+ All Categories
Home > Documents > Ankersmit on historical representation. Resemblance, substitution and exemplification

Ankersmit on historical representation. Resemblance, substitution and exemplification

Date post: 12-Dec-2016
Category:
Upload: chiel
View: 213 times
Download: 1 times
Share this document with a friend
19
This article was downloaded by: [Fordham University] On: 14 September 2013, At: 15:51 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 Ankersmit on historical representation. Resemblance, substitution and exemplification Chiel van den Akker a a History Department, VU University Amsterdam, The Netherlands Published online: 15 Aug 2011. To cite this article: Chiel van den Akker (2011) Ankersmit on historical representation. Resemblance, substitution and exemplification, Rethinking History: The Journal of Theory and Practice, 15:3, 355-371, DOI: 10.1080/13642529.2011.588521 To link to this article: http://dx.doi.org/10.1080/13642529.2011.588521 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, sub- licensing, systematic supply, or distribution in any form to anyone is expressly
Transcript

This article was downloaded by: [Fordham University]On: 14 September 2013, At: 15:51Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,UK

Rethinking History: The Journalof Theory and PracticePublication details, including instructions for authorsand subscription information:http://www.tandfonline.com/loi/rrhi20

Ankersmit on historicalrepresentation. Resemblance,substitution and exemplificationChiel van den Akker aa History Department, VU University Amsterdam, TheNetherlandsPublished online: 15 Aug 2011.

To cite this article: Chiel van den Akker (2011) Ankersmit on historical representation.Resemblance, substitution and exemplification, Rethinking History: The Journal ofTheory and Practice, 15:3, 355-371, DOI: 10.1080/13642529.2011.588521

To link to this article: http://dx.doi.org/10.1080/13642529.2011.588521

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all theinformation (the “Content”) contained in the publications on our platform.However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, orsuitability for any purpose of the Content. Any opinions and views expressedin this publication are the opinions and views of the authors, and are not theviews of or endorsed by Taylor & Francis. The accuracy of the Content shouldnot be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions,claims, proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly in connectionwith, 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, sub-licensing, 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

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

Ankersmit on historical representation. Resemblance,

substitution and exemplification

Chiel van den Akker*

History Department, VU University Amsterdam, The Netherlands

Ankersmit has argued that the substitution theory of historicalrepresentation should be preferred to the resemblance theory. Historicalrepresentations do not resemble the past; they take the place of theabsent past and function as a substitute for it. Contrary to Ankersmit, Iwill argue that there is no reason to prefer the substitution theory to theresemblance theory. Moreover, the substitution theory cannot accountfor the relationship between historical representations and the past. Athird theory, the exemplification theory of representation, will be putforward to account for that relationship.

Keywords: Ankersmit; Danto; representation; epistemology; identity;narrative

Introduction

According to Frank Ankersmit, historical representations take the place ofthe absent past by functioning as a substitute for it (2001, 80–1). In this,Ankersmit is not alone. Other philosophers of history favour thesubstitution view on historical representations too (Munslow 2003, 158).Although Ankersmit’s notion of historical representation has beenscrutinized (Zammito 2005; Zelenak 2009), his substitution view ofrepresentation has not. This is remarkable, since the substitution viewgoes to the heart of Ankersmit’s view on historical representation.

Ankersmit argues for the substitution theory by comparing it to its rival,the resemblance theory, according to which a representation resembles whatit represents. I will distinguish six arguments that purport to show that thesubstitution view is to be preferred to the resemblance view. I will agree withnone of them. Moreover, I will argue that the substitution theory cannotaccount for the relationship between historical representations and the past.This does not mean that discussing Ankersmit’s arguments is nothing otherthan a scholastic exercise to keep our minds busy for as long as it lasts.

*Email: [email protected]

Rethinking HistoryVol. 15, No. 3, September 2011, 355–371

ISSN 1364-2529 print/ISSN 1470-1154 online

� 2011 Taylor & Francis

DOI: 10.1080/13642529.2011.588521

http://www.informaworld.com

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

On the contrary; there is much to learn even by simply trying to penetratethe depths and richness of Ankersmit’s arguments. We will be coveringimportant ground, discussing problems of epistemology, identity andnarrative.

Our discussion of the substitution and resemblance theory of representa-tion will not limit itself to Ankersmit’s treatment of the subject. Ankersmit’smain influence is Arthur Danto’s theory of representation. Danto, in turn, ismainly influenced by Nelson Goodman. Therefore, Danto and, to a lesserextent, Goodman will also figure prominently in this essay.

The rejection of the substitution theory of historical representation willnot result in a rehabilitation of the resemblance theory of historicalrepresentation. Therefore I propose a third theory: the exemplificationtheory of representation. According to the exemplification theory, therepresented becomes an example of what the work expresses. A medallionrepresenting Louis XIV as the sun and expressing omnipotence turns LouisXIV into an example of an omnipotent king. This theory of artisticrepresentation is developed by Goodman. Danto favours this theory as well,and that is something Ankersmit somehow missed. I will argue that theexemplification theory of representation applies also to works of history.

In the conclusion I will point out that the exemplification theorynowhere contradicts Ankersmit’s narrative philosophy of history as putforward in his Narrative logic (1981). It is even better suited to account forthat narrative philosophy than his own substitution view.

1. Representation and epistemology

Ankersmit’s first argument to prefer the substitution theory to theresemblance theory is that words in contrast to figurative paintings, donot resemble what they represent (2001, 222). This is, of course, true, but amoment’s thought will make us realize that this argument cannot be used tofavour the substitution view, for words are no substitutes for what theyrepresent anymore than words resemble what they represent. The same istrue of strokes of paint. Strokes of paint do not function as a substitute forwhat they represent either, just as paint does not resemble what itrepresents.1 I do not pick an apple from a painted bowl of fruit, nor do Ithink that apples consist of red and yellow paint. We are clearly on thewrong track and using the notions resemblance and substitution in anerroneous way if we wonder whether words and paint resemble or functionas a substitute for what they represent.

Let us for the moment adopt a more helpful way of using both notions.Resemblance is a relation between representation and represented; therefore,if words resemble what they are about, it must mean something like this: theimage or idea that words invoke resembles what those words represent.Similarly, the image that strokes of paint invoke resembles what those

356 C. van den Akker

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

strokes of paint represent. A representation consisting of words does not onlyrepresent objects and events in the world but may represent words used byothers too. In case of a representation of a speech act, resemblance may meansomething like this: the content of the words (representation) resemble thecontent of the speech act (represented). In these instances, resemblance isused in such a way that words do resemble what they represent. In a similarhelpful way we may use the term substitution. The content of words is asubstitute for the content of the represented speech act and the image thatwords and paint invoke is a substitute for what those words and strokes ofpaint represent. This is how Ankersmit seems to be using the notion ofsubstitution. Before we decide whether the helpful way of using the notionresemblance is a viable alternative to Ankersmit’s interpretation, we need tolook at his second and third argument first.

According to Ankersmit’s second argument favouring the substitutiontheory of representation to the resemblance theory, the substitution theory iscompatible with the resemblance theory while the reverse is not, for:

there is no inconsistency in saying (1) that all representation is substitution and(2) that in pictorial representation resemblance is sometimes (or perhaps evencustomarily) relied upon in order to achieve a believable substitute for therepresented [because of ‘certain (naturalist) traditions or conventions ofpictorial representations’]. On the other hand, one cannot do the opposite andassert (1) that the resemblance theory is correct and (2) that texts,nonfigurational art, or parliaments ‘represent’ a represented. (Ankersmit2001, 223)

We can agree with Ankersmit that parliaments do not resemble the peoplethey represent, but that is not the resemblance that is at stake here, for thelaws of parliament resemble the laws that the people would make. That iswhat parliaments are all about. Thus, the resemblance and the substitutiontheory are compatible in this case. We can also agree with Ankersmit thatnon-figurative art does not represent objects, so there is nothing for them toresemble. This is a problem for the substitution theory too, for if there isnothing for non-figurative art to represent or resemble, there is nothing forthem to be a substitute for either.2 Thus, in this case the question ofcompatibility does not even arise. This leaves us with the question ofwhether texts resemble what they represent. This is Ankersmit’s firstargument all over again. Apparently, texts or words taking the place of therepresented by functioning as a substitute for it is regarded as a normal andsensible thing to say, while saying that texts or words resemble what theyrepresent is considered to be nonsensical. However, words do not take theplace of the represented anymore than words resemble what they represent.Moreover, if we say that words do not resemble what they represent, we arealso obliged to say that paint does not resemble what it represents. In thatcase, the question of compatibility does not arise either. If we useresemblance and substitution in the more helpful ways that I have

Rethinking History 357

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

suggested, then the resemblance theory is compatible with the substitutiontheory (and vice versa). Hence, we have no reason to prefer the substitutiontheory to the resemblance theory.

My critique of Ankersmit’s first two arguments only purports to showthat those arguments cannot be used to show that the substitution theory isto be preferred to the resemblance theory of representation. I do not want tofavour either of them, as will become clear in the rest of this essay.

What we should note is that the more helpful way of using the notion ofsubstitution degrades the notion to that of standing for. This is what seemsto be happening in the quoted passage of Ankersmit above. If so, the notionof substitution is identical to that of representing and for that reason, isredundant, since representing means standing for. The helpful notion ofsubstitution, then, is not helpful, after all.

We should also note that the helpful way of using the notionresemblance gives rise to certain epistemological problems, for the relation-ship between mental representations (such as words invoking images) andthe reality that those representations resemble is what epistemology is allabout. Here, we arrive at Ankersmit’s third argument. This argument willalso turn out to be unconvincing.

According to the third argument, the disadvantage of the resemblancetheory is its tendency ‘to reduce problems of representation to thosetraditionally investigated by epistemology’. If we want to explain howresemblance accounts for the relation between the representation and therepresented, then resemblance will be ‘the representational analogue of thecriteria of truth that epistemologists always discuss’ (Ankersmit 2001, 223).To this, Ankersmit is opposed. According to him, the statue of MarcusAurelius at the Piazza del Campidoglio3 and the historical being, Aurelius,both belong to the inventory of the world. They have the same ontologicalstatus (Ankersmit 2001, 11 and 81). It follows that: ‘the relationship betweenthe represented and its representation – a world-to-world relationship –could never be modelled on the relationship between world and language(which is the exclusive domain of interest for the epistemologist)’(Ankersmit 2001, 225). Aurelius and his statue can both be described.This relationship between language (description) and the world (e.g. humanbeings, dead or alive, statues) is the epistemologist’s domain of interest.Thus far, we need not disagree, but we should not forget the differencebetween Aurelius and his statue: a statue is a representation and Aurelius isnot. Calling the relationship between Aurelius and his statue a world-to-world relationship, then, is misleading, for this world-to-world relationshipis asymmetric: Aurelius is not related to his statue as his statue is related toAurelius. The human being, Aurelius, is a necessary condition of his statueand that is obviously not the case the other way around. It follows that thehuman being, Aurelius, and his statue do not have the same ontologicalstatus.

358 C. van den Akker

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

Certainly, the statue is an object and belongs as such to our ontology,but as a representation, the statue has an ontological status only in so far asit functions as a representation, and this functioning is an event rather thanan object. Therefore, the relation between Aurelius and his statue is not aworld-to-world relationship, but a world-to-representation relationship, andthis is the relationship that epistemologists worry about. The relationshipbetween the represented and its representation thus can be modelled on therelationship between world and language, contrary to what Ankersmitthinks. Moreover, we do not have to make a difference between language(description) representing reality and representation such as statuesrepresenting reality to surpass epistemological worries. As long as werealize that representing reality is a matter of habit and not of fit, traditionalepistemological worries need not worry us.

Resemblance gives rise to epistemological worries, since it can beunderstood as a measure of realism, but if realism of representation is amatter of habit, a traditional standard of representation of some culture at agiven time, these worries do not arise, for then we are no longer bothered bythe realist assumption that there is a ready-made or antecedent world forour representations to fit or resemble. Resemblance is not a relation betweena ready-made world and a representation (Goodman 1976, 37–9). The morehelpful way of using the notion resemblance turns out to be unhelpful aswell. The images that words or strokes of paint invoke do not resemble anantecedent or ready-made world.

Two issues must now be dealt with. First, if there is no reality to fit orresemble our mental and material representations, should we not abandonthe resemblance theory of representation altogether? Second, if representingreality is a matter of habit, may we still call the relationship betweenAurelius and his statue a world-to-representation relationship? I will dealwith these issues consecutively.

First, wemust realize that to see representing reality as amatter of habit asGoodman does, is in agreement with the view put forward by Ankersmit andDanto that a representation gives us the represented (Ankersmit 2001, 82; seealso Section 2 below). We only have representations and no ready-madeworld (that is or has been) with which to compare our representations. It doesnot follow that we cannot say that a representation resembles aspects of thepast or the world. I may still say that the statue of Aurelius resemblesAurelius, not because I have travelled back in time to compare the statue withthe real Aurelius or because I know of other statues representing Aurelius; Ionly mean to say that the statue resembles those other statues. I may say thatthe statue of Aurelius resembles Aurelius because it is in agreement with mystandard of what counts as a realistic representation. These standards varyfrom culture to culture and from time to time.

The realist might object by showing his hands and kicking a stone.Indeed, here we have two chunks of reality that we may resemble and

Rethinking History 359

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

imitate by showing our hands and kicking our stones, but resembling andimitating is not representing! A painter picturing a bowl of fruit in all itsfinesse aims at representing the bowl of fruit; he will not merely aim atresembling or imitating its colours and forms. A dancer trying to represent aswan only by resembling or imitating swan-like behaviour will probably failto do so. For one thing to resemble another, the two things and someone tonote the resemblance must be present. For one thing to represent somethingelse, only the first and someone to interpret it as a representation – the eventI talked about above – need to be present. The point is this. The concept ofresemblance apart from the concept of representation establishes arelationship between two actual things, while the concept of resemblancein relation to the concept of representation illuminates the relationshipbetween the representation and the represented more or less successfully,depending on whether or not it is in agreement with some standard ofrealistic representation. Resemblance is not a condition but a possibleconsequence of representation.

We no longer need to hesitate in calling the relationship between therepresented and the representation a world-to-representation relationship.Representations represent a represented and understanding a representationis understanding the relationship between the representation and therepresented, even if the latter is absent and given by the former. It is truethat Aurelius and the past precede their representations, but it is equally truethat Aurelius and the past do not precede their being represented.

Let us return to Ankersmit’s arguments favouring the substitution viewof representation to the resemblance view.

2. Representation and identity

Danto makes a distinction between representation in the sense of beingpresent again, as, for example, Dionysus was believed to be present inperson during certain rituals, and representation in the sense of standing inthe place of something else, as, for example, representatives of parliamentstand for us (Danto 1981, 19). According to Ankersmit, the resemblancetheory cannot account for representation in the sense of being present again,for in this sense of representation, the representation and the represented areidentical, while resemblance presupposes non-identity: only two differentthings can be said to resemble one another (Ankersmit 2001, 226). Thisholds for the substitution theory as well. Something can only function as asubstitute for something else. It is nonsense to say that Dionysus functionsas a substitute for himself whenever he was present at some ritual. Both theresemblance and the substitution theory cannot account for representationin the first sense of being present again. Furthermore, in all argumentspreceding and following this section, representation is used in the secondsense of standing for something else, as symbolizing. Resemblance and

360 C. van den Akker

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

substitution are only of use, if at all, in the second sense of representation.We still have not found any reason to prefer the substitution view ofrepresentation to the resemblance view.

The question whether there is resemblance and/or substitution wheneverthere is a representation, raises, again, a problem for the resemblancetheory. Ankersmit:

For think of Gombrich’s renowned hobbyhorse. The average table will moreclosely ‘resemble’ a real horse than a hobbyhorse, and yet, as Gombrich sorightly insisted, the hobbyhorse more successfully represents a real horse thana table (for the simple reason that the hobbyhorse may function in the eyes ofthe child as a real horse, whereas the table will not). (Ankersmit 2001, 226)4

I do not see why a table more closely resembles a real horse than ahobbyhorse. Presumably a table is less successful in representing a horsebecause galloping with a table is more difficult than galloping with a stick (orhobbyhorse). This cannot be decisive, since there is no reason to assume thata table cannot function as a horse and is less successful therein than a stick (orhobbyhorse).5 As Ernst Gombrich points out, any ridable object would do(Gombrich 1978, 4). This is precisely what Danto’s theory of representationfollowing Goodman all is about: anything can be a representation ofanything else (Danto 1981, 72–3; Goodman 1976, 5; Ankersmit 2001, 227).6

Just as anything can represent anything else, anything can resembleanything else in indefinite ways. Even a fly representing the World TradeCentre, to use Danto’s example, looks like the World Trade Centre in atleast some respects. The point is this: whenever there is a representation, wecan ask ourselves whether concepts such as substitution and resemblancecan help us to illuminate the relation between the representation and therepresented. If there is a representation, I am certain that we will always becapable of finding some resemblance between the representation and therepresented, for anything is like anything else in at least some respects. Theresemblance thus found need not be illuminating. It probably only is inthose cases where the representation is considered to be realistic in the sensegiven to it above.

The substitution view is not much better equipped to deal with casessuch as a fly representing the World Trade Centre. Since a representationstands for the represented, representations always, in a sense, take the placeof something else. A fly may take the place of the World Trade Centre, butthe notion of substitution is hardly illuminating in this case. It merelyrephrases the fact that there is a representational relationship in the firstplace, for in what sense does the fly function as a substitute for the WorldTrade Centre other than standing for/representing it? In most cases,substitution amounts to nothing more than ‘standing for’. If so, thesubstitution theory of representation cannot explain the relationshipbetween the representation and the represented. Below, I will point out

Rethinking History 361

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

that some types of representation that do function as a substitute for whatthey represent. Historical representations will not be one of them.

For a proper understanding of the notion of representation, it isimportant to note that resemblance is neither a sufficient nor necessarycondition of representation (Goodman 1976, 5). A triangle representing apyramid does not do so because it sufficiently resembles one. Since anythingmay stand for anything else, resemblance is not a necessary condition ofrepresentation either (even though, with the help of some imagination, thereis possibly some resemblance to be found; but that only proves thatresemblance is an extremely vague and highly flexible concept). Neither issubstitution a sufficient nor necessary condition of representation. Here,too, we must make a distinction between the use of that notion apart fromand in relation to the notion of representation. If I substitute a broken glasswith a new one, the new glass does not represent the old broken one, and if Irepresent a pyramid by means of a triangle, the triangle does not necessarilytake the place of the pyramid and function as one; it merely stands for it.Now let us return to Ankersmit.

Ankersmit’s fifth argument to favour the substitution theory ofrepresentation to the resemblance theory runs thus:

What we should note is that the substitution theory differs from the resemblancetheory by avoiding the suggestion of a conceptual hierarchy between thenotions of representation and of identity. The resemblance theory inadvertentlycreates such a hierarchy by presenting resemblance as the link between identityand representation: if A sufficiently resembles B, what A represents is (or isidentical with) B. Put differently, when A represents B it can be tailored by thecriteria of resemblance in such a way that the result of all the tailoring will be(identical with) the represented B all over again. And this is different with thesubstitution theory: this theory does not posit such a conceptual hierarchybetween identity and representation. (Ankersmit 2001, 227)

Here Ankersmit seems momentarily to have forgotten that resemblespresupposes difference: only two different things can resemble one another.Since resemblance precludes identity, the resemblance theory does not createa link between the notions of identity and representation, so it cannotsuggest a conceptual hierarchy between these notions in that it makes thenotion of identity secondary to that of representation. Hence, we still haveno reason to prefer the substitution theory to the resemblance theory. Theargument is false for another reason. The resemblance theory states that if Arepresents (or imitates) B, then A resembles B (in at least some respect).Therefore, we cannot argue on behalf of the resemblance theory, asAnkersmit does, that ‘if A sufficiently resembles B, what A represents is (or isidentical with) B’. That is not what the resemblance theory says. As notedabove, resemblance is neither a sufficient nor necessary condition ofrepresentation. A portrait will probably resemble its sitter, but the portraitdoes not represent the sitter because it resembles the sitter.

362 C. van den Akker

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

Maybe Ankersmit would respond by pointing out that I use ‘identity’ inthe sense of ‘being identical’ and not in the sense of ‘the identity of a person’,as, for instance, a person’s portrait represents the identity of the sitter, but inthis sense of identity too, the identity of the person as portrayed by theportrait only resembles the identity of the person after the representationalrelationship is established – or, after we can say such things as ‘the identityof the person as portrayed’. There is no suggestion of a conceptual hierarchybetween the notions of representation and identity in this case either.

The passage following the one just quoted is relevant because it explainswhy the substitution theory does not suggest a conceptual hierarchy betweenrepresentation and identity (a hierarchy which, as I have explained, is absentin the resemblance theory as well):

Representation is the representation of identity because identity only comesinto being by representation, and vice versa; there is not, first an identity,which is or could, next, be represented – whether in agreement with certaincriteria of resemblance or not – no, representation and identity both come intobeing at one and the same time. (Ankersmit 2001, 227)

Since identity and representation both come into being at one and the sametime, there cannot be a conceptual hierarchy between these two terms. Inthis passage the notion of identity is used in the second sense as the identityof a person. A statue should represent the identity of the person who isrepresented (Ankersmit 2001, 232). Do representation and the identity of aperson or thing come into being at one and the same time? Obviously, theexistence of Aurelius is a necessary condition for his statue, but that is notthe issue here. Ankersmit points at the fact that if we want to know theidentity of Aurelius, we have no other choice but to represent him. Ofcourse, Aurelius’s public identity is, as is common practice for sovereigns,fabricated. That does not alter the fact that representation and identity comeinto being at one and the same time. The public identity of Aurelius and hisprivate identity no doubt diverge, as will his historical identity; in any case,identity and representation come into being at one and the same time. Evenin antiquity, to be is to be seen.

Although we can agree with Ankersmit’s (and Danto’s) insight thatrepresentation and identity both come into being at one and the same time,we still have not found any reason to prefer the substitution theory to theresemblance theory of representation or vice versa. It is time to have a lookat our third theory, the exemplification theory of representation.

3. Representation and exemplification

The sculptor of the equestrian statue of Aurelius tells us how Aureliusshould be seen. We know from medieval descriptions that originally aconquered barbarian king kneeled underneath the horse’s front right leg.

Rethinking History 363

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

Aurelius had just returned from one of his successful conquests. With hisgesture, he may be greeting the Roman people or show clemency to theconquered king. Striking is his somewhat distant but patient facialexpression. The statue expresses his grandeur, his dignity as a stoic bringerof peace. That is what the statue is a metaphor for and how Aurelius shouldbe seen. ‘What it expresses is metaphorically exemplified’ (Goodman 1976,85). Or, ‘what a work expresses is what it is a metaphor for’ (Danto 1981,194). Does the statue of Aurelius look like the real Aurelius? It probablydoes. Is the statue a substitute for the real Aurelius? Yes, in the sense thatour attitude towards the real Aurelius should be the same as our attitudetowards his statue, an attitude of awe and submissiveness. The statuetransfigures Aurelius, as Danto would put it. The real Aurelius becomes anexample of a dignified, stoic bringer of peace by means of his statue.

What we should note is that none of the theories of representation iscompletely helpless when it comes to illuminating the equestrian statue ofAurelius. Resemblance helps us to determine what the statue represents: avictorious Roman on a horse. The victorious Roman and his horse arerepresented in accordance with well-known standards of representation.7

The substitution theory helps us to explain the raison d’etre of the statue: thestatue functions as a substitute for the real Aurelius. However, it is theexemplification theory that is most illuminating, for if we would not knowwhat it is the statue expresses, we would not know why it functions as asubstitute for Aurelius in the first place.

The exemplification theory adds something important to the analysis ofrepresentations such as statues. Resemblance and substitution illuminate,more or less successfully, the relationship between the representation andthe represented. Artistic representations not only stand for what theyrepresent (as all representations do); they express something about it too,thereby transfiguring the represented. This is what only the exemplificationtheory can explain. Therefore, the exemplification theory should bepreferred to both the resemblance and the substitution theory ofrepresentation in the case of works of art, and, as I will now argue, in thecase of works of history too.

Not only do works of art metaphorically exemplify, works of history doso too. For instance, Jacob Burckhardt expresses the thesis that in fifteenth-century Italy, man became conscious of his individuality. Man becomingconscious of his individuality is not something which can be found infifteenth-century Italy, nor can it be found in its remains. Burckhardt canonly show his thesis by means of examples. Leon Battista Alberti, Leonardoda Vinci, paintings, sculptures, ideas, architecture, even mockery and thirstfor glory; all are examples of man’s individuality. Although each aspect ofthe past can be identified and found as such, only relative to a historicalthesis can they be considered examples (of that thesis). By means ofBurckhardt’s work, Da Vinci, mockery, and so forth, become examples of

364 C. van den Akker

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

man’s individuality, in the same way that by means of the statue, Aureliusbecomes an example of a dignified, stoic bringer of peace. In this sense,works, artworks (Danto 2000, 132) and historical writings, embody themeaning they express. ‘What is expressed is metaphorically exemplified’,Goodman has taught us, or, as Danto has put it: ‘what a work expresses iswhat it is a metaphor for’. Burckhardt’s historical thesis is called theRenaissance. Fifteenth-century Italy as a Renaissance is a metaphor for manbecoming conscious of his individuality. Paintings, sculptures, architecture,inventions, human beings, and such things as mockery and thirst for glory,are all examples of this thesis. Burckhardt’s work has transfigured the past.

In his work, Burckhardt describes the fifteenth century. His workresembles fifteenth-century Italy in the sense that it is in agreement with ourstandard of realistic representation. Does his work function as a substitutefor fifteenth-century Italy? I do not think that it does. A work of historydoes not function as a substitute for the past in the same way that a stickfunctions as a substitute for a horse, for what would be the counterpart of‘riding’ in the case of Burckhardt’s work functioning as a substitute forfifteenth-century Italy? Maybe the work functions as a substitute in the sameway that Aurelius’s statue functions as a substitute. Our attitude towardsfifteenth-century Italy has changed as a result of Burckhardt’s work. We seefifteenth-century Italy in a new light, just as we see Aurelius in a new lightafter seeing his statue. This new way of seeing fifteenth-century Italy mightbe taking the place of an old way of seeing, but it does not take the place ofthe past by functioning as such. The statue takes the place of therepresented, while the work of Burckhardt does not.

Aurelius’s statue can be said to function as a substitute for the realAurelius, while this cannot be said of works of history. I believe that thesubstitution theory only makes sense if the representation represents anabsent body, whether that is a man, animal, or (man or animal-like) god. Iwill return to this in the conclusion. There is still one last argument ofAnkersmit’s to consider.

4. Magritte and the human condition

In his sixth and last argument, Ankersmit submits the resemblance andsubstitution theory to a test. The test is to explain Rene Magritte’s well-known painting La condition humaine.8 This painting from 1933 pictures aroom with a view. In front of the window in the painting stands anotherpainting. This painting in the painting represents a landscape that wouldhave been seen, were it not that the landscape is hidden from view by thatpainting. So here we have a representation in a representation. The paintingrepresenting a landscape is part of a painting of a room with a hidden view.

The painting representing the landscape can be dealt with from the pointof view of the resemblance theory as well as from the point of view of the

Rethinking History 365

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

substitution theory. According to Ankersmit, there can be no differencebetween the painting in front of the window and the view from the window.The painting in front of the window is a substitute for the landscape and itresembles the landscape (this is problematic to the extent that the view andits representation are identical, while resemblance presupposes non-identity;I will come back to this later). Ankersmit writes:

It is true that the relevant part of Magritte’s painting is identical with itself (aspart of the painting), but this part of the painting has two functions to fulfil: (1)to function as the landscape itself (in the way that paintings, according to thesubstitution theory are required to function as what is represented by them),and (2) to function as the painting of the landscape (i.e., as a representation ofthe landscape) . . . Now, needless to say, the difference between these twofunctions is exactly analogous to that between (represented) reality and itsrepresentation. But the difference of functions disappears if we see the paintingwith the eyes of the adherent of the resemblance theory: for there is nodifference in the resemblance relationship between (1) the landscape itself andthat part of the painting for which the painting is the substitute and (2) thelandscape itself and its representation as presented by the painting in the room.(Ankersmit 2001, 229)

In short, the painting in the painting fulfils two functions: (1) it functions asthe landscape itself and is as such a substitute for the landscape; and (2) itfunctions as a representation of the landscape. The distinction between thesetwo functions is absent from the point of view of the resemblance theory, sothe resemblance theory fails the test and the substitution theory does not. Iwill now argue why this is not so.

Remember our stick or hobbyhorse. The stick functions as a horse justas the painting in the painting functions as a landscape. The stick is nohorse, just as the painting in the painting is no landscape. The stickrepresents the horse and because it represents the horse it is a substitute forthe horse. This is, if we would follow Ankersmit, also true of the painting inthe painting. Because the painting in the painting represents the landscape, itis a substitute for the landscape. Saying that the painting in the paintingfunctions as the landscape itself – the first function (1) – is identical withsaying that the painting in the painting represents the landscape. Is this notwhat the second function (2) says? According to the second function (2), thepainting in the painting represents the landscape. Therefore, Ankersmit’sdistinction between the two functions is highly questionable. The painting inthe painting is a representation of the landscape and functions as asubstitute for the landscape. The fact that the painting in the painting hidesthe landscape from view is not important. The painting need not have beensituated in front of the window, but, for instance, underneath it. If that werethe case, the painting would still represent the landscape and function as asubstitute for it, even though we could see the landscape through thewindow. The fact that the painting in the painting hides the landscape from

366 C. van den Akker

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

view does not imply that it is a substitute for the landscape in the sense ofthe substitution theory: replacing one thing for another is substitution, butthat does not give us representation. Substitution is neither a sufficient nornecessary condition of representation.

Maybe we have missed the point. The landscape as painted by thepainting in the painting is identical with the landscape that the painting inthe painting hides from view. This is an extraordinary case. We never saythat a picture is identical with what it represents: we would be confusing theone for the other and that is not what we normally do. An idol taking theplace of a god is not identical with the god; it only has an identical function.A stick is not identical with a horse, even though it can have an identicalfunction. The fact that the landscape hidden from view is identical with thepainting of the landscape is only possible in special cases like Magritte’spainting, but then the issue is not representation; the issue is whether or howa representation can hide our view. We can even question whether thepainting in the painting represents the landscape. Since the landscape hiddenfrom view is identical with the landscape as represented by the painting inthe painting, there is no way we can distinguish between the representationand the represented. This is what we always should be able to do, accordingto the substitution and resemblance theory of representation.

To understand Magritte’s painting, we should take the painting as awhole into account. The picture consists of a room with a view and that viewis obstructed by the painting in the painting representing a landscape that isidentical to the obstructed landscape. We can identify what we see wheneverwe look at a painting, just as we can identify what we see whenever we lookthrough a window. Is Magritte not making the point that looking at apainting is identical to looking through a window? Magritte himself tells usthe following:

the tree represented in the painting hid from view the tree situated behind it,outside the room. It existed for the spectator, as it were, simultaneously in hismind, as both inside the room in the painting, and outside in the reallandscape. Which is how we see the world: we see it as being outside ourselveseven though it only is a mental representation of it that we experience inourselves. (Quoted in Ankersmit 2001, 228)

I do not think that we see the world ‘as being outside ourselves’, we simplysee all kinds of objects and events happening in the world. Perceiving apainting is no different from perceiving the world surrounding it (that is, thewaves of light hitting the retina are no different; there might be a differenceof processes in our brain caused by the three-dimensional interpretation of atwo-dimensional picture, but that is not a philosophical issue, but an issue tobe dealt with by physics, neurology, and biology). The interesting question iswhat the painting expresses, what it metaphorically exemplifies. It is clear tome that Magritte wants to say something about figurative painting as such.

Rethinking History 367

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

On the one hand, art is one of the things in the world; we see a painting onan easel in front of a window. On the other hand, art is expressive andtherefore not merely one of the world’s objects. Is the painting of a roomwith a painting in front of a window not a metaphor for perception? Is it nottrue that the painting exemplifies that seeing a figurative painting is likeseeing through a window? Is not that part of the human condition? At least,that is how Magritte sees the world. Through the painting, figurativepaintings of views through a window become an example of the humancondition.

5. Conclusion

The best argument against the resemblance view is perhaps made byGombrich. He has pointed out that function precedes resemblance(Gombrich 1978, 3–5). Something needs to function as a representationbefore the question of resemblance arises. We have said nothing else:resemblance is no condition of representation. One could argue that thisfunctioning is taking the place, and therefore that, in the end, thesubstitution theory of representation is correct and should be preferred tothe resemblance theory of representation, but substitution is only plausiblein some cases. Most of the time it amounts to nothing more than standingfor, merely rephrasing the fact that there is a representation. One importantlesson we should learn from our wanderings in the land of representations isthat not all representations are alike. The substitution view is only plausiblein such cases as described by Gombrich:

The clay horse or servant, buried in the tomb of the mighty, takes the place ofthe living. The idol takes the place of the god. The question whether itrepresents the ‘external form’ of the particular divinity or, for that matter, of aclass of demons is quite inappropriate. The idol serves as the substitute of theGod in worship and ritual – it is a man-made god in precisely the sense that thehobby horse is a man-made horse; to question it further means courtdeception. (Gombrich 1978, 3)

We cannot generalize the substitution view to all representations. A subwaymap is no substitute for the subway and a label on a tube of paint is nosubstitute for the paint in the tube, although the subway map representssubway stations and routes and the label represents the paint in the tube.Only representations that stand for absent bodies function as substitutes forthem. The examples which Gombrich gives us illustrate this, as does thestatue of Aurelius.9

Historical representations are no substitutes for what they representeither. Historical representations, like works of art but unlike subway mapsand Gombrich’s substitutes, express their meaning. So Jean AugusteDominique Ingres’s painting of Napoleon as Charles the Great expresses

368 C. van den Akker

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

the grandeur and dignity of the emperor. Through the painting, Napoleonbecomes an example of an exalted emperor. Napoleon as Charles the Great isa metaphor for his grandeur and dignity. Similarly, Jonathan Israel’sportrayal of late seventeenth- and early eighteenth-century Europe as aRadical Enlightenment expresses the coming into being of modern man, freedfrom the boundaries of ecclesiastical authority and divine legitimacy ofaristocracy andmonarchy, secularizing all institutions and ideas and believingin a freedom to do so. Late seventeenth- and early eighteenth-century Europeas a Radical Enlightenment is a metaphor for the coming into being ofmodern man. Israel shows his thesis on the past with numerous examples:radicals such as Baruch Spinoza and Frederik van Leenhof, the new eruditejournals, ‘universal’ libraries, literary clubs, lexicons, tea- and coffee-houses,and the deistic travel novel. On Spinoza, his prime example, he writes:

the early publication of the work [the Tractatus] is both intriguing andhistorically significant, being a key factor in the early diffusion of Spinozism inEurope outside the Netherlands. It also illustrates the extraordinary adroitnessof Spinoza’s publisher, Jan Rieuwertsz the elder (c.1616–87), and the early roleof the radical philosophical underground as a mechanism of clandestine bookdistribution. (Israel 2002, 279)

The early publication is not only an example of the adroitness of Spinoza’spublisher, but, together with the clandestine book distribution and diffusionof Spinozism, an example of the Radical Enlightenment itself. We canidentify and describe the Tractatus, its date of publication and its diffusion;but only through Israel’s work are these objects and events transfigured intoexamples of the coming into being of modern man. That is what his workexpresses and what it is a metaphor for. Only the exemplification theory ofrepresentation can account for this special relationship between works ofhistory and the past they represent.

In the end, Ankersmit need not disagree with the exemplification theoryof historical representation. In his main work Narrative logic (1981),Ankersmit introduces the notion of ‘narrative substance’. Narrativesubstances can best be understood as works. Like works, narrativesubstances embody theses on the past (Ankersmit 1981, 2). Narrativesubstances ‘do not refer to identifiable ‘‘things’’ or their aspects in historicalreality. They have a purely expository function’ (Ankersmit 1981, 117). Or,as I have put it, works of history show the past by means of examples. Thetheses they express cannot be found in the past, nor can they be found in itsremains. One can only give examples of them. Ankersmit puts it thus: factsin historiography have an ‘illustrative’ use (Ankersmit 1981, 147). Thequoted passage of Israel above nicely proves this. Finally, like metaphor,narrative substances individuate a point of view, ‘a perspective in terms ofwhich reality should be analyzed or viewed upon’ (Ankersmit 1981, 228).What a work expresses is metaphorically exemplified and what it is a

Rethinking History 369

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

metaphor for. Radical Enlightenment is a metaphor for late seventeenth-and early eighteenth-century western Europe expressing the coming intobeing of modern man. That is how the late seventeenth and early eighteenthcenturies should be analysed or viewed. Aspects of that part of the past –Spinoza, literary clubs, and so on – should be seen as examples of thecoming into being of modern man. This, then, is the narrative logic ofhistoriography that the exemplification theory of historical representationcaptures.

Acknowledgement

I would like to thank two anonymous referees for their helpful and stimulatingcomments on an earlier version of this essay.

Notes

1. Ankersmit is not consistent in his use of resemblance. He writes: ‘representa-tions, such as paintings, will ordinarily ‘resemble’ each other (in the sense of allbeing pieces of canvas with paint spots on them) more than what they‘represents’ (such as castles, human individuals, etc.)’ (Ankersmit 2001, 310, n8).This argument was first made by Goodman (1976, 5).

2. Although non-figurative art may not represent actual objects, it can still beconsidered representational in the sense that it is about something. Danto betsthat of every work of art, he can tell what it is about (Carroll 2002, 19). Forinstance, if an abstract painting exemplifies linearity, it represents and is aboutlinearity, without representing an actual object, person, or such thing as alandscape. Here, we already anticipate our third theory of representation, theexemplification theory.

3. This statue is a weatherproof copy of the original, which can be found at thecourtyard of the Palazzo Nuovo.

4. It is misleading to compare a table with a hobbyhorse, since a hobbyhorse is arepresentation and a table is not. It would have been better to compare a stickand a table. See also Gombrich (1978, 7): ‘It needed two conditions, then, to turna stick into our hobby horse: first, that its form made it just possible to ride on it;secondly – and perhaps decisively – that riding mattered.’

5. The reason why a stick functions as a horse whereas a table does not is a matterof habit. One can imagine a society in which people do not eat at a table or sitdown at one to converse. One eats and converses while standing, holding a stickto represent the fact that they do so. Tables are just there to put the dirty disheson. In this imaginary society, parents would give their children the tables to playwith, not the important sticks.

6. We may add that talk about tables resembling horses is using resemblance in thesense apart from the concept of representation. This is not what the resemblancetheory is about. The resemblance theory states that we understand arepresentation because it resembles what it represents

7. Here, we should keep the distinction between the use of the concept resemblanceapart from and in relation to the concept of representation in mind. Determiningwhat a statue represents by means of resemblance is using the concept ofresemblance in relation to the concept of representation: A represents B becauseA resembles what our standard of representation considers to be B. The sculptorprobably tried to imitate or resemble a real horse and the features of Aurelius.

370 C. van den Akker

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13

Here, resemblance is used apart from the concept of representation. Resemblingand imitating is not necessarily representing, although it can be used in service ofit, following the habits and practices of the day.

8. As far as I know, at least four paintings of Magritte bear the title La conditionhumaine. They are painted in 1933, 1935, 1944 or 1945 and 1948. Also L’appeldes cimes (1933), La belle captive (the title of two paintings, one from 1950 andone from 1967) and Les promenades d’Euclide (1955), depict a painting in apainting representing a landscape.

9. More historical examples of this type of representation can be found in CarloGinzburg (2002).

Notes on contributor

Chiel M. van den Akker is a postdoctoral researcher at the VU UniversityAmsterdam and teaches historical theory at that university. He has written in Dutchabout narrative, truth and interpretation, and historical representation.

References

Ankersmit, F.R. 1981. Narrative logic. A semantic analysis of the historian’s language.Meppel: Krips Repro.

Ankersmit, F.R. 2001. Historical representation. Stanford, CA: Stanford UniversityPress.

Burckhardt, J. n.d. Die Kultur der Renaissance in Italien. Wenen: Phaidon-Verlag.Carroll, N. 2002. Introduction. In Theories of art today, ed. N. Carroll, 3–24.

Madison: University of Wisconsin Press.Danto, A.D. 1981. The transfiguration of the common place. A philosophy of art.

Cambridge: Harvard University Press.Danto, A.D. 2000. Art and meaning. In Theories of art today, ed. N. Carroll, 130–40.

Madison: University of Wisconsin Press.Ginzburg, C. 2002. Representation: The word, the idea, the thing. In Wooden eyes.

nine reflections on distance, 63–78. London and New York: Verso.Gombrich, E. 1978. Meditations on a hobby horse and other essays on the theory of

art. London and New York: Phaidon Press.Goodman, N. 1976. Languages of art. An approach to a theory of symbols.

Cambridge: Hackett.Israel, J.I. 2002. Radical enlightenment. Philosophy and the making of modernity

1650–1750. Oxford: Oxford University Press.Munslow, A. 2003. The new history. Harlow: Pearson Education.Zammito, J. 2005. Ankersmit and historical representation. History and Theory 44:

155–81.Zelenak, E. 2009. Exploring holism in Frank Ankersmit’s historical representation.

Rethinking History 13: 357–69.

Rethinking History 371

Dow

nloa

ded

by [

Ford

ham

Uni

vers

ity]

at 1

5:51

14

Sept

embe

r 20

13


Recommended