You are on page 1of 19

This article was downloaded by: [Universidad Autonoma de Barcelona] On: 03 April 2014, At: 01:04 Publisher: Routledge

Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Rethinking History: The Journal of Theory and Practice


Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rrhi20

Frank Ankersmit's narrative substance: A legacy to historians


Peter P. Icke
a a

History Department , University of Chichester , UK Published online: 27 Oct 2010.

To cite this article: Peter P. Icke (2010) Frank Ankersmit's narrative substance: A legacy to historians, Rethinking History: The Journal of Theory and Practice, 14:4, 551-567, DOI: 10.1080/13642529.2010.515809 To link to this article: http://dx.doi.org/10.1080/13642529.2010.515809

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly

forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Rethinking History Vol. 14, No. 4, December 2010, 551567

Frank Ankersmits narrative substance: A legacy to historians


Peter P. Icke*
History Department, University of Chichester, UK

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

This essay is not intended as a critique of Ankersmits theorisations, neither in the general nor in the particular. Rather, it will focus on Ankersmits concept of the narrative substance at an explanatory level only, expressly in order to advance the arguments herein. To be more precise, I will explore two very signicant points which, it appears to me, unquestionably arise out of this specic element of Ankersmits wider thesis. In doing so I hope to illustrate the inescapable centrality of language in historical theory a matter which has been pushed to one side, or wholly excluded, following the recent fashionable growth of interest in memory and experience studies. Keywords: Ankersmit; narrative; colligation; metaphor; symbolism; language

Frank Ankersmit is currently Professor of Intellectual History and Historical Theory at the University of Groningen in Holland. Over the last 27 years he has published a substantial body of literature relating to, and exploring the consequences of, recent developments in historical theory. That is to say, to be more precise, the development of historical theory from its radical re-assessment occasioned by the so-called linguistic turn of the 1970s/early 1980s up to its present day (2010) interest in memory and experience studies. However, for the purposes of this paper my particular interest is in Ankersmits very rst book Narrative logic (Ankersmit 1983) and, to be even more specic, I am focusing here on its central and dening element, the narrative substance. First, however, some brief prefacing remarks in relation to the work of Hayden White, with whom Ankersmit is so often linked, and then a couple of more general negative comments. With regard to White, it seems obvious that one cannot and Ankersmit certainly does not ignore Whites huge contribution to the eld of historical theory and indeed Ankersmits work could in some ways be seen as a development of it.
*Email: peter@ickep.fsnet.co.uk
ISSN 1364-2529 print/ISSN 1470-1154 online 2010 Taylor & Francis DOI: 10.1080/13642529.2010.515809 http://www.informaworld.com

552

P.P. Icke

Nevertheless, I think it would be a mistake to imagine that Ankersmit might have found in White a model for his own work, for whilst he presents a distinctively personal philosophical style of argument which, broadly speaking, carries him towards the same conclusions as those expressed in, for instance, Whites Metahistory (1973), he has always been (and indeed is) original and complex in his own right. Furthermore, I should point out that before the publication of these two books (Metahistory and Narrative logic) historical theory concerned itself almost exclusively with the philosophical problems associated with historical research and thus (despite White) there was arguably no comprehensive philosophical analysis of the narrative writing of history per se. Ankersmits purpose in writing Narrative logic was, as he insisted, to put this matter right and in so doing he demonstrated that contemporary historical theory was inadequate to its undertaking; that it failed to satisfactorily account for the nature of the narrative form which was generally taken to arise unproblematically out of the product the facts and singular statements of historical research. Ankersmits intervention in historical theory accordingly corrected that established position by proposing that the hitherto presumed transparent process of historical writing was nothing less than opaque and, furthermore, that almost everything that really mattered in historical theory took place between the completion of historical research and its writing up into narrative form a process which Ankersmit referred to as the trajectory from evidence to text. It was the express purpose of Narrative logic to scrutinise these matters through philosophical enquiry essentially an enquiry into the nature of the narrative form, the nature of narrative knowledge that emerged from it and the role of linguistic instruments in the structure of that knowledge. Thus the possibility, or otherwise, of correspondence between the past and its narrative representation might be determined. But now to my negative comments. This paper is, as I have already indicated, about the originality and permanency of Ankersmits thesis vis-a`vis narrative substances with some contextualising comments about the continuing signicance of its linguistic/textual expression in the current post-textual/post-linguistic days in which we are alleged to be living, aided not least by Ankersmits own progression beyond such linguistic and textual concerns (in Ankersmits case of the sublime kind). This paper is not therefore, and should not be read as, a defence of Ankersmits theorisations. Rather, I am using Ankersmits narrative substance or, to be more precise, two very signicant points which arise from it in order to illustrate the inescapable centrality of language in historical theory; a matter which seems to have been overlooked or pushed aside with the recent substantial growth of interest in memory/experience studies. In short, there is no intended critique of Ankersmit in this essay there is simply an explanation of Ankersmits narrative substance which serves as a vehicle for my argument.

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Rethinking History

553

The paper is divided into three sections. In Section 1 I explain what Ankersmit meant by the term narrative substance and describe its mode of operation. In Section 2, following a selective expansion of the arguments presented in Section 1, I consider what I take to be the central and lasting importance of the narrative substance for historians. And nally in Section 3 I relate the notion of the narrative substance to a wider (and current) philosophical perspective. These prefacing remarks now made, let me outline what I take to be Ankersmits proposal for narrative substances.

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

1.

The narrative substance

Ankersmits argument (as laid out in Narrative logic) is articulated around three propositions or theses, which he himself called his three essential pillars which, briey put, go as follows: (1) The past has no narrative structure within it; any such apparent structures can only occur in the forms of the narratives themselves. (2) Narrative substances are the primary narrative entities in narrative accounts of the past. (3) There exists a relationship between the narrative substance and metaphor and, moreover, the narrative use of language is fundamentally metaphorical. Now, point (1) and the second aspect of point (3) are not my immediate concern in this paper and it will suce to make just the following brief comments about them in order to justify my marginalisation of them at this point. Referring then to point (1), Luis O. Mink long ago (and famously) pointed out the obvious and clear distinction between life and action on the one hand and its narrativisation on the other. That is to say that our thoughts and actions may well be the subject of narratives that we tell ourselves, but the thoughts and actions themselves are not narratives. Or, to put it another way, we may well be the historians of ourselves but not at the moment we act. In short, as Mink puts it, Stories are not lived but told (1987, 60). Referring now to the second part of point (3), the fundamentally metaphorical nature of language is for me a given. The point here is that new or strange phenomena are always put under description in terms of familiar or presumed to be understood previously described phenomena which, in their turn, have been subject to the same familiarisation process, and so on in an apparently endless regression. Hence, it would appear that the language with which we choose to put our human version of the world under description, and thereby deduce our so-called knowledge of it, comprises a complex of receding metaphors all the way down.

554

P.P. Icke

Having made these points I may now proceed with my explanation of Ankersmits concept of narrative substances and I do so by rst answering the question what precisely does Ankersmit mean by narrative substances and where does this idea come from? Well, I think that the best way to get into Ankersmits essential position is to understand how the concept of the narrative substance emerged out of an old idea which Ankersmit had appropriated and then ingeniously reshaped for his own purposes. By which I mean that what Ankersmit so successfully did was to appropriate W.H. Walshs exhausted notion of the colligatory concept, sever it from all its objective referential entanglements, re-describe it as a linguistic instrument and then relocate it in language where it had really always belonged. Colligation, as construed by Walsh, was essentially a form of contextualisation; that is to say that past events are placed in a context through the establishment of a web of interconnections with other events in a manner that reveals the larger historical wholes within which the individual events, as its constituent parts might be thought to have been explained. Walsh expected that colligation would oer the historian a tool of common usage in the sense that it might operate in its own right and independently with regard to the existing positivist and idealist theories of historical understanding which had been essentially articulated through C.G. Hempels Covering Law Model and R.G. Collingwoods Re-enactment Theory.1 Now, a number of historical theorists had, over several years, buried themselves in the nuts and bolts of Walshs theory in what turned out to be an endless and unproductive search for some kind of workable understanding of it. Ankersmit then stepped into the frame with a brilliant idea which nally closed down the debate, for he appropriated Walshs colligatory concept, cut it loose from its problematic direct referential linkage with the past and re-described it under a new name, the narrative substance, as a linguistic device; a device which was, and is, fundamental to the logical structure of narrative accounts of the past. In essence Ankersmits postulation of the narrative substance rests on the observation that, taken in isolation, a constituent statement of a narrative account of the past has a double and not a single function. As a statement it refers directly to some aspect of the past or thing in the past, however, as a component of the narrative viewed as a whole, it serves as a property of a picture of the past. Thus Ankersmit draws attention to the clear distinction between the narrative statements referring to as opposed to its being about some aspect of the past. Where a statement, or indeed a single word, performs this double function within the narrative it might not be clear which of these two functions is operative at any one time. Take, for example, the narrative use of the word Napoleon, which could refer directly to the man himself empirically acquired biographical details perhaps or the word could be embedded in a component of a coherently individuated picture of Napoleon from the unique viewpoint of the narratives author

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Rethinking History

555

who favours a particular interpretation of Napoleon as this or that type of man the authors Napoleon. One could say, then, that there are two subjects intrinsic to the word Napoleon. First there is the narrative subject the actual man about whom veriable statements can be made which refer directly to the past and in the second place the narrative substance or the narrative picture of a man generated within the logic of the text which cannot be subject to truth claims precisely because this picture is gured by the historian within the substance of the narrative itself and does not refer to anything outside it to which correspondence criteria might be applied; it is just an opinion governed by a point-of-view which is not informed by the past itself. Ankersmits subject/substance distinction is of importance because the conation of these two functions can lead to confusion and unjustied claims with regard to the status of the narrative viewed as a whole; claims such as, for instance, the facts are right so the story must be right. Such claims disregard the ontological distinction between the content and the form of narrative accounts of the past. Now, as Ankersmit points out, narrative substances might be associated with specic names (for instance The Renaissance, The Cold War, The Enlightenment and so on), but often this will not be the case. Nevertheless, this does not mean that pictures of the past are not in use when such generally accepted terms fail to appear. Rather, the essential point here is that it is the narrative substance (governed by the historians uniquely individuated point-of-view) which gives rise to and embodies narrative meaning, and that this function should not be confused with, conated with, or reduced to, the capacity of the narratives individual statements to describe and, therefore, refer to the past. On this basis Ankersmit stakes claim to a third logical entity in addition to the subject and predicate of propositional logic and it is his contention that this new entity, the narrative substance, carries the narrative meaning of historical accounts of the past and, as such, constitutes the primary logical entity within those narrative accounts. When a narrative substance acquires a name (for example The Renaissance) and this name subsequently enters into common usage, an erroneous presumption becomes attached to it. That is, it acquires the putative status of a known and denable entity which is assumed to inhabit the past itself; it is taken (as Ankersmit would say) as part of the inventory of the past such that, as a consequence, the clearly demonstrable case that there are as many dierent Renaissances as there are historians who write on the subject is lost from view. Of course in practice the narrative substance (regardless of any discursive familiarity leading to its presumed ontic status) will lack any clear consensual identity because there will always be some measure of disagreement over the appropriate make up of its individuating statements. Furthermore, Ankersmit submits, narrative substances or pointsof-view, in order to possess identity at all, will require the presence of other competing points-of-view. That is to say that identity itself is a relational and

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

556

P.P. Icke

thus relative matter which arises out of dierences; the notion of the existence of just one single point-of-view is arguably oxymoronic because such singularity would constitute a postulated truth and not signify a proposal about it. Consequently, because any particular historians pointof-view will strive to exclude competing points-of-view, it follows that the narrative substance gains its identity negatively by virtue of what it rejects by what it is not. And, crucially important here, it should be noted that competing collections of narrative representations of past events depend for their existence on the various narrative substances within them. Moreover, there could be no historical debate without them for such debate would then be restricted to the empirical level of reference which, on its own, could not constitute a history at all: no narrative substance, no history. Turning now to the rst part of the third pillar of Narrative logic (to recall his argument here that there is a relationship between narrative substances and metaphor) Ankersmit argues that one cannot overlook the conspicuous similarity that exists between the modes of operation of both the narrative substance and metaphor such that the narrative substance can be seen to be of a fundamentally metaphorical kind. He builds his argument to that eect in the following way. Narrative statements (as previously established) have a double function they describe the past (such description being subject to empirical falsication) and they collectively constitute an image of the past (an image, seen from a particular point-of-view which logically cannot submit to any form of falsication). There is therefore a clear analytical distinction between these two quite dierent uses of language, the empirical referential language of description and the aesthetic language of individuation and yet, Ankersmit argues, these two language functions will always be conated within the narrative itself. And, Ankersmit points out, this double function, which typies narrative statements is also typical of metaphorical statements. Take, for an example Shakespeares metaphor All the worlds a stage from As You Like It. The meanings of the words world and stage in isolation (their signication) would, in Ankersmits argument, comprise the literal content of the phrase. However, taken in the whole, the phrase also constitutes an invitation to the reader or audience to visualise the world from a particular point-of-view which, in this case, eects the transfer of the attributes of the (theatrical) stage onto the world. This metaphorical use of language which is crucial for all meaning making thus endows Shakespeares phrase with a surplus of meaning which accordingly exceeds (is in excess of) its literal meanings. Ankersmit refers to this excess as the scope of the metaphor (more of this in a moment), which allows him to make the signicant observation that the metaphor is not part of reality itself but should rather be seen as a linguistic gure which, in common with the narrative form, proposes a particular view about reality. This, then, is the basis of Ankersmits central argument which ties the gure of metaphor to

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Rethinking History

557

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

that of narrative substances and thence to the narrative form. In so doing he draws attention to the most conspicuous logical property shared by the metaphor and the narrative seen as a whole; namely, that their meanings can be said to be carried by their propositional and not by their descriptive modes of articulation. It can be seen from the foregoing that Ankersmit takes the view that (to use his own words) descriptive statements have no narrative compass of their own. It is only the historians unique point-of-view that endows purposefully selected narrative statements with their collective capacity to illustrate that very same governing point-of-view from which those statements were selected. Seen this way, there is an internal circularity which characterises the make-up of all historical narratives and which, accordingly, arms the autonomous nature of the narrative meaning carried by the whole historical text. This is how Ankersmit puts the circularity argument in Narrative logic:
The dependence of separate descriptive statements upon the points of view individuated by them when taken together, their compasslessness when taken in isolation, underlines . . . the circularity so characteristic of historical knowledge. Isolated individual [narrative] statements . . . may indicate all conceivable directions only a narrative point-of-view can give them narrative direction, yet this point-of-view only comes into being thanks to those helpless descriptive statements. (Ankersmit 1983, 218)

Furthermore, with regard to facts and narrative meanings he continues that


. . . the variety of narrative meanings one and the same statement may have in dierent narratives suggests that what the historical facts are (i.e. what is expressed by narrative statements) always depends on what narrative use is made of the narrative statements in question. Thus there are no facts devoid of narrative interpretation in narratives. (Ankersmit 1983, 2189)

Now, the foregoing relativism would appear to permanently preclude the possibility of any denitive selection criteria that might be put to use in order to identify the best of any collection of variously construed historical narratives each purporting to properly (truly) represent the same past event. However, skirting this relativistic conclusion, Ankersmit himself proposed a selection strategy that rested on his notion of scope which, accordingly, he extended beyond the metaphor to encompass the narrative form as a whole (and which somewhat deected attention away from this relativistic diculty). The scope of a metaphorical statement then is, as already explained, wider than its literal descriptive content because its descriptive content, in isolation and lacking a point-of-view, fails to project any familiarising structure of comprehension onto that part of reality in question. Hence Ankersmit concludes that

558

P.P. Icke

. . . it seems only reasonable to say that the most successful metaphorical statements are those in which the dissimilarity between scope and descriptive content has been maximalized. Scope-maximalization is the goal of the metaphorical dimension in language. (Ankersmit 1983, 221)

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Following this line of argument, but now applying it to the narrative as a whole (which, as already shown, possesses meaning in excess of the sum of its constituent descriptive parts), Ankersmit concludes that a narratives merit must, as was the case with the metaphor, turn on the degree to which its scope has been maximalised. Indeed, whenever historical narratives are compared it is precisely their scopes and not their points-of-view that form the basis of that comparison because there is no common ground against which the comparisons of intangible points-of-view could take place. Consequently, Ankersmit contends that the best narrative among competing narratives must be the one that has most eectively and courageously maximalised its scope. To quote Ankersmit directly; Fertility and not truth is our criterion for deciding upon the relative merits of narratives (Ankersmit 1983, 223). Ankersmit is therefore making his preferred selection from competing narrative accounts of past events on the basis of aesthetic considerations alone his preference does not, by any measure whatsoever, turn on considerations relating to empirical truth claims. This, then, is Ankersmits essential thesis with regard to narrative substances and I now want to consider its signicance and consequences for historians. 2. The unavoidable consequences of the narrative substance for the discourse of history Ankersmits central thesis, then, focuses clearly on the sometimes observed yet scarcely worked distinction between historical research and historical writing. The product of historical research is typically expressed in the form of veriable facts and singular statements that refer directly to the past, while the product of historical writing is expressed in the form of narrative proposals about the past that necessarily refer internally to their own narrative substances and not to the past itself. Historical writing, seen as a coherent whole, is therefore an internally referenced imaginative product of an aesthetic kind and as such cannot possibly be subject to validation on the basis of truth claims at the level of the singular statement. So clear does this point appear to be that it is interesting, curious even, to note here bearing in mind that (a) histories are necessarily constituted, and thus carry their meanings, in culturally dependant story forms and (b) that it is undeniably the case that such story forms cannot be found in the past itself that Ankersmits helpful proposition, which identied the narrative substance as the referent of its own manifest story form, received the negative reaction

Rethinking History

559

that it did. After all his proposition presented immediate solutions to a number of problems within historical theory; not least it oered a very plausible explanation for the apparently inexhaustible growth in the production of dierent, sometimes radically dierent but always dierent, historical accounts of a single vanished past a past evidenced only through a nite common base of traces remaining from that single past. The narrative substance thus functions as a kind of lter which processes available facts and narrative statements according to their usefulness with respect to the particular narrative meaning intrinsic to the narrative substance in question. Or, to rephrase, it is from the point-of-view which governs the historians chosen narrative substance that a preference for a particular fact or statement is expressed. This preference will be for facts and statements which contribute to the individuation of that very same narrative substance or point-of-view which, in turn, is itself an expression of the historians predisposition to a story form of a particular kind. Hence in this circular fashion Ankersmit shows that the writing of history is an autonomous activity in the extent to which narrative meaning (meaning carried by the narrative viewed as a whole) is referenced internally to its own point-of-view rather than to the past itself. On this understanding it follows that the facts and singular statements arising out of historical research do not, and cannot, somehow gure forth the truth about the past. Rather, they constitute the reservoir of raw unprocessed (and therefore historically meaningless) materials from which the historian makes a preferred selection in order to individuate a pregured notion about the causality/direction/ meaning and so on, of the past. Having imagined and individuated a history, coherently articulated about its own governing but arbitrary narrative substance, some might see t to defend that meaning as if it were part of the past itself and not a projection onto it of an imaginary narrative proposal about an ultimately unfathomable past which, of course, never possessed a narrative form in the rst place. To complete this section of the paper, I now want to highlight the two most signicant points2 which, in my view, logically follow from Ankersmits position in relation to narrative substances. My rst point is that (on Ankersmits view and mine) the autonomous linguistic construct that we call history cannot provide access to any meaning, direction or purpose located in the past itself; indeed, the discourse of history has little at all to do with the past in that sense. What emerges from Ankersmits argument is therefore the defensible proposition that historians, for purposes of their own, linguistically generate and then project imagined meanings onto a past that neither notices nor cares about them. History is one way trac nothing ows back from the past but the echo of the imaginative gure of history itself. There is, in short, an immutable disconnect (a dissonance) between history and the past which history purports to adequately represent. It therefore follows that the notion that

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

560

P.P. Icke

history could constitute some sort of bridge to the actuality of the past is fundamentally awed and cannot be sustained. And my second point, concerning narrative meanings and values, is that it should be understood that the facts and statements which comprise a narrative proposal about the past, taken in isolation, are in a manner of speaking vacant; that is to say that as empty signiers they are devoid of narrative meaning and direction, or compassless (to quote Ankersmit). It follows that it is only relative to their appropriation and organisation into narrative form under the governing rationale of a narrative substance that facts and statements have their narrative dependant (and hence relative) meanings and values impressed onto them in accordance with the synthetic narrative point-of-view which they have been purposefully selected to individuate. Or, in short, the meanings and therefore the values attached to historical facts/statements are to be found only in relation to their uses when organised into narrative form. Furthermore, even if it were possible (and it is not) to nd the denitive history of some past event a kind of transcendental history which would embody the capacity to reject all competing accounts of that same event and place the undisputed intrinsic meaning of that past event before us all there yet remains an insurmountable problem. That is (as can be deduced from Ankersmits argument) that any happening or fact, either past or present, does not entail, intrinsically within itself, a xed value which might lead to a xed ethical imperative or response to that happening or fact. Values/meanings are not in happenings/facts; they are always projected onto them in relation to enculturation, ideology, context (and so on) which, taken together, constitute a point-of-view: value and meanings are indisputably a relative matter. It follows therefore, that since values are neither in nor entailed from facts, that any perceived imperative or actual response to the facts of a situation or happening can be seen to be resting on a wholly arbitrary rhetorical position. Thus, in any particular instance, one might ask whose values are they that are in play, how and why are such particular values privileged above all others and in whose interests do they operate? Understood in relation to the above two points, the writing of history and the shape or structure of the meaning that it carries can be seen to arise out of an arbitrary unsanctioned organisation of the past an internally functioning organisation which draws both on its own narrative substance(s) and on language and linguistic devices in order to generate its manifest (and meaningful) story form. Accordingly, Ankersmit sees such organisation of the past as a violation of the past . . .
. . . the [t]elling of the past . . . is unavoidably a violation of that past in order to eect such a narrative organisation of the past an organisation that is not ska 1998, 78). intrinsic to the past itself. . . . (Doman

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Summing up, one could conclude that the evidential traces of the contingent actions and situations which once constituted the actuality of some past event

Rethinking History

561

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

are drawn/gured together into some or other historical order by the historian through the imposition of some or other story form upon them. This story form is a linguistic construct informed solely by its own narrative substance. It is a product of the enculturation of its own time and it cannot be found in the past itself. The past is thus violated by the very act of historicisation; that is to say that the past is pushed around, beaten into shape and suitably appropriated in order to t an arbitrary story form which is retrospectively imposed on it whence to serve some human purpose to which the past itself is indierent. Moreover, this process of constituting the historical past backwards, which Hayden White sees as a process of retrospective ancestral constitution, is (as he puts it) what historical consciousness is all about. Eliminate it and historical systems would not exist at all.3 For the nal section of the paper I am now going to relate these matters back to a more general philosophical position with regard to the historicisation of the past. A position which eectively precludes the possibility that these matters might be put aside or discounted (not least by Ankersmit himself) through arguments which seek to dismiss the linguistic and turn and hence the narrative substance in historical theory as passe irrelevant. And, moreover, that in its wake we should now embrace forms of real presence and real experience which are seen in some quarters to lift history out of the realms of both the empirical/epistemological and the linguistic/textual into the new light of direct historical experience(s).4 My argument is that Ankersmits own position in relation to narrative substances, as I have outlined it here, is so strong that such a negation of it (that is through bypassing/superseding the linguistic/textual or, indeed, ignoring it altogether) is a massive mistake. These considerations of course open up an extensive discursive eld which is beyond the scope of this essay and it is for this reason that, on this occasion and in order to best illustrate my point, I intend to narrow that eld by invoking some pertinent arguments which I have extracted from Martin Daviess recent publication Historics (2005). It should again be understood that it is not my purpose here to comment on or critique Daviess overall thesis; I am simply appropriating, or stripping from it (in a Rortian sense), whatever is of use in the furtherance of my own argument. 3. A broader philosophy

In his book Historics, Davies (2005) articulates an uncompromising critique of the historicised culture in which we humans live (the already historicised world) and he censures the generally accepted condence in historical knowledge as a foundational discipline in the humanities. He does not comment much on the fundamentally illusory nature of the historical discourse; this is for him a given a starting point or presupposition for which there already exists a vast body of supportive and persuasive

562

P.P. Icke

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

literature (White 1973, 1978; Ricoeur 1984, especially 91174; Ankersmit 1983; Cohen 1986; and many others). Rather, taking all this as given, Davies examines the consequences of living in a world which is obsessed with the notion of historical knowledge to the extent that this notion in itself comes to underpin both our individual and collective understandings of identity and our understanding of society. For the purpose of this argument I turn to the last section of Variation Four of Historics, entitled History as symbolic re-enactment wherein Davies focuses on the central function of tradition and legacy (embodied in our already historicised world) in the legitimisation of current practice or, in short, the function of history and its symbolism as an armative ideology. Within our (human) historicised world, Davies argues, there is nothing which cannot become or has not already become a historical symbol the past is all around us in our historic environment which is a collective memory containing stories written in stone, brick, wood, glass, steel; stories inscribed in eld patterns, hedgerows, designed landscapes and so on . . . (Davies 2005, 233). Accordingly, the past which no longer exists can only (putatively) make itself known through some kind of existing historical text within the historicised semiotic system. There is, therefore, no extra-textual referent for the historical text, no knowable objective reality, nothing beyond the semiotic-system which describes it. Davies, quoting Michael Oakeshott, is in agreement with Oakeshotts notion that . . . the illusion of an extra-textual referent boils down to a purely self-referential textual coherence (Davies 2005, 234). Thus humans, as symbol using language animals, consciously operate reexively through systems of representations sign-systems (notations) of human making without which human objective reality would be unthinkable and thus unknowable. This world of words creates the world of things or, as Hayden White expressed it in his Introduction to Tropics of Discourse, . . . tropics [meaning tropological language use] is the process by which all discourse constitutes the objects which it pretends only to describe realistically and analyse objectively (White 1985, 2). Returning again to Historics, Davies, drawing on George Steiner, continues
. . . the past is thinkable and knowable only through the current semiotic or symbolic system. Our sense of the past, not as immediately, innately acquired reexes, but as a shaped selection of remembrance, is radically linguistic. History, in the human sense, is a language-net cast backwards. . . . which reinforces the illusion of an objective past as a substantive extra-textual referent. (Davies 2005, 235)

The past (to paraphrase Davies again) takes extra-textual priority because the historicising reex is already intended in the texture of signs and symbols that is, in the grammatical structures that make up the

Rethinking History

563

language-system. Whatever the historical reex produces is, in some form or another, already known already there which is why history has nothing new to teach us. Consequently, what history authorises Davies characterises as
. . . the constant reproduction of the already existing, morbid culture of traumatic experience, phantom memories and petried objects. It sanctions a mentality that defers automatically to a traditional value-system that puts a selection of dead saints, cardinals, archbishops, and earls before the generations of the living. It thereby sustains a pessimistic, misanthropic ideology that always has seen prelates and nobles as the repositories and guardians of conservative truths. (Davies 2005, 235)

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

One day (Davies concludes) historys fatal arrogance might provoke opposition: the dissociation of humanity from history, the institution of a new language and a new thought-style for the humanities (Davies 2005, 235). It is against this broad backdrop that I can now more readily position Ankersmits narrative substance. For, seen as a self referencing creative linguistic instrument an enabling device which both constitutes and explains its object of historical enquiry internally in relation to itself it ts into the general framework of thought exemplied by Davies in his book Historics. This framework is articulated about the understanding that we humans live in a rhetorical world of human making which is constituted in the language we use to describe it. Any thing that lies outside our capacity for its linguistic description could not be linguistically related to other things and thus could not (as Richard Rorty argues) be talked of or thought of. Or, as Rortys Nietzsche points out,
. . . that things posses a constitution in themselves quite apart from interpretation and subjectivity, is a quite hopeless hypothesis; it presupposes that interpretation and subjectivity are not essential, that a thing freed from all relationships would still be a thing. (Rorty 2007, 111)

Thus, to step outside a language of description is to step into an ineable void. A void which, in a converse sense, nds itself in serious want of a language of description (and any familiarising language would do, providing that it has sucient utility value to allow us humans to get around in it). There is, then, no escape from language and I therefore cannot agree with Michael Roth who, in his recent indexical review entitled Ebb tide (2007), maintained that following the demise of the linguistic turn
. . . the massive tide of language. . . . has receded . . . [and] we are now able to look across the sand to see what might be worth salvaging before the next wave of theory. . . . begins to pound the shore. As language recedes. . . . etc/ etc. (Roth 2007, 66)

For here the point is that arguably the medium of language does not (nor cannot) itself recede it is not (in Roths sense) tidal: language cannot

564

P.P. Icke

succumb (to use his metaphor) to the moons gravity and thereby leave us beached outside its reach in an ineable void. As symbol using language animals we are always continuously and totally submerged in language like it or not, that is how it is. Ankersmits narrative substance, then, understood as a linguistic instrument, both uses and inescapably resides inside Steiners language net. And when that net is guratively speaking cast backwards the nature of the historical discourse which emerges from it can, in essence, be explained in relation to the matrix of the narrative substance which, in the rst place, draws attention to and arms the fundamental and categorical structural divide between the past and its narrative representation and, second, suggests an autonomy of narrative function which can be shown to be primarily governed by a narrative logic informed by linguistic rules and devices rather than the actuality of the past itself. The historical narrative cannot therefore be taken as some sort of window through which the actual past might be viewed exactly as it was. Rather, the historicisation of the past has to be seen as a proposal (as a representation of an aesthetic kind), cut adrift from the past itself because it is unavoidably mediated through a complex of self-governing narrative structures which, stricto-sensu, lack any points of reference located in the past. Moreover, it should also be noted that, as a proposition, the historical narrative rests on an argument (as all propositions do) and an argument can neither be true nor false arguments can only be valid or invalid. Thus the notion of truth, in this sense, has no bearing on or relevance to the debate on the function and nature of the narrative form in historiography. In conclusion I oer two brief endnotes, the rst relating to the substance of my central argument and the second constituting a reexive comment on the nature of all arguments my own included. Point One: Although hinted at in this essay rather than being argued for at length, I think that Ankersmits own recent move beyond the linguistic and the textual is prohibited by his own theory of narrative substances. That is to say that if his theory/argument is valid, and I have argued here that it is, then this best Ankersmit (as dened by his own notion of the narrative substance) is so good that various eorts (including his own) to go beyond it cannot succeed and hence his own recent moves towards experience and away from language are destined to be still-born. This matter, which has a bearing on the many contemporary moves towards experience and presence is not, of course, the subject of this paper. For, as I have explained, it is not my primary intention here to critique Ankersmit, but rather to use the Ankersmit of Narrative logic (with particular reference to his narrative substances) to rearm the centrality of language in historical theory. If I have made my point, then the upshot here is that this best Ankersmit is not only good enough for us all but that he is also (in relation to his excursions into historical experience) too good for his own good.

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Rethinking History

565

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Point Two: This is an elementary one in that it might be assumed of any academic paper. Namely, that of its relative positioning which, given all that Ankersmit has to say about the relativistic nature of any point-of-view, needs some comment here. With regard to my own point-of-view, then, all the foregoing represents my current position on the matters that I have raised and this position, in a Rortian sense (once again), is constituted within the structure of my own end language. Now, some have misunderstood Rortys end language concept and, accordingly, they nd in it a contradictory foundationalist (or xed) position resting uneasily within a broader relativist philosophy. But this argument is awed, for an end language is not xed. Rather, it continuously reshapes in relation to new rhetorical proposals (new metaphors) and thus any end language can only aord a temporary or quasi-foundation for argument pending better metaphors to come. In this sense no argument can ever reach closure and this consideration also applies to, and hence governs, the status of the arguments (albeit arguments which are to me the best that are available right now) which I have put forward within this paper. For, as I have already pointed out, even the very best arguments can only be expressed in relation to the always temporary rhetorical positions adopted by their authors. Seen thus, no argument can possibly transcend its permanent and inescapable status as work in progress. The historical world is a world without closure. Acknowledgements
I would like to thank Keith Jenkins for his critical evaluation of my argument and his invaluable support and enthusiasm for it. I also want to mention Jonathan Coope, whose constructive comments following his reading of an early draft of this paper have helped to shape its nal form.

Notes on contributor
Following his 31-year career in the world of Civil Aviation as an Airline Pilot, Peter Icke returned to the far more hazardous world of academia; he is currently completing a PhD thesis on Frank Ankersmit under the supervision of Professor Keith Jenkins at the University of Chichester.

Notes
1. 2. For Hempels Covering Law Model see Hempel (1965) and for Collingwoods Re-enactment Theory see Collingwood (1994, 282302). Also see Walsh (1942, 1967) for a more detailed description of colligation. While Ankersmit would (I think) be in agreement with my rst point, as outlined here, it is unlikely that he would be in agreement with my second point because, in contradiction to his earlier theorisations, it is his current belief that there exists within the historical narrative a continuity between fact and value. Ankersmit accordingly nds that . . . historical representation truly presents us with the much sought-after trait dunion between the is and the ought. We begin with merely a set of true statements and move then, automatically and naturally, toward an answer to the question of how to act in the future. The transition is

566

P.P. Icke
completely natural, and at no stage can we identify a point where pure knowledge becomes pure action . . . fact and value, the is and the ought, are merely the extremes on a continuous scale . . . (Ankersmit 2001, 94). Now this position of Ankersmits, briey suggested in Narrative logic and expanded in subsequent publications (most recently during an interview with Ranjan Ghosh, see Ghosh 2007) appears to me to embody the rm assumption that there is a denite link (an entailment), of a xed and permanent kind, between fact and value. For, in order that true statements might move automatically and naturally toward an answer to the question of how to act in the future those true statements must have xed values embedded within them a multiplicity of possible values could only suggest a multiplicity of possible actions and thus fail to answer the question of how to act in the future. But, the idea that facts have values xed within them is at variance with Ankersmits own central concept concerning narrative substances which relies on the particular understanding that facts and individual statements are, as he puts it in Narrative logic . . .compassless when taken in isolation . . . only a narrative point of view can give them a narrative direction . . . the variety of narrative meanings one and the same statement may have in dierent narratios suggests that what the historical facts are . . . always depends on what narrative use is made of the narrative statements in question. Thus there are no facts devoid of narrative interpretation in Narratios (Ankersmit 1983, 21819). It seems, therefore, that for the purpose of his seminal narrative substance argument, Ankersmit took the relativistic view that the meanings, and therefore the values, attached to historical facts are to be found only in relation to their uses when organised into narrative form. However, he contradicted and undermined this earlier position when he proposed his more recent continuous scale fact/value argument which can only operate on the basis of a single xed relationship between any particular fact and value. Such a foundationalist position would not rest easily with my own position as expressed here. This view is expressed by both Hayden White and Hans Kellner. See the Introduction in Re-guring Hayden White (Kellner 2009). For example the works of the Groningen School (that is, works mostly emanating from or inspired by the Centre for Metaphysics, Groningen, which was set up by Frank Ankersmit and Eelco Runia to study historical experience and related phenomena). For general coverage of the subject see essays by Runia (2006), Ankersmit (2006), Bentley (2006) and others (History and Theory 45, no. 3).

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

3. 4.

References
Ankersmit, F.R. 1983. Narrative logic: A semantic analysis of the historians language. The Hague: Martinus Nijho. Ankersmit, F.R. 2001. Historical representation. Stanford, CA: Stanford University Press. Ankersmit, F.R. 2006. Presence and myth. History and Theory 45, no. 3: 32836. Bentley, M. 2006. Past and presence: Revisiting historical ontology. History and Theory 45, no. 3: 34961. Cohen, S. 1986. Historical culture. Stanford, CA: University of California Press. Collingwood, R.G. 1994 [1936]. The idea of history. Oxford: Oxford University Press. Davies, M.L. 2005. Historics: Why history dominates contemporary Society. London: Routledge. ska, E. 1998. Encounters: Philosophy of history after postmodernism. Doman Charlottesville: University Press of Virginia.

Rethinking History

567

Downloaded by [Universidad Autonoma de Barcelona] at 01:04 03 April 2014

Ghosh, R. 2007. Interdisciplinarity and the doing of history: A dialogue between F.R. Ankersmit and Ranjan Ghosh. Rethinking History 11, no. 2: 22549. Hempel, C.G. 1965. Aspects of scientic explanation and other essays in the philosophy of science. New York: Free Press. Kellner, H. 2009. Introduction. In Re-guring Hayden White, ed. F.R. Ankersmit, E. Domanska and H. Kellner, 18. Stanford, CA: Stanford University Press. Mink, L.O. 1987. Historical understanding. Ithaca, NY: Cornell University Press. Ricoeur, P. 1984. Time and narrative, Vol. 1. Chicago, IL: University of Chicago Press. Rorty, R. 2007. Philosophy as cultural politics. Cambridge: Cambridge University Press. Roth, M.S. 2007. Ebb tide. History and Theory 46, no. 1: 6673. Runia, E. 2006. Spots of time. History and Theory 45, no. 3: 30516. Walsh, W.H. 1942. The intelligibility of history. Philosophy 17, no. 66: 12843. Walsh, W.H. 1967. An introduction to philosophy of history. London: Hutchison. White, H. 1973. Metahistory: The historical imagination in nineteenth-century Europe. Baltimore, MD: John Hopkins University Press. White, H. 1978. Tropics of discourse. Baltimore, MD: John Hopkins University Press.