The Discovery of France: A Historical Geography, from the Revolution to the First World War
by Graham Robb
Reviewed by David A. Bell
The New Republic Online
For a book of "historical geography," The Discovery of France has received remarkable attention and acclaim: long and appreciative reviews in British and American newspapers, the title of "notable book of the year" from the New York Times, rapturous applause in the New York Review of Books, and so forth. The reason is not hard to see. Graham Robb is an engaging and gifted writer, known for his enjoyable and instructive biographies of Hugo and Rimbaud. Moreover, The Discovery of France is the sort of history that seems almost to have disappeared from the world of professional academic historians: written in a light and pleasant style, crammed with colorful and unexpected details, it offers what seem like tantalizing glimpses into a vanished, forgotten past.
All the more pity that it is actually a distressingly bad book. Robb has set out to uncover the true story of France's cultural diversity, and to show how cartographers, surveyors, travelers, and officials of the French state started to come to terms with this diversity between 1789 and 1914. But he falls headlong into some of the biggest traps that await overly enthusiastic historians. He presumes without justification that source material drawn helter- skelter from widely differing regions and periods belongs to a single story -- a melodramatic saga of a "lost world." He therefore elides crucial regional and chronological distinctions, while ignoring some rather large and important phenomena -- the Counter-Reformation, absolutism, proto-industrialization -- that fail to fit his thesis. And he accepts deeply tendentious accounts of French life, marred by ideology or romance, as if they were transparent descriptions of it. As a result, when he claims to be mapping historical terra incognita, he is often doing little more than recycling nineteenth-century myths. And he fails to put any of what he is doing in a larger historical or geographical context, again and again greeting with delighted surprise what are in fact well- known elements of European social and cultural history.
What makes matters worse is that these mistakes were so unnecessary. Over the past century, some of the greatest names in European and American scholarship -- Fernand Braudel, Emmanuel Le Roy Ladurie, Natalie Zemon Davis, Laurence Wylie, Alain Corbin, Eugen Weber, and many others -- have produced shelves' worth of books on virtually every subject that Robb touches on. Robb cites some of this work in a scattershot way, but he has learned little from it, and he acknowledges hardly any of it. Indeed, he goes so far as to write in his introduction that "For some time, it had been obvious that the France whose literature and history I taught and studied was just a fraction of the vast land I had seen" -- a statement that, if true, means that he either failed to read or failed to understand most of the secondary works in his own bibliography.
He wanted to write a book, Robb says, "in which the inhabitants were not airlifted from the land for statistical processing, in which 'France' and 'the French' would mean something more than Paris and a few powerful individuals." It is a fine objective, but dozens of books in his own bibliography -- by Braudel, Corbin, Weber, Pierre Goubert, Robert Darnton, and many others -- have done precisely this. Robb does seem to have paid close attention to one of them, Weber's pathbreaking study Peasants Into Frenchmen: The Modernization of Rural France, which appeared in 1976, for he repeats a good many of its arguments about how a culturally isolated countryside became truly integrated into the nation only in the late nineteenth century. But he ignores nearly all of the vigorous debates that Weber provoked (I participated in some of them), which ended up overturning or modifying many of his initial conclusions. And so Robb ends up trying to resuscitate a vision of pre-modern French rural life that serious scholarship has by now thoroughly discredited, while missing the ways in which this scholarship has rendered the subject not just far more complex, but also far more interesting than its Romantic eulogists ever suspected. The Discovery of France is a case study in why academic history, for all its flaws, still matters.
Am I being too harsh? Robb's book, after all, is not an academic monograph. Indeed, he presents it as a sort of whimsical, historically informed guidebook that will help travelers discover France on their own. He calls it the fruit not only of "four years in the library" but also of "fourteen thousand miles in the saddle," traveling the length and breadth of the country on a bicycle. His title refers not just to efforts by historical geographers to map and to describe the country between 1789 and 1914, but also to his own physical and intellectual journey. And the book is certainly lively. Yet Robb's readers have the right to expect not just an engaging account of the French past, but a well- informed and accurate one, particularly given his insistent claims to be overturning conventional wisdom and revealing hidden secrets about history. Robb presents his own research as a sort of Romantic quest that allowed him to recover truths that the dusty scholars missed or forgot. It is a charming (and ancient) conceit. The trouble is that when it comes to a subject as large and complex as the one he has taken on, it is nearly impossible for knight-errant research of this sort to come to any truly important or original conclusions, no matter how many years in the library and miles in the saddle the researcher puts in. If you don't rely upon and engage with the collective accomplishments of serious scholarship, you are going to get the story wrong. Not every way of discovering France is equally good.
The problems arise above all in the first part of the book, in which Robb gives his overview of what he calls the "undiscovered continent" of pre-modern France -- a continent "that had yet to be fully colonized." He casts the story as one of a vast and colorful diversity, which he contrasts to the monotone sameness that casual readers and travelers supposedly encounter in the France of our own day. Like Eugen Weber, he explains that customs, dress, and even language once varied from region to region, and sometimes from village to village. He vividly describes lost "tribes" such as the "Colliberts" who lived in the marshes of the Atlantic coast, rarely setting foot on dry land; and the "Cagots" who constituted a virtual caste of untouchables in the southwest; and the "Polletais" fisherfolk of Dieppe who "spoke a dialect that was barely recognizable as a form of French."
Yet Robb fails to take this diversity seriously. For, unlike Weber, he soon starts dealing in generalizations that cover the entire country and several centuries of history. Robb actually knows quite well that this sort of approach is roundly ahistorical: in his first chapter, he writes reprovingly of existing scholarship that the people "who made up the 'rural' three-quarters of the population received historical rather than anthropological attention only when they began to think of themselves as French." (I would like to see Robb make this charge stick against any real scholar of the subject of the past forty years.) But he then proceeds to practice precisely what he preaches against, repeatedly implying that all parts of pre-modern French rural society lived according to essentially the same unchanging rhythms, until the advent of a modern state and economy wrenched them into history some time after 1800.
Here are some examples of the way Robb proceeds. When talking about pre- modern Catholicism, he offers the blanket assertion that "Jesus Christ was a relatively minor figure," and characterizes popular religious practice as everywhere dominated by pre-Christian and quasi-Christian beliefs: "a world of saints and fairies." He claims that French peasants -- all of them, presumably -- took "season-long siestas" in the winter, spending much of their time asleep, like hibernating animals. He insists that only the inhabitants of the Parisian basin thought of themselves as "French," and adds: "There was no deep-rooted sense of national identity."
In fact, each of these large generalizations collapses as soon as one looks more closely at the actual history. Calling pre-modern Catholicism heavily pagan ignores the Counter-Reformation, which not only did a great deal to formulate the dubious thesis of "pagan survivals" in the first place (so as to justify a massive enterprise of evangelization to overcome them), but also introduced hugely significant changes into the practice of rural religion, bringing much but not all of rural France surprisingly close to what would now seem conventional forms of Catholic observance. This transformation has been the subject of long and careful examination and debate by historians such as Jean Delumeau and Dominique Julia, and religious sociologists such as Gabriel Lebras (on the subject, all of whom Robb fails to cite). As for the winter "hibernation," it not only plays down the considerable work that French peasants did in the winter repairing tools, making clothes, tending to animals, and traveling outside their villages as hired laborers; more importantly, it also ignores the crucial phenomenon of proto-industrialization. As historians such as Liana Vardi have shown, in some areas of northern France already in the eighteenth century a majority of rural households had abandoned subsistence farming to become weavers, producing textiles on a fairly large scale as part of extensive commercial networks based in the cities.
Robb's big assertions about national identity do not jibe with the fact that even in the middle of the eighteenth century French military victories were cheered across the country, including in many rural areas. Or with the fact that millions of rural French people proudly asserted their identity as "French citizens" in the festivals of the French Revolution. Robb does not consider the possibility that these men and women might have stressed their regional identity when dealing with Parisians, but their French identity when dealing with foreigners, as the historian Peter Sahlins showed in his book Boundaries: The Making of France and Spain in the Pyrenees (also not cited by Robb). Sahlins's fascinating work, which would have both enriched and complicated Robb's story, suggested that a strong sense of national identity first emerged in the borderlands -- for instance, as villages sought the support of the French state in property claims against neighbors living on the other side of international frontiers. Again and again, scholarship on the subject has uncovered dynamic and interesting historical patterns that Robb's overarching generalizations simply cannot encompass.
Robb runs into these problems not only because he ignores so much important historical writing on pre-modern France, but also because of the way he reads his sources. Throughout his book, he adopts a frustrating and contradictory strategy of reading: he chides earlier observers for perpetuating stereotypes and myths about the peasantry, but then turns around, when it suits his purpose, and cites some of the most myth-laden sources as empirical descriptions of reality. Thus he oddly asserts that "some towns and villages were flourishing democracies when France was still an absolute monarchy," a claim belied by any number of histories of French rural life. Robb bases his claim largely on an account of a Picard village called Salency, written by a Norman merchant who visited it not long before the French Revolution. He is apparently unaware that Salency had been a fashionable destination for proto Romantic pilgrims ever since the novelist Félicité de Genlis "discovered" it in the 1760s, and enthused about the pristine pastoral charm of its annual "rose festival." Salency's villagers benefited considerably from this attention, but before the Revolution they still lived under a heavy burden of feudal law, which among other things forbade them from marrying outside their parish, and subjected them to the rule of their overlord and parish priest. The merchant knew all about Salency's cult status, and was eager to add to its legend.
Throughout his book, Robb repeatedly quotes travelers' accounts, novels, polemical histories, and political petitions without asking certain elementary questions about them. How were their descriptions shaped by generic convention? How were they influenced by their author's politics or aesthetics? What were the circumstances of their composition? Not surprisingly for a biographer of Victor Hugo, Robb draws with particular delight on Romantic literature for source material. But of course the Romantics liked nothing better than to portray rural life as isolated from the pernicious effects of the city and commerce, and dominated by the rhythms of the natural world, and maintaining a close connection to a wild, ancient pre-history. These preconceptions derived directly from their artistic and philosophical ideals, and colored everything they wrote on the subject. Robb correctly chides some of his authors for their condescension toward the peasantry, but otherwise he tends to read their characterization of rural life with a surprising lack of skepticism. And so he ends up perpetuating many myths that the Romantics devised.
Admittedly, it can be very difficult to grasp the complex realities of pre- modern French culture, and it is worth looking more closely at one particular chapter of The Discovery of France to see why. The fourth chapter is amusingly titled "O Óc Sí Bai Ya Win Oui Oyi Awè Jo Ja Oua" -- all different ways of saying "yes" in different French dialects and local languages. The point is to illustrate the idea that pre-modern France was a crazy quilt of different tongues, from Basque and Breton to Occitan (the distinct Romance language that dominated southern France, of which Provençal is the best-known dialect) and dialects of French itself. As a result, a great degree of misunderstanding occurred, often with comic results.
This characterization of French linguistic history is largely correct in a basic sense -- although it is also utterly familiar to scholars at least since the publication of Ferdinand Brunot's fundamental Histoire de la langue française-des origines à 1900 in the early twentieth century. (Robb does not seem to know it, although nearly all his secondary sources draw their evidence largely from it.) Yet the characterization is also woefully incomplete because of two further points, which historians and sociolinguists have often made, but which Robb largely misses. The first of these, as the existing scholarship amply recognizes, is that different local languages and dialects had greatly different histories and roles in French society. The German of Strasbourg and the Flemish of Dunkirk were closely related to literary and administrative languages spoken just across the border, and prevailed in towns and cities as well as the countryside. Occitan existed as a literary and administrative language until the progress of royal absolutism pushed it aside in favor of French. Breton and Basque did not exist as commonly written languages until modern times, and were not much spoken by local social elites since the middle ages. Some non-French languages, in other words, were exclusively oral, limited to peasants, and easily looked down upon as patois (the derogatory word comes from patte, an animal's paw, suggesting a whiff of the barnyard). Others, above all Alsatian German, managed to compete with French in many realms.
Secondly, the best-known sources have a strong tendency to exaggerate the extent of pre-modern linguistic diversity and mutual incomprehension. Most important here is the man whom Robb seizes as the central figure in his discussion: the famous radical priest Henri Grégoire, who in 1790 sent a questionnaire on local languages to correspondents throughout France and received a fascinating set of responses. (Most of them were published in the 1880s, and have received intensive study.) In 1794, at the instigation of the revolutionary government, Grégoire then composed a legislative report calling for the "eradication of patois" and the establishment of standard French as the country's sole language. In his report Gregoire stated categorically that only three million of France's nearly twenty-eight million people spoke proper French, while listing no fewer than thirty patois that supposedly competed with it. This evidence seems to corroborate Robb's thesis perfectly.
But here's the rub: Grégoire's work was anything but a neutral description of linguistic reality. A less well-known legislative report on language from the same year found serious linguistic difficulties only in four small peripheral regions: Brittany, the Basque country, Alsace, and Corsica. Its author, a Jacobin named Bertrand Barère, himself came from an Occitan-speaking area, but he did not regard the Occitan-French divide as an important one. The reason for the difference is that Grégoire had much higher standards than Barère for what it meant to "speak" a language. As a radical priest, he wanted to accomplish the political equivalent of a religious conversion with the peasantry, something that demanded a very high level of mutual comprehension. From this perspective, it was indeed true that very few people in France spoke sufficiently good French. But the historical evidence suggests that most speakers of Romance dialects could conduct basic market transactions in French, obey basic French military orders, understand the gist of French sermons or court proceedings, and so forth. Even in Celtic Brittany, French was widely understood, although barriers to comprehension were much higher.
In sum, claims to vast degrees of linguistic difference have to be taken with a great deal of salt. But Robb gives credence not only to Grégoire, but also to later groups that had an even more obvious interest in exaggerating France's language problems, notably educational officials seeking to justify vast projects of language instruction and regionalist militants seeking to justify political autonomy on the basis of linguistic difference. The scholarly advances in this field have depended in part upon relatively technical work in sociolinguistics, but it has been clearly summarized by historians, whom Robb could easily have consulted had he not been so enchanted by his own Romantic vision.
In the second part of his book, Robb recounts how educated, largely urban elites discovered his "undiscovered continent" between 1789 and 1914. They did so, he stresses, largely to pave it over, reducing its colorful and deeply human diversity to a bland and uniform sameness. "Pave over" should be taken rather literally, for Robb places a great deal of emphasis on transportation, and is fascinated by the speed at which news and people traveled. At the time of the French Revolution, he notes, even high-speed messengers averaged only seven miles per hour over long distances. Road construction allowed for some improvement (it actually started well before the Revolution), but only the railroads brought real change. In 1828, France had fourteen miles of railroad track; sixty years later it had 22,000. And as the railroads connected previously isolated areas to major cities, an aggressively centralizing state began sending its "colonial" forces into the interior. "The Third Republic," Robb writes, sought to "erase local cultures," notably by "[massing] its pedagogical army on the wild frontier of language."
It is an attractive story (and also a close cousin to the one that Eugen Weber told thirty years ago). But once again, it is more than a little incomplete and misleading. As Tocqueville recognized a long time ago, the central state had a significant presence in rural France well before the nineteenth century. Meanwhile, even without railroads, a surprisingly large number of French people routinely traveled great distances long before the nineteenth century. The historian Olwen Hufton has shown that millions of oftendesperate French men and women took to the roads each year in the 1700s, walking tens or hundreds of miles to find work. Shepherds took flocks across international borders. And as for artisans and servants, their established migration routes led not only to major cities, but even across the Atlantic, to French Canada and the Caribbean. Elsewhere Robb repeats the familiar claim that Third Republic schoolteachers harshly punished children for speaking patois in the classroom. Yet the historian Jean-Francois Chanet has shown that such stories have been exaggerated. Moreover, some parents wanted their children punished for speaking patois, because they recognized the enormous social advantages that came with fluency in the national language.
Once again, one can only wish that Robb had engaged more seriously with the dusty old library. Hufton's magnificent study of the eighteenth-century poor contains the sort of stunning original material that begs for a writer of Robb's talent to convey to a larger public. Her vision of millions of men and women struggling to get by through an "economy of makeshifts," many of them tramping at their tragically slow pace through the French countryside, often reduced to petty theft or prostitution to survive, is as colorful and as powerful as Robb's vision of millions of men and women living their entire lives in virtual isolation, rarely encountering strangers or leaving the valleys of their birth. But Hufton's tale has the advantage of being supported by the evidence.
The Discovery of France has charmed many people because it is not, in truth, a discovery at all. For all its claims to novelty, its Romantic vision of a lost world is a very familiar and oddly comforting one. We like to see the past as a wild, colorful English garden that contrasts sharply with our own electrically lit, tightly scheduled lives. We like to imagine a world that may have been brutal and unfair, but was also close to nature and full of mystery and color and life -- a world of saints and fairies, populated by idiosyncratic isolated "tribes" like the Colliberts and the Polletais. Above all, it is delightful to think that we can get tastes and glimpses of this world by pedaling through charming stretches of the French countryside, with someone like Graham Robb as our entertaining guide. If only it were so. But the work done by scholars since the Romantic era has shown that the French past was really a very different sort of place -- strange and mysterious and beguiling, but not in the way that the Romantics imagined. And we can hope to grasp its reality only through hard digging in the archives and the libraries, and through the toil of historical analysis. A bicycle will not take us there.
David A. Bell is a contributing editor at the New Republic.