collective subjectivities narrate-to-postpone-indefinite-death. And even if silence appears to signify death, the silence of a dead collective might not be as irreversible as the silence of a dead individual. In that sense, it is necessary to allow for a more open-ended structure of temporality when developing an analytics to be employed on policy narratives.

But there are other 'ends' than death. For one, there is fulfilment. Or rather; Heidegger sees fulfilment in death – but there are other fulfilments to approach. One is the realisation of ones true identity in Utopia. Laclau introduces the concept of 'myths' in his theoretical framework as "'spaces of representation', which are designed to make sense of and suture dislocations." (Howarth 2004:261) Kølvrå develops the concept to designate a specific form of narrative with a specific kind of end: "Political myths .. always involve ... an eschatological dimension, because ... they project utopia into the future." (Kølvrå 2009:37) Kølvrå' s Lacanian point is that fulfilment in Utopia is impossible – but the desire to fill the lack and reach Utopia is what drives narration.

Agreed: The urge to reach out towards Utopia is in many cases an effective motor for narratives of collective identity. But is it necessarily the only one? Perhaps there are these overarching narratives holding each large scale collective together. But perhaps there are not. Perhaps there are only smaller narratives suggesting how to go on;

leading to 'lesser' ends than death or Utopia. The smaller narratives may recur to grander narratives. Perhaps they have a better chance to catch on if they do so. But do

they always have to relate to grand narratives? Does one necessarily refer to one organizing suture waiting in the (impossible) end, every time one acts politically?

Kølvrå bases his argument on White who is investigating the desire for having reality be endowed with a meaning, so that we may tell a story of it in which "events seem to tell themselves" (1980:8). He finds historiography to be "an especially good ground on which to consider the nature of narration and narrativity because it is here that our desire for the imaginary, the possible, must contest with the imperatives of the real, the actual." (1980:8) Kølvrå paraphrases White to the event that "the problem with narrating history is that history as the eternal succession of events in the world does not end, but narrative as a form must end – it must at some point conclude" (2009:35;

cf. White 1980:26). And the reason is that for a narrative to work, it needs to have a point: "The demand for closure in the historical story is a demand ... for moral meaning, a demand that sequences of real events be assessed as to their significance as elements of a moral drama" (White 1980:24).

I do not have any issues with this account as such; what I wish to contend concerns, firstly, scale, and, secondly, the problems involved in conceptualizing the future as part of history and the spatial metaphoric imported in this way. To specify these problems, we need to look closer at the argument White makes.

To me it seems that both the problems concerning scale and the choice of spatial metaphoric pertains to the way in which White generalizes from history writing: As an introite to his focus on the role of narrative, White enumerates a series of diacritica of good history writing:

by common consent, it is not enough that a historical account deal in real, rather than merely imaginary, events; and it is not enough that the account in its order of discourse represent events according to the chronological sequence in which they originally occurred. The events must be ... narrated as well, that is to say, revealed as

possessing a structure, an order of meaning, which they do not possess as mere sequence. (1980:9)

More specifically, "it must honor the chronological order of the original occurrence of the events of which it treats as a baseline that must not be transgressed in classifying any given event as either a cause or an effect." (1980:9; italics added) In his concluding remarks, White returns from the specific analysis of history writing to the more general problematique of his introduction and of the title of his essay – "The Value of Narrativity in the Representation of Reality" to say that "this value attached to narrativity in the representation of real events arises out of a desire to have real events display the coherence, integrity, fullness, and closure of an image of life that is and can only be imaginary." (White 1980:27; italics added)

As indicated above, two points are to be made here: One concerning size, and one concerning the specificity of an end.

Firstly, at what size may an 'end' have the critical effect of closure? Which is to say:

At what scale does narrative kick in? Death may be a Big end suitable for a Big narrative of individual identity; and Utopia may be a Big end suitable for a Big narrative of collective identity. And ideal type Modern history writing, focusing on the overall political-social order (cf. White 1980:15) needs big ends to narrate big moral points.

But is not every articulation of a cause to an effect a narrative – beginning with the cause, ending in the effect? Narratives may be found on a much smaller scale aiming at much smaller ends – in every utterance connecting a cause to an effect. "It may be useful, then to define the action of a narrative as the representation of an instance, no matter how small, of ... events and existents in a chain of temporal causality or at least contingency" (McQuillan 2000b:8 paraphrasing Seymour Chatman). Such small scale narratives play a central role in the interaction of identity politics, whether or not they add up to or articulate big narratives.

Secondly, even if one never knows when Death will arrive and if one knows that one never ever arrives in Utopia, the direction is in both cases unambiguously decided. In narration, according to White, one projects a line towards a single point.

This kind of singular end is not particularly apt for an analysis designed to open up alternative futures.30 This may be achieved, contrarily, by pointing out the articulation of the hegemonic projection as but one among other possibilities.

Singular projection is part of the construction of a future – but it cannot stand alone;

it needs another metaphor to play up against. 'Horizon' is such a metaphor.

But how are the two metaphors – projection of a point and horizon – to be combined?

McQuillan suggest that "the subject can construct a present as a distinct ontological region of reality ... by placing an imaginary horizon on the boundless and differential syntagm of narrative" (McQuillan 2000b:20). So you have an endless line projected;

you babble along adding events to events – and then you cut it off by 'drawing down the curtain' to form a horizon. The result is that the projection stops at the horizon.

In Critchley's rendition of Heidegger, the relation is the opposite: "to understand what it means to be an authentic human being, then it is essential that we constantly project our lives onto the horizon of our death." (Critchley 2009a) In this version, the horizon is there – and we need to direct our story to reach it at a specific point. The story must be made to end at the horizon.

Both these ways of conceptualizing a horizon, however, produce points.

In the way Koselleck develops the spatiality of the future-as-expectations, the horizon is not just producing a point – it is producing a line, a line allowing for the projection of a series of alternative points. Koselleck's description of the aspect of the future as expectation begins like Heidegger's in saying that "expectation ... takes place in the

30 Cf. Neumann & Øverland (2004).

today; it is the future made present; it directs itself to the not-yet, to the nonexperienced" (1985:272). But Koselleck does not continue the list describing the future as the not-yet and the nonexperienced to end with the single point of Death. In stead the list describing the future widens out as it directs itself "to that which is to be revealed. Hope and fear, wishes and desires, cares and rational analysis, receptive display and curiosity: all enter into expectation and constitute it." (1985:272) As "that which has yet to be made is spread over minutes, hours, days, years, and centuries; ...

only the individual parts are visible" (1985:272), hence

it is more precise to make use of the metaphor of an expectational horizon ... The horizon is that line behind which a new space of experience will open, but which cannot yet be seen. The legibility of the future, despite possible prognoses, confronts an absolute limit, for it cannot be experienced. (1985:273)

The absoluteness lies not in an end point of Death but in a limit of experience.

Figure 2.3 illustrates how the first aspect time – 'the future' – may be conceptualized as the horizon of expectation visible from the present.

By insisting on the future of a narrative to direct itself towards a horizon in the sense of a line of possibilities, weight is lifted from the shoulders of the future. The weight is tilted backwards; towards the past and, not least, towards the present. Death may be articulated – as may Utopia – but it is a possibility rather than a necessity.31 But the obligation to grasp the events together into a narrative comes not from Death but from articulation. The two remaining subsections of this section develop the spatial structure of the two remaining aspects of time – past and present – suitable for the analysis of policy narratives.

31 In chapter 3, the dissertation returns to how necessity may be articulated in narratives through specific conceptualizations of the future.

Present:

Ontology

Future as Horizon of Expectation

Figure 2.3 The 1st aspect of time in policy narrative

The future as horizon of expectation visible from the present ontology