Literary composition and the early medieval historian in the nineteenth century

Ian Wood

Print version

Ranke may not seem the obvious way to begin a discussion of the relationship between history and nineteenth-century historical novels which take the early Middle Ages as their subject. After all, in the preface to his History of the Latin and Teutonic Nations from 1494 to 1514, first published in German in 1824, he commented that 'A strict representation of facts, be it ever so narrow and unpoetical, is, beyond doubt, the first law.' Moreover, the German scholar had scarcely any impact on the writing of early Medieval History, despite his importance for early modern and modern historians. Although he is frequently described as the mentor of Georg Waitz, the latter was already working for the MGH before he attended Ranke's seminar: and although Ranke appears in Waitz's autobiography, it is as a friend and a hill-walker, not as a teacher. In fact scholars working on Antiquity and the Middle Ages were already paying attention to source criticism long before Ranke came along.

Alongside his reputation for being a stickler for facts, Ranke was also concerned of course which how things really were: wie es eigentlich gewesen. The phrase appears in the preface to the History of Latin and Teutonic Nations. One insight into what Ranke meant may perhaps he gleaned from Philip Ashworth, the translator of the History, who visited the old German scholar shortly before he died. In the course of the interview Ranke remarked: 'Great as is the respect and veneration in which I hold Sir Walter Scott, I cannot help regretting he was not more available for the purposes of a historian than he is. If fiction must be built upon facts, facts should never be contorted to meet the ends of the novelist. What valuable lessons were not to be drawn from facts to which the great English novelist had the key; yet, by reason of the fault to which I have referred, I have been unable to illustrate many of my assertions by reference to him.'

The historical novel, then, if fully footnoted, so that one could identify what was accurately recorded, could be used by the historian as source material. Manzoni, the Risorgimento novelist, dramatist and cultural theorist, in his treatise Del romano storico, made a series of observations about the historical novel, as well as epic and tragedy, which connect at a number of points with Ranke's yearning to use Scott, whom the Italian called 'the Homer of the historic novel'. 'How many times', he asked, 'has it been said, and even written, that the novels of Walter Scott were truer than history!' Surely, wie es eigentlich gewesen. The historical novel, as a mixture of history and invention ought, according to Manzoni, to be an impossibility (we are not far here from Virginia's Woolf's description of biography as a bastard art), but could be a success in the hands of a master. More important, from our point of view, he also pointed out that historical novels became increasingly based on fact, and indeed started to include footnotes, from the eighteenth century onwards. Manzoni made a comparison with drama: 'Shortly after the middle of the last century, a French actor or actress (I do not know which) introduced a general reform in costuming to make it conform to the time in which the dramatic action was set.' In relation to all this he used the concept of the verisimilar (verosimile).

In what follows, I shall not be looking for accurate facts in historical novels written in the nineteenth century. Rather, I wish to consider a number of literary works written by men who were also scholars, to see whether there could be a serious interpretative purpose in the historical novel. Essentially I shall look at three novels: Chateaubriand's Les Martyrs, Sismondi's Julia Sévéra, ou, l'an quatre cent quatre-vingt-douze, and Dahn's Kampf um Rom, although I will also look briefly at two dramatic works: Kaiser's Attila, König der Hunnen and Manzoni's Adelchi. The shadow of Sir Walter Scott will, of course, be ever present, but as none of his novels deals with the end of Rome or the post-Roman period, he will remain a shadow.

Scott, although the greatest historical novelist of the period was by no means the first. One might categorise some English Gothic novels as being attempts to conjure up the past. More scholarly, and perhaps more important, was Marmontel's Bélisaire, published in 1767. This tells the story of Belisarius, the general of Justinian, who fell from grace and was, in some traditions, imprisoned and blinded. The novel recounts the blinded hero's return home. As he begs his way back to his castle he meets a number of his earlier opponents, including the Vandal king Gelimer and also the Bulgars, who are overcome with pity for him, and take vengeance by killing one of the Romans who had betrayed him in Italy. The news of his progress reaches Justinian, who decides to follow him incognito, with the young general Tiberius. On reaching his castle, Belisarius's wife dies in a moment of joy and grief, on seeing him blind. One critic unkindly suggested that she was very sensible to die at that moment, because it meant that she did not have to listen to the political discourses between Belisarius and Justinian that take up the rest of the book. Marmontel disarmingly states in the preface that he knows that it was only a popular tradition that Belisarius was blinded (actually it is a remarkably early tradition, as shown in the early nineteenth century by Lord Mahon, though probably still a fiction): but he claims that everything else in the novel is derived from historical fact, essentially taken from Procopius and Agathias, though not, he insists, the Secret History, which he did not regard as an authentic work of Procopius.

Marmontel's claim to historical accuracy is certainly not justified, although he clearly knew his sources very well, and makes frequent allusion to them. In fact, however, the novel was inspired by a famous image of the blind Belisarius, originally by Van Dyck, but known to Marmontel from an engraving by Abraham Bosse, and the idea of writing it came from Diderot. Its purpose was not to say anything about the past, but rather to debate the Bourbon present. Marmontel himself had been imprisoned in the Bastille on the false charge of having slandered the duc d'Aumont. To read Marmontel in the hope of discovering historical verisimilitude, to use the notion stressed by Manzoni, would be a disappointing experience, despite the claims of the preface.

A much more ambitious, and indeed bizarre, attempt to conjure up the past was Chateaubriand's Les martyrs, begun in 1802, just as he was completing the Génie du Christianisme. Les martyrs is set at the end of the Great Persecution: it tells the story of Eudore, a young Christian, and Cymodocée, daughter of a pagan priest, who, as a descendent of Homer was responsible for the upkeep of the shrine of the great poet. Most of the first half of the novel is taken up with the exploits of Eudore. He had been a soldier, whose postings had taken him through much of the Empire, including a spell in Rome, where he had been educated with Ambrose and Augusine, and a brief period as slave to Clothilde, the wife of Pharamond, before he became commander in Armorica, where the druidess Velleda fell for him – the episode would be the basis for the libretto of Bellini's Norma. In the course of his wanderings he ended up excommunicated, and as a result was subject to a long period of penance. It was as a penitent that Cymodocée saw him and fell in love with him. She, however, was lusted after by the governor Hiéroclès. Eudore saved her from the governor, and sent her off to Jerusalem, where she entered service with Helena. He himself went to Rome where he represented the Christians before Diocletian, who is presented as wise but weak, and as a result was persuaded to sign the edict of persecution and then to abdicate by Galerius and Hiéroclès. Eudore warned Constantine to flee to Britain, and was then tried, condemned, and imprisoned with just about every other martyr who ever existed. The outbreak of persecution meant that Helena could not protect Cymodocée, who headed into the Judaean desert, and was baptised in the Jordan by Jerome. She then made her way to Rome to join Eudore in the Colosseum, where he managed to place a ring on her finger before they were eaten by wild beasts.

It is difficult to do justice to the narrative of Les martyrs. Not least because alongside this worldy story there are scenes in Heaven and Hell, and even a brief trip to Purgatory. The earthly scenes are heavily footnoted: they are underpinned with a barrage of citations of classical and early Christian sources, as well as references to objects and sites that Chateaubriand knew or had read about. In the preface Chateaubriand insists that there is a factual basis for what he has to say. He does, however, admit, disarmingly, that there is a certain amount of apparent anachronism (which is a slight understatement, seeing that the material he draws on stretches from the Archaic Age to the fifth century A.D.), but he urges the reader not to be disturbed by any apparent discrepancy, arguing that Augustine and Jerome behave as they would have done had they been there, and that if one thinks that a figure like Pharamond, Clodion or Mérovée, is out of place, one should just imagine that it was another person of the same name as the individual known from Fredegar: there were, after all, lots of Pharamonds, Clodions and Mérovées in history (Chateaubriand's view, not mine!). For the modern reader the historical anachronisms may seem easier to stomach than the scenes in heaven and hell, which are unfootnoted, but which are partly dependent on Paradise Lost, as Chateaubriand states in the preface. Indeed the whole work is conceived as a confrontation between Homer and Milton. Heaven and hell are every bit as real as the Roman Empire in the novel: effectively the world conjured up by Chateaubriand is that of religious art, which in precisely the same decade was being revived with, for instance, plans for the decoration of the dome of the Panthéon, which was to have represented Napoleon and Josephine in heaven with Charlemagne, St Louis and the shrine of Geneviève, Clovis and Clothilde: the scheme, of Antoine Gros, had to be modified to represent the restored Bourbons in place of the emperor and empress. There were those whose religious understanding meant that their reading of history was cosmic. And if Les martyrs for us lacks historical truth, Manzoni's verisimilitude, for Chateaubriand it was concerned with eternal truth: the work represented an attempt to put the arguments of the Génie du Christianisme into a novel, so as to reach a wider audience. It is a text that lies at the heart of the catholic revival of the nineteenth century, and would inspire a long line of catholic historians, most of whom have been forgotten, but who were of significance for the development of Church, and more broadly religious history: Frédéric Ozanam, founder of the Société de Saint Vincent de Paul, and therefore himself a saint, as well as author of a major work on the early medieval Church, Montalembert, author of Les moines d'Occident, and the Prince de Broglie, president of France, and author of a very creditable account of the fourth century Church, which, although it was attacked by the abbé Guéranger, founder of Solesmes, was defended than none other than Pius IX.

But while Chateaubriand had his devotées (and we shall return to yet another in a minute), he also had his opponents, of whom the most interesting was perhaps Jean Charles Léonard Simonde, better known by his adopted name of Sismondi. He set about writing his historical novel, Julia Sévéra, ou quatre cent quatre-vingt-douze in 1819/20, just after he had begun writing his Histoire des français, twenty-nine volumes of which would appear between 1821 and 1842. The first volumes of his Républiques italiennes had appeared in 1807/8, though the work was not completed until 1817.

Stripped to its bare bones Sismondi's narrative is not unlike Chateaubriand's. Boy meets girl, they fall for one another, but they are only united after a series of picaresque adventures. In this instance the boy is Félix Florentius and the girl Julia Sévéra: her other suitor is none other than Clovis. She is pushed forward by the catholic clergy in order to ensure his conversion. The objection of the lovers leads them to be kidnapped by the agents of bishop Volusianus of Tours in a wonderful scene set in a ruined castle, which has more than an echo of the English Gothic novel. Félix's objection leads to him to be exposed in church as a demoniac: Volusianus exploits the liturgy to whip up public opinion in a scene which deserves full operatic performance. Félix's subsequent attempts to get justice lead to his falling in with the Bagaudae (bands of dispossessed and dissidents), before he and Julia are finally united.

This being Sismondi, rather than Chateaubriand, religion comes off badly: as a Genevan Calvinist Sismondi had no love for the Catholic episcopate, which he saw as the enemy of all liberty. Yet equally, and more important for us, being Sismondi a very considerable amount of detail is devoted to the socio-economic reality of the period. We are told a good deal about slavery, and the depiction of the Bagaudae is as well founded as any scholarly treatment of the subject: essentially we are given a close reading of Salvian by a man for whom the inadequacies of Adam Smith had been exposed by a tour of the industrial cities of the North of England.

In his preface Sismondi is explicit about his aims. He wished to present the conditions of peoples, the relations of inhabitants, dominant opinions and domestic habits. He seems to have intended to write a series of novels setting out such aspects of the past for a whole sequence of periods, but in the event he only produced the one for the late fifth century, where he brought together Romans, Armoricans, Bagaudae, Franks and others. He explains that he wished to set out an image of society as it really was, and not to present a particular religious or political system. The comment on a religious system is surely a swipe at Chateaubriand. Not that he really avoided setting out his own anticlerical and libertarian religious and political views. Even so, we can, I think accept that the novel really was an attempt to present his interpretation of late Roman and early Frankish society to a broad audience.

As for that audience, we might guess that it included his mother, but also his wife and her friends. He had married Jessie Allen, sister-in-law of Josiah Wedgwood, and aunt of Charles Darwin, in 1819. It would seem that he also wrote his History of the Fall of the Roman Empire, a very smart modification of Gibbon, for Jessie and her friends, for it appeared in English in 1834 months before it appeared in French. That an English audience was envisaged for Julia Sévéra might be deduced from the use of epigraphs from contemporary sources, in the manner, as Sismondi put it, of the Scottish novelists (he clearly had Scott in mind). But in any case Julia Sévéra appeared in both French and English, though the French edition came first. The English text actually appeared at more or less the same time as the publication of the first American review, which announced that this was a book that was so bad that it most certainly did not deserve translation. This is a little unkind: it may lack the verbal glories of Chateaubriand: indeed much of it is in rather plain prose (though the scenes where Sismondi’s own feelings about liberty and catholic oppression break out are rhetorical tours de force. But it is a serious attempt to present the early Middle Ages as Sismondi understood them (which was pretty well) to the general public. And as he comments in his preface, most of the characters are attested in the sources: the events are rarely in conflict with what was known of the period in the first half of the nineteenth century, and there is only one unquestionable anachronism (which is admitted by Sismondi): the hermit, Senoch, in whose cell Félix and Julia are united, lived fifty years after the supposed date of the story.

At almost exactly the moment that Sismondi turned to write up the fifth century in novel form Augustin Thierry embarked on a rather more radical combination of history and fiction. As Thierry himself explained, it was reading Les martyrs that inspired him to write about the past, and it was Ivanhoe that provided him with a model of how to write. The two influences came together most fully in his Récits mérovingiens. The first of these appeared as journal articles in 1829, but they were gathered together in 1840. These works of Thierry can perhaps best be described as 'faction'. They are painstakingly accurate reworkings of sections of Gregory of Tours, to which very little has been added except description and direct speech. Essentially Thierry presents his material as if it were a novel, but the accuracy of the reproduction of his source material is exemplary. And, it is underpinned with a vast apparatus criticus, quoting large sections of the source material in Latin, and adding discussion of points of detail – all of which, in the later articles, was done from memory, for Thierry suffered terribly from disease which left him blind and crippled, and able only to quote to an amanuensis.

But while Thierry cannot be faulted in his presentation of his material, his interpretation essentially lay in his choice of episodes, which consistently brought out the worst in the Franks, and of the Merovingians in particular. Thierry saw himself as both a Gaulois and a rotourier: in certain crucial respects the two categories overlapped. To be a Gaulois in France was not to be a Frank: and for generations, but above all since the publications of Henri de Boulainvilliers in the 1720s, the Franks had been associated with the aristocracy, while the Gaulois had been thought of as essentially the peasant class of France. In the eighteenth century aristocratic writers had been keen to present themselves as descendents of the Franks. During the Revolution this association was inevitably rather less popular, and there was a growing desire to be thought of as a descendent of the Gaulois. It was this feeling that Thierry encapsulated. Essentially he reversed the old picture of the arrival of the Frankish champions of aristocratic equality, and instead stressed the extent that they were oppressors. He could have found the material to effect that reversal of ideas in the writings of Boulainvilliers, Montesquieu and Mably, but it was, in fact, English historiography of the Norman Conquest and Scott's depiction of the oppression of the Anglo-Saxons in Ivanhoe that led him to present the Merovingians in unremittingly dark terms.

It was a vision of the Frankish past that had a remarkable impact on the realisation and representation of the early Middle Ages. The Récits merovingiens would be illustrated above all by Jean-Paul Laurens in 1880, but the notion of the degenerate Merovingians became a popular subject for painters from the middle of nineteenth century. Laurens was one of the artists commissioned to decorate the Panthéon, which has a number of medieval cycles: depicting Genèvieve, Clovis, Charlemagne and Jeanne d'Arc. The Huns, the Franks, and the Alamanns are all presented as out-and-out savages (although among the clergy surrounding Clovis in the depiction of his baptism by Paul-Joseph Blanc one can make out Antonin Proust, Léon Gambetta and Georges Clémenceau!). Not that all the images in the Panthéon present the period as savage. By the 1870s when the designs were being drawn up, an alternative reading of the early Middle Ages was becoming established: that of Fustel de Coulanges, with its emphasis on Roman continuity, which found its presentation in the paintings executed by Puvis de Chavannes.

Important in all these, however, Laurens as much as Puvis, is the painstaking attempt to be accurate. The paintings present the latest in archaeological reconstruction, and they portray fibulae, and other jewelry, as well as weapons, as they had been revealed by excavation. These paintings are in many respects the equivalents of Thierry with his extensive citation of sources.

Thierry's writings were enormously popular, and not just in France. Much of his work (which included a history of the Norman Conquest of England) was translated into English, and the Historical Essays (including the Récits mérovingiens), which were first collected together in French in 1840, appeared five years later in an English translation printed in Philadelphia.

Although the first of the articles which make up the Récits mérovingiens was not published until 1829, Thierry's model of oppression had already been set out in earlier articles gathered together in 1827 as the Lettres sur l'histoire de France. The first of these had started to appear in the Censeur Européen in 1817. Two years later Alessandro Manzoni, who was in Paris at the time, came across Thierry's work. He saw in the story of oppression that Thierry had derived from Scott and behind him the Anglo-Saxon scholar Sharon Turner, a model that he thought he could apply to Italian history. He did so initially not in a historical novel (the Promessi Sposi would come later, in 1828), but in a pair of dramas, Il conte di Carmagnola and the Adelchi. This latter was concerned with the early Middle Ages. It tells the story of Charlemagne's repudiation of his Italian wife, his invasion of Italy, and the defeat of the last Lombard king, Desiderius. In historical reality Desiderius's son Adelchis managed to escape to Constantinople, where he plotted, without success, to reverse the Lombard defeat. In Manzoni's presentation Adelchi is killed fighting the army of Charles, but only after first realising that the Lombard State was doomed because of its oppression of the indigenous Romans.

Unlike Thierry, Manzoni did alter history to suit his argument, though without the death of Adelchi there would have been no tragedy, and the end would as a result have been less dramatically satisfactory. As Manzoni argued in his essay on the historical novel, which is expansive enough to include both epic and also tragedy, the story had to be paramount. At the same time, as we have seen Manzoni noted the increasing determination of writers of historical fiction to research their subject matter thoroughly, and even to cite source material, either in the text itself or in footnotes. In the case of the Adelchi, Manzoni went one stage further, writing his Discorso sopra alcuni punti della storia langobardica in Italia. This is a detailed examination of the evidence for the Lombard settlement in Italy (which ultimately comes down to two sentences in Paul the Deacon), which interprets that material as showing that the settlement amounted to major oppression of the Italian people, who were effectively enslaved – the model is ultimately Thierry's for the treatment of the Gauls by the Franks. The Discorso looks at the whole history of Lombard dominance, but above all at the early years of their settlement: the Adelchi presents the conclusions of the Discorso for a literary audience, setting out the picture of oppression largely in two great choral odes, in the speeches of the priest Martin, who guides Charlemagne's troops, and in Adelchi's final descriptions of the state of Italy under Lombard rule.

Manzoni clearly had a political motive in setting out this interpretation, for Lombards come to represent all foreign invaders of Italy, the removal of whom was central to the Risorgimento. Not surprisingly the interpretation offered by the Discorso became the dominant reading of the Lombard period, albeit with some modifications, notably at the hands of Carlo Troya and Cesare Balbo.

Equally important for Manzoni (and indeed for Troya) was the idea that the papacy might play a role in the unification of Italy: and indeed papal influence is an important factor in prompting Charlemagne to invade the peninsula in the Adelchi. Hadrian himself makes no appearance, for his representation on stage was not permitted in Italy. He could, and did, appear in Germany, where Zacharias Werner gave him a central role in his drama Attila, König der Hunnen of 1809. Werner, who was for a while seen as the more likely literary successor to Goethe, particularly in the circle of Mme de Staël, wrote a number of historical plays, notably about the Baltic Crusades. All his works were extremely well researched: the Attila shows a remarkable knowledge of all the available sources. In Werner's hands the play becomes a proto-Wagnerian spiritual drama. At the time of writing Werner himself was still a protestant, although his spiritual leanings had already led to a breach with Goethe, and he would become a catholic, and indeed would be the dominant catholic preacher in Vienna at the time of the Congress, when he would denounce his earlier theatrical works. The Attila would attract the attention of a number of composers: in particular Verdi saw the opportunity to transform the work into another Risorgimento piece depicting Italy under threat of foreign invasion. Verdi would also add a scene in which hermits fleeing from the sack of Aquileia found Venice – a not-overly subtle allusion to the fact that the work was commissioned by La Fenice: though Verdi's librettists did base the scene on a passage in Sismondi's work on the Italian republics. And an attempt at verisimilitude was made in the original production, in that Verdi wanted authentic designs. What this amounted to was sending the designer to the Vatican to look at Raphael's famous fresco of the encounter of Attila and Leo – though in the opera Leo had to be designated simply as an old man.

Over the position of the pope and more generally of catholicism Manzoni and Sismondi were at loggerheads. Sismondi's study of the Italian Republics was as important in providing historical ammunition for the Risorgimento as was the Adelchi, but Sismondi had no time either for the papacy or for the Catholic Church, which prompted Manzoni to attack his stance over religion in his Osservazione sulla morale cattolica.

Manzoni's stance may have held sway in Italy, but it was challenged strongly by more northerly writers, perhaps most interestingly by Felix Dahn. Dahn's twelve-volume Könige der Germanen, which came out over the period from 1861 to 1911 is a very precise study, which is remarkable for its caution. Dahn knew his sources extraordinarily well. He wrote significant studies of Procopius (he was in the vanguard of those arguing for the authenticity of the Secret History) and Paul the Deacon, and his analyses of law in his study of kingship are impeccable. Unlike most German contemporaries, and indeed a number of English scholars, in his academic he does not overstate the Germanic nature of his information. Occasionally he pauses to say that there must be a Germanic aspect to what he is dealing with, but since he cannot find it in the sources he has to leave it out. His Germanic enthusiasms, and they were considerable, found a different outlet: in plays, poems, opera libretti and a novel. Of these, the most important is his novel Ein Kampf um Rom, which deals with the failure of the Ostrogothic kingdom in Italy.

Essentially Dahn provides a narrative account of events in the Italian peninsula from the death of Theodoric to the final defeat of the Ostrogoths at Busta Gallorum. The narrative largely follows Procopius and Agathias, although love and hate interests are added, while a shadowy senator, Cethegus, who does appear in Procopius (where he is indeed somewhat sinister) becomes the representative of the Italians rejecting both Ostrogothic and Byzantine rule. In addition, there are two rather more bizarre episodes, when Vikings turn up: on the second occasion to take the remaining Goths and their dead leaders to Thule, thus affirming their ultimate association with the Germanic North.

Dahn's novel is certainly a good yarn: good enough to have been made into a film in the 1950s, starring Orson Welles. It is not, however, taken seriously: certainly not as history. Commentators have noted its significance as a roman-à-clef, something that Dahn himself makes explicit in his Autobiography. As he explained it, Justinian was Napoleon III, Theodora was Eugénie, the Goths were the Austrians, and the scheming Cethegus was Pius IX. First dreamt up in Ravenna in 1858, begun in Munich in 1859, but shelved for much of the 1860s, and only completed in 1874, these parallels, however, became more and more stretched. After all, by 1874 Napoleon and Eugénie were no longer forces to be reckoned with, even if the Austrians had indeed been driven out of Italy.

Emphasis on the narrative and on its connections with political events in the 1860s have, I suspect, blinded readers to a more significant aspect of the work, which is not just clear from the bizarre role played by the Vikings, but is also closely tied to the representation of the leaders of the Goths as Germanic heroes. This is the Germanic reading of the early Middle Ages that Dahn felt he could not put into his more scholarly work: arguably this is what he thought was the more accurate representation of the early period: we are back with Ranke's wie es eigentlich gewesen, or with Manzoni's claim that Scott's novels were truer than history. For Dahn we may suspect that the picture he presented in Ein Kampf um Rom was historically more ‘true’ than the scholarly picture presented in Die Kőnige der Germanen.

Moreover, Dahn's attempt to represent Ostrogothic Italy was carefully executed. Like Sismondi, Dahn spent time to explain the position of the peasantry. The rights of the Ostrogoths over land are carefully established: their treatment of the subject population is the dealt with in more than one episode in the novel. What we see is not oppression, but rather a pattern of justified ownership which caused some, largely unjustified, resentment. It is not hard to see in Ein Kampf um Rom the mirror image of Manzoni's Adelchi. The Goths are not oppressors: Romans rather than Goths are the trouble makers.

Dahn does not claim to be answering Manzoni, and indeed there is no clear citation of the Adelchi. But given his theatrical interests, both as an author and indeed as the son of an actor and an actress, and given his linguistic facility and his love of Italy, it is hard to believe that he did not have the Adelchi in mind when constructing Ein Kampf um Rom.

The works of literature that I have been considering are rather more complex works than is usually acknowledged – with the exception of the Adelchi, whose scholarly credentials are openly proclaimed because of the Discorso. But, as we have seen, all these works, even Marmontel's Bélisaire, are based on a thorough knowledge of the sources – something that Manzoni thought was central to the historical novel, as it had developed by the end of the eighteenth century. The group of works is certainly a rather special one: I have concentrated after all on works of literature written by men who had some claim to be historians: Chateaubriand, Thierry, Manzoni, and Dahn. All of them had axes of one sort or another to grind: Chateaubriand was pushing for a religious revival, Thierry was arguing the case for the Gauls against the Franks, Manzoni, like Verdi, was promoting the Risorgimento, and Dahn was arguing for Germanic tradition. And to some extent these works were allowing their authors to go beyond the bounds of conventional history, in pushing their precise readings.

Equally important is the impact of these works on the historical imagination, not so much of scholars (though one would be wrong to assume that scholars are immune to such influences), but of the wider public. All these writers were addressing an audience beyond any academic readership. I am aware that to some extent the concept is not entirely valid for the nineteenth century. Le grand publique might have read Chateaubriand's Génie du christianisme: and a sizeable number of people may have read sections of Sismondi's histories. The publication history of Manzoni's Discorso does not suggest a best seller: nor indeed does that of Adelchi, although the choral odes were set to music, and must have been more widely heard. How many people other than academics ever read Dahn's Könige der Germanen? But we have a pretty good idea how of the scale of the readership of Ein Kampf um Rom. 84,000 copies were printed in the first 18 years of publication: it became even more popular in the run up to the First World War, and in July and August 1914 it was the most popular confirmation present for boys: more popular, that is, than the Bible. By 1938 it had sold 615.000 copies: by 1950, 750,000. These are colossal numbers for the time.

What this must have meant was that the German understanding of the Ostrogothic period was the image of Ein Kampf um Rom – and not that of the relevant volume of the Könige der Germanen. So too, the dominant image of the Merovingians in France was that of Thierry's Récits mérovingiens. It would appear, however, that Sismondi's Julia Sévéra never really took off. As for the history of the persecutions, that was surely dominated not by Gibbon, but by Chateaubriand. And alongside these literary representations there are the images presented to the public through etchings and paintings: Bosse's Bélisaire, which actually inspired Marmontel: Laurens' illustrations for the Récits mérovingiens: even the designs for Verdi's Attila, derived from the Raphael fresco. The Raphael frescoes would never have been a very good guide to what Attila and the Huns looked like: nor is Bosse's Bélisaire to be relied on. But there is rather more to be said for Laurens' illustrations for Thierry's essays. Ranke was clearly right to be suspicious of Walter Scott: nevertheless, these works did make some attempt at achieving verisimilitude. If one believed in the nineteenth century that one could come close to wie es eigentlich gewesen, Thierry and Laurens were not a bad place to start.