Migration and the French Colonial Atlantic as Imagined by the Periodical Press, 1740–61

Among the major colonial powers of the early modern Atlantic world, France stands as an anomaly for the notoriously small numbers of colonists it sent to its empire, especially to continental New France. In 2006, historian Yves Landry suggested that a negative image of the colonial world in French print media might have had a dampening influence on the enthusiasm of potential colonists, but no studies to date have been conducted to confirm or invalidate his hypothesis. This article presents the results of a dual-scale analysis, part digital and part traditional, of a corpus of Old Regime periodicals. Extracted from Gallica (the French National Library’s online archive), this corpus of approximately 75,000 printed pages features the three semi-official periodicals published under King Louis XV: the newsoriented Gazette, the literary Mercure de France, and the scholarly Journal des savants. Text mining such source material poses significant methodological challenges, due in no small part to the irregular spelling of the early modern era and to the poor quality of the Optical Character Recognition (OCR) data obtained from documents nearly three centuries old. After a brief outline of the techniques applied to alleviate these problems, the article presents a three-layered model of the French colonial Atlantic as represented by the periodical press. The base layer, outlined by the common knowledge about America displayed in such evergreen types of content as word games, poetry, and stage plays, shows an unchanging image of the New World as an alien land filled with riches and glory for the few, mortal threats for the many, and the best opportunities, perhaps, set aside for anyone but the French. A second layer, made up of the periodicals’ coverage of the slow production of knowledge through science and exploration, edulcorates this picture to some extent by showing that America is in the process of being domesticated, but that this process is very much still in its infancy. Buyer, beware. Finally, the top layer, represented by the Gazette’s news coverage, shows a French colonial world that is dominated by Britain, virtually invisible to the reading public in peacetime, and fraught with chaos at every moment. It seems that the overall picture of French America painted by the periodicals would have had little appeal for the reading public, a finding that is 1 A doctoral scholarship provided by the Fonds de recherche du Québec — Société et culture financed this research. Journal of European Periodical Studies 4.1 79 largely compatible with Yves Landry’s original hypothesis, at least for the two decades leading to the Seven Years’ War.

Migration and the French Colonial Atlantic as Imagined by the Periodical Press, 1740-611 the Spanish sent to theirs, a proportion that falls to a fifth when compared to the British and to a mere five percent when compared to the Portuguese.4 In a 2006 study of the migration to Canada, historian Yves Landry lists some of the common explanations for this phenomenon: a perceived lack of opportunities for enrichment, the French army's unquenchable thirst for restless men, and a royal government equally afraid of depopulating the metropole and of filling a colony with Protestants who had very good reasons to leave. He also suggests (but does not investigate) a new one: bad press. By transmitting to the reading public a negative image of Canada as a cold and inhospitable place, print media would have served as a deterrent to colonial ambitions.5 This paper tests Landry's hypothesis, extended to the whole early modern French empire, by applying text and data mining techniques to a corpus of periodicals published between 1740 and 1761 and comprising some 75,000 pages. As we will see, the results of this enquiry are compatible with Landry's hypothesis, although the relatively small circulation of pre-Revolutionary periodicals implies that their influence must have been limited. These results also allow us to characterize the imagined geography of America as constructed by the most important periodicals of the late Ancien Régime.
The core of this article, which follows a short description of the corpus, the methods applied to it and their respective limitations, is made up of three parts. First, an analysis of the news published in the Gazette shows that French America is depicted as a dangerous place, and that it is largely invisible in peacetime. Second, the periodicals' coverage of the production of contemporary knowledge about America shows that the so-called New World's most appealing qualities are to be found in rival empires rather than in lands under French control. Finally, the common wisdom embodied in the periodicals' coverage of the arts and even in games shows French America as an eternally alien land where opportunity is restricted to the few and unpleasantness abounds for the many. As these angles of approach build upon one another, a rather foreboding portrait of French America emerges: one that would have been unappealing to most readers and unlikely to incite them to migrate.

Corpus
The corpus under consideration for this study is made up of the three most prominent periodicals published in France during the reign of King Louis XV: the Gazette, the Mercure de France, and the Journal des Savants. It covers the period between 1740, when the French print industry and its potential influence on the public started growing rapidly, and 1761, when the British conquest made French migration to Canada impractical. The weekly Gazette enjoyed by far the largest circulation of any French periodical at the time: 15,000 to 17,000 copies in 1760.6 With some copies read aloud in cafés and private reading cabinets, and others passed from one reader to the next along sometimes lengthy chains of ownership, the Gazette may have reached an audience of 100,000 people or more, not all of them literate, with peaks of readership reached in wartime.7 Its popularity was rooted in its exclusive right to the publication of first-run political news. Content-wise, however, it had little in common with today's newspapers: Gazette articles were unsigned, varied in length between a few lines and several pages, and often discussed several unrelated topics. In the 14 July 1753 issue, a summary of Parisian news featured a baptism, an obituary, a cardinal's visit to Court, a banquet followed by fireworks, a military inspection tour, a list of colonial goods recently disembarked at Marseille, an account of recent privateer activity, a quote for Compagnie des Indes stock, and the results of the royal lottery.8 Elsewhere in the same issue, an article originating from the German town of Regensburg merely stated: 'The Diet has ordered that from this day forward, the Police Magistrate will forbid private individuals' coaches from displaying the honour marks assigned exclusively to the Ministers' coaches.'9 To a modern eye, the result is akin to reading a chronicle or a collective diary.
The Mercure de France was as close to a general purpose monthly as anything that could be found in the eighteenth century. A typical issue would include over 200 pages of poetry, book reviews, reports from academies, songs, open letters, excerpts from the classics, advertisements, embellished reprints of news previously reported by the Gazette, and word games. Not everyone agreed with this eclecticism; already in June 1721, the Mercure's editor had to defend this policy, stating that 'opinionated censors […] insist on treating as a defect the variety that gives it its character. Do they not know that this journal is made for everyone, and that it must provide dishes for every taste?'10 Censors' opinions notwithstanding, the Mercure was widely read among the upper classes, with a circulation estimated at 1,600 copies in 1763.
Finally, the Journal des Savants was the Western world's first scholarly publication, beating the Philosophical Transactions of the Royal Society to the press by a few months in 1665 and inspiring some sixty imitators by the end of the seventeenth century.11 Its circulation is estimated at 1,000 copies. The Journal printed excerpt-heavy, long form book reviews intended for readers who could not obtain the source material themselves, along with open letters from scholars and news from the publishing industry. Its reviews tended to be cordial, as its writers came from the same academies as the authors whose work they discussed, but a sarcastic streak occasionally seeped through. When the Abbé Prévost invoked Plato's mention of an unknown continent to the west of Atlantis to suggest that the Ancients might have been aware of America, the Journal's reviewer noted that, according to Plato, Atlantis was the size of Asia and Africa combined, and that Prévost 'did not reflect that an island of [such size] located between Europe and America would have left little room for the sea'.12 Digitized versions of all three periodicals, in PDF and in plain text generated through Optical Character Recognition (OCR), are freely available from Gallica, the French National Library's online portal. Unfortunately, the archive is incomplete for the Mercure, whose digitized run ends in 1758, and the Journal, for which a consecutive set of annual compendia is only available from 1751 to 1759. Nevertheless, the amount of data available for digital analysis is considerable, at approximately 75,000 printed pages, the Mercure accounting for two-thirds of the total.
As a model of the information available to the French reading public at the time, this corpus is of course incomplete. It does, however, contain the three most important periodical sources of information that the government wanted the people to be able to read. Therefore, it features the internal coherence, explicit boundaries and lack of ambiguity suggested as preconditions in Michael Piotrowski's framing essay elsewhere in this issue.13 Further, as tightly controlled, semi-official publications, these periodicals would be expected to reflect the royal authorities' attitude towards the colonies, which as mentioned earlier was lukewarm at best. Finally, the periodicals' message would have reached a fairly diverse audience with a corresponding diversity of potential interests in the colonies. While most of the readers of the Journal des savants and the Mercure belonged to the upper classes of society and would have been unlikely to entertain the idea of permanent migration to America, some of them would at least have been open to investing in colonial plantations as absentee landlords or to send second sons across the Atlantic on temporary fortune-making ventures. Meanwhile, the Gazette achieved significant penetration of the middle and lower classes, including skilled artisans who would have been highly desirable immigrants from a colonial standpoint and struggling workers who might have been tempted to cross the ocean in search of better lives. Thus, we would expect the message transmitted by the corpus to be relatively coherent and (at least potentially) influential.

Methodology
Text mining the Gazette, the Mercure de France and the Journal des Savants poses significant challenges. From a twenty-first century perspective, all early modern texts suffer from three types of defects described elsewhere by Piotrowski: spelling variance caused by the absence of standards, changes in spelling over time, and transcription glitches caused by faulty OCR or by hand-keying errors.14 French historical texts are especially prone to the second of these defects, with such ubiquitous words as 'avoit' ('had'), 'étoit' ('was') and even 'françois' ('French') having exchanged an 'o' for an 'a' in the nineteenth century. Further, OCR quality is irregular for the three periodicals studied in this article because some of the original documents were printed with non-standard type, damaged by centuries of exposure, or preserved only as misaligned or cropped microfilms.15 A variety of context-dependent strategies to deal with these issues, from crowdsourcing transcriptions to sidestepping the raw OCR results using hand-crafted metadata, have been described in scientific literature.16 A full discussion of the strategy employed in this study is beyond the scope of the present article and can be found elsewhere.17 Suffice it to say that bespoke software based on Levenshtein's algorithm allowed the recovery of several hundred instances of some twenty-three keywords related to the colonial Atlantic ('Canada', 'Amérique', etc.) that had been written in unusual ways in the sources or damaged by OCR; these keywords had themselves been selected to minimize the repairs necessary to render the text mining process effective.18 Keyword instances were identified in 1,184 articles from the Gazette, 595 from the Mercure, and ninety-four from the Journal. These articles were extracted from the corpus and submitted to visual inspection, which revealed that the Gazette articles' cities of origin and years of publication did not follow expectations; specifically, a vast majority of these articles were sent to the publisher from outside France and most of the articles of French origin were published during wartime. (A purely digital process would not have noticed this pattern because cities of origin are italicized in the Gazette and nearly unintelligible to OCR.) Much of the analysis in this paper will be based upon this metadata, which was corrected by hand. Unfortunately, no similar metadata patterns emerged from the Mercure or from the Journal. Digital analysis of the Mercure and Journal articles therefore had to be limited to simple methods applied to the raw OCR data, like collocates and keywords in context. Since the OCR process yielded relatively poor results for the Mercure and the Journal, all digital results were then validated visually.  2014). Various types of searches focusing on specific patterns and/or fuzzy logic can also be used to mitigate OCR errors. The Viral Texts project, which looks at sequences of five words ('5-grams') in order to identify possible article reprints between English-language periodicals, has employed one such strategy to trace an overall picture of text migrations between periodicals even though the strategy admittedly misses a significant share of the reprints; see David A.

News from Overseas: America as a Foreign and Dangerous Land
For French readers of the Old Regime, the Gazette was the primary source of fresh news. Other French periodicals were locked out of the news market by the Gazette's royal privilege, which merely allowed them to reprint stale content from the Gazette after weeks of delay (and for a fee). French-language gazettes published in the Netherlands or in the papal enclave of Avignon were under no such restrictions, but their circulation was limited to a few hundred (or, in the case of the Gazette d'Avignon, a few thousand) copies by high postal rates and by transit delays that made much of their content obsolete by the time it reached its audience. Certainly, few Frenchmen who could afford to subscribe to a foreign periodical would have done so without also subscribing to the Gazette for a fifth or a tenth of the price.19 As a weekly presence, the Gazette had a major influence on readers' mental image of the world. In the case of colonial America, some of this influence may have been a consequence of the origins of the articles discussing the colonies, the distribution of these articles in time, the specific colonies they discuss, and the vocabulary they employ to do so.

Article Origins: A View from Away
A map of the origins of the 1,184 Gazette articles that discuss colonial America (Fig. 1) shows a surprising pattern: sources based in London account for an absolute majority of the dataset and for three times as many articles as all of France combined. London dominates even the subset of articles that explicitly mention the French colonies of Guadeloupe, Martinique, and Canada, or the cities of Québec and Montréal; the only exception is Saint-Domingue, which appears roughly as often in articles written on both sides of the English Channel. This is not without consequence, as many of the London articles seem to be clippings from the British press, sent by the Gazette's local correspondents and translated so closely that they even refer to British ships, colonies and armies in the first person. Three months after the British capture of Québec during the Seven Years' War, for instance, the Gazette prints that French troops will soon besiege General Amherst in the city: 'La garnison que nous avons dans cette Capitale du Canada, est résolue de se bien défendre; mais la Place est en trop mauvais état pour résister longtemps à une attaque vigoureuse.'20 Thus, the Gazette presents a British view of American events to the French public more often than it presents a French one. The Iberian peninsula also accounts for more articles than France, partly because of frequent mentions of the prince and princess of Brazil in high society news; this sign of rhetorical appropriation of the American space has no counterpart on the French side, as there is neither a prince of Canada nor a duke of Saint-Domingue to imprint the importance of France's colonial domains in the readers' imagination.
Further analysis of the metadata reveals that none of the keywords occurs more frequently in articles written in France than in news briefs from abroad (Fig. 2). Neither 'America' as a general concept, nor the word 'colony', nor any of the colonial establishments themselves, French or otherwise, are discussed primarily from a French point of view. The Gazette prints news about Jamaica more often than about all of

Temporal Distribution: French America as a Theatre of War
Plotting a time series of Gazette articles originating in France (Fig. 3) reveals another odd pattern: these articles are printed far more often when the colonies are at war (1744-48 and 1754-61) than when they are at peace (1740-43 and 1749-53). This is a trend that the entire corpus of 1,184 articles does not exhibit. Nor is it a statistical accident: a chi-squared test shows that the probability of such a skewed distribution being the result of random variation is less than seven in ten billion.22 Fig. 3 Gazette articles about America originating in France, divided by year of publication When dates of publication and places of origin are cross-referenced, we observe that peacetime articles are five times more likely to come from London than from France, that Spain takes over second position, and that even Portugal occupies a close fourth. French colonies appear in a mere fifty-two peacetime articles of all origins combined, compared with 170 articles for Jamaica, Philadelphia and Boston. Canada, in particular, is virtually invisible, with only fourteen mentions in the entire peacetime subcorpus: seven instances of the keyword 'Canada', two of which appear in articles about the compensation paid by the British crown to its subjects after an invasion; seven instances of 'Québec'; and none whatsoever of 'Montréal'. A study of collocates (i.e., words that appear within a short distance of a keyword) further reinforces the martial and foreign nature of the articles about America. Words that appear particularly often within five places of the word 'colonie' ('colony') include both 'anglaises' ('English') and 'françaises' ('French'), but in absolute numbers the former appears three times more often than the latter. Of all the words that appear in the same sentence as 'colonie', the most common is 'troupes' ('troops'), with sixty-six occurrences. Furthermore, the most important collocates for 'Canada', within a distance of ten places or within a single sentence, include 'sauvages' ('savages'), 'troupes', 'marquis' (which refers either to French general Montcalm or to Governor General Vaudreuil), 'fort', 'Vaudreuil', 'Montcalm', 'men' (as in soldiers), 'détachement' ('detachment'), 'ennemis' ('enemies'), 'camp', 'général' and 'anglois' ('English'). And while Louisbourg, the major French fortress on Isle Royale (today's Cape Breton, Nova Scotia), appears in the corpus too rarely to carry much statistical significance, it is worth noting that, among the four words collocated with it more than ten times, 'guerre' ('war') is the most common, followed by two forms of (naval) 'ship'. For the Gazette's reader, there is very rarely such a thing as a peaceful continental New France. The Caribbean colonies enjoy a slightly higher profile, with seventy-eight peacetime mentions, mostly about shipping to and from the metropole. But mentions of natural disasters, shipwrecks, and crimes give the reader the impression that peacetime in the Caribbean can be a relative term. An article sent from Nice announces the arrest of an 'impostor who, under a false title, abused the confidence of several people in Martinique'23 while the Parisian news for 15 July 1752 mention the wreck of the Soliman, inbound from Saint-Domingue to Le Havre, whose crew managed to escape to a nearby ship.24 Collocates for 'Martinique' include, along with the expected 'navires' ('ships'), 'vaisseau' ('vessel') and 'chargé' ('loaded'), the word 'conduit' ('conducted'), that only appears in the context of a ship being conducted somewhere against its will after being captured by privateers, and the words 'pris' ('taken') and 'prises' ('prizes'), also associated with piracy and naval combat. Thus, regular readers of the Gazette are likely to perceive the French Caribbean, like New France, as a land of violence and danger.

Thematic Structure and Mental Distance
Approaching the corpus from the opposite direction, that is, looking at article origins as a function of thematic content instead of the reverse, allows us to refine the overall picture. A k-means clustering of the corpus's metadata into five clusters shows that three of the clusters, including the ones characterized by the presence of the keyword 'Amérique' (' America'), by the presence of the keyword 'colonie' ('colony'), and by the presence of British colonies (primarily Jamaica), are overwhelmingly British in origin.25 A fourth cluster, characterized by the presence of Brazil, understandably shows Spanish and Portuguese predominance. The fifth cluster (Fig. 4) is a catch-all in which French colonies appear more often than their foreign rivals, but not overwhelmingly so; even here, articles of French origin are in the minority (Fig. 5). Nowhere is a colonial topic of any kind discussed primarily as a French phenomenon. The copresence graph in Fig. 6 summarizes the conclusions that can be drawn from the Gazette metadata. Each vertex represents a keyword, a city of origin, or a date of publication. Edges represent the forty pairs of vertices that coexist in the largest numbers of articles; the width of an edge is roughly proportional to the number of articles that share both metadata items. A central clique containing London, America, the keyword 'colonie' and (to a lesser extent) Jamaica dominates the corpus. Most colonial establishments, including the French colonies of Québec, Canada and Guadeloupe, are linked to London but not to Paris or to any other French city. Fig. 6 A copresence graph that shows the dominant association between London, America, and the colonial world within the corpus While every French colony in the graph is linked to London, not a single foreign colony is linked to France. Similarly, ten years of publication are linked to London, indicating relatively numerous discussions of colonial America in articles received from the British capital; none are linked to Paris. The Gazette thus suggests to its readers that the mental distance between France and its colonies is much larger than that between Britain and its own empire, or even that the French empire is somehow a more important presence in London public discourse than in its French counterpart. It is but a short step from there to the conclusion that France had already lost the (mental) battle for America, long before the Treaty of Paris that put an end to the Seven Years' War.

Production of Knowledge: America as Land of (Some) Opportunity
In the Mercure and the Journal, curiosity about America abounds. Knowledge about the New World is published, debated, applied, and contested as quickly as it is produced. From this bubbling cauldron emerges a picture of America as an unfinished land, lacking Europe's sophistication but definitely worthy of attention: a place on the rise. Two primary types of content promote this message and thus imprint upon the reader a sense of slow but durable (and promising) change: maps, atlases and the descriptive articles that sometimes accompany them; and scholarly discussion of the New World's natural resources.

Mastering the Land through Cartography
By the mid-eighteenth century, European geographical discourse had fully integrated the concept of a world divided into four parts: America, Asia, Africa, and Europe. Scholarly treatises reviewed in the periodical press were explicitly divided into four chapters or four sections along those lines;26 the four parts of the world had even become a kind of idiom, used in all manners of contexts. L'Univers énigmatique (The Enigmatic Universe), a book of devotion reviewed in the Journal in 1759, claimed that 'Deism is anathema in Africa, in Asia, in Europe, in America, for all mankind'.27 The plans for the proposed Place de Mars (now Place de la Concorde in Paris) featured images of the four parts on four of its eight sides, the others being dedicated to the seasons.28 Even the publishers of card games replaced the traditional suits with the four parts of the world, and individual card designs with facts and figures, in order to teach geography in a pleasant manner.29 And where one part of the world was mentioned in the Mercure, the others often followed. All three Old World continents are on the list of the ten most frequent collocates for 'America'; twenty-one of forty-six occurrences of 'Asia' and thirty-three of sixty-six occurrences of 'Africa' can be found within ten places of it.
Geographers and their readers were well aware, however, of the knowledge imbalance between these four parts of the world. For example, when Robert de Vaugondy published an essay on the history of geography, reviewed by the Mercure in 1755, only five of the one hundred maps included in the book represented the New World, whereas Europe accounted for seventy-two.30 Thus, any new tidbit of information about America excited curiosity, not to mention publishers' greed. Nowhere was this more obvious than in the frequent announcements of new maps, plans, and atlases, some of them accompanied by travel narratives and other explanatory texts, that could be found in the Mercure. The corpus contains over thirty of these semi-advertorial articles of all shapes and sizes, in which the editors hyped the 'details that one could not have seen in any of [the maps] published before.'31 Utility was the prime argument pushing sales. In 1751, navigators and merchants were invited to acquire a new map of Saint-Domingue in which coastlines were printed in black and prevailing winds in red, thanks to the careful alignment of two plates on the same sheet of paper, one after the other. A most delicate process, wrote the Mercure's editor, because 'if the slightest disturbance were to occur, the winds would no longer align with the latitudes and longitudes of the map, which would become false and would be of no use to navigators'.32 Left unwritten in such advertisements, but undoubtedly understood by the less adventurous part of the readership nonetheless, was the suggestion that whomever left for the New World with anything older or less precise would be risking their lives.
In such contexts, the meaning of 'utility' was often stretched to include maps that could be used as props to enliven news reading. Several maps depicting the locations of current fortifications and colonial establishments were published during the Seven Years' War, ostensibly to provide readers with a convenient way to visualize the narratives of current events found in the Gazette. This undoubtedly reinforced the image of colonial America as a battlefield.33 Despite these maps' relatively high prices (an eight sheet map of North America announced in 1756 cost between twelve and eighteen livres, depending on paper quality, which amounted to 150 to 250 percent of the price of a one year subscription to the Gazette),34 demand seems to have been high enough to encourage some cartographers to rush their work to beat competitors to market, or at least to advertize upcoming maps before they were ready for print.35 And with so much of America still barely explored, neither the periodicals nor (presumably) the public would hold the cartographers to unreasonable standards of perfection, as long as they were forthright about their sources. The Journal was particularly well disposed towards cartographers who 'accompany […] the maps they bring to light with treatises that serve as vouchers for their work.'36 For example, in 1756, Jean-Baptiste d'Anville received praise for the way in which he explained how he calculated Quebec City's longitude, based on new measurements taken in Boston and on estimates of the distances between several fixed points in North America; the editor further lauded his honesty as d'Anville 'does not hide the fact that not everything enjoys the same degree of certainty' and admitted that certain parts of his map had been drafted based on no more than 'a simple notion' obtained second or third hand from Indigenous peoples.37 Thus, maps of France's colonial domains may have enhanced, rather than diminished, the overall aura of uncertainty about them. An aura that some would have read as exciting, others as dangerous.

A Space to Inventory
Regular readers of the Mercure and the Journal were also exposed to a vision of America as a land of botanical plenty. At various times, the periodicals reported that a surgeon in Martinique had found a way to extract indigo from a new plant species, more resistant to wind and insects than previously known sources;38 that Canadian squashes exhibited such interesting nutritional properties that botanists were trying to acclimate them in France;39 that a beer brewed from Canadian yew and berries had 'soft and benign' purgative properties;40 that the leaves of a tree found in Louisiana yielded a tea that purified the blood, eased miction and softened the pain of gout;41 or that Canadian fir beer cured scurvy.42 If a grumpy writer could sometimes be found to dismiss the need to know plants that 'being so foreign to our lands, seem to regret their homes if we judge by their dejection',43 the message transmitted to readers was usually that 'plants are a perpetual school of [the blessings of ] Divine Providence'.44 This tantalizing vision was reinforced by the suggestion of mineral riches yet unimagined hidden in the vastness of America. The discovery of the new metal platinum in the Spanish colonies inspired an academician writing in the Journal to exclaim: ' After all that men have done to rip from the Earth its treasures, from Nature its secrets, it would have been hard to believe that a new species of substance would overthrow our systems'.45 A flurry of articles about the new metal's properties and applications ensued, undoubtedly whetting the readers' appetites. For if there was one entirely new and unknown metal in America, why not more?
However, the sober reader could not have helped noticing that the new discovery had been made hundreds of leagues away from the nearest French outpost, in Spanish lands that already harboured the lion's share of the New World's gold and silver. Indeed, when the Mercure published a dictionary entry on mining, it was the diamonds of Brazil and the gold of Peru that featured in prominent places, not the meager resources of French America.46 When New France had proven so frustrating to mineral hunters that 'false as a diamond of Canada' had entered the vernacular, was it worth taking the risk of death, disease, and isolation to seek a fortune that would most likely remain a mirage, or at best turn into a recipe for some sort of medicinal brew?

Eternal America: Culture and Collective Consciousness
When we shift our focus towards the kind of Journal and Mercure content that reflects the age-old common wisdom about the New World that everyone in polite society was expected to know, the picture becomes both clearer and less appealing. Indeed, while this content tends to show some opportunity for the elite, it spends far more time on the perils of isolation, cultural alienation, and an unforgiving climate that strike everyone equally. Combined with an enduring image of America as a distant and alien land, this would have done little to encourage migration to the colonies.

Opportunity for Some
The potential benefits of trying one's luck in the French colonies were mostly to be found in the Mercure's enhanced reprints of old news originally published in the Gazette. In order to provide value to its readers without infringing on the Gazette's monopoly, the Mercure would often augment a necrological notice or a marriage announcement with biographical data regarding the glorious history of the people involved. The rewards that these texts advertized, however, tended to be restricted to the nobility and somewhat intangible: for every young woman who married a Norman gentleman 'born in Martinique, where he has considerable property',47 there was a widow whose husband's famous ship captain ancestor 'died in America'48 and a marquis whose grandfather served as commander of a regiment 'for the service of [His Majesty] in Canada' in the 1670s.49 While the prestige of high level colonial service (or, for the unfortunate, of an honourable death abroad) would not have been inconsequential in Ancien Régime society, few potential emigrants would have had access to such honours.
Indeed, few representations of colonial opportunity would have seemed attainable even to the majority of elite readers. When a young female character in a stage play asked her lover to elope with her to Canada (before she discovered that he was a servant and abandoned him), it was because her father, 'once Governor of New France, has left an immense fortune'50 there; rarefied air to say the least. Readers who would have disposed Periodical Press, 1740-61 of enough capital to emulate the rich La Rochelle merchants whose exploits were celebrated by the local academy,51 or who could have been enticed to bring such capital to the foreboding swamps of Louisiana to cultivate silkworms,52 were undoubtedly few and far between. Meanwhile, opportunities for the humble rarely made an appearance in the periodicals. A rare case involved a poor migrant woman named Goëling, on her way to Philadelphia (and not to a French colony) in search of a better life, who was only mentioned in the Mercure because the writer of an open letter approved of the innovative way in which a doctor had just amputated her arm.53 At best, the hopeful upper middle class reader may have been encouraged by an essayist who recommended seeking employment 'abroad or in the colonies' where 'money is not required of an associate, but only good character and talent',54 as long as he had the proper connections. Poor workers devoid of any such connections would have found very little reason to indenture themselves in the hope of making a better life for themselves at the end of their contracts, even in the unlikely event that the printed material would have found its way into their hands through those of a recruiter.

Repulsion for All
However unlikely the siren songs of riches and glory, periodicals rarely missed an occasion to counter their potential attractiveness with tales of woe. Some of these tales focused on the poor reputation of the colonists with whom an emigrant would be forced to mingle in America. A story published in 1755 recalled a past when 'we used to talk much […] about these colonies that we send to America and that serve to purge the State […] of several miserables who, while not guilty enough to deserve death, were guilty enough to make society wish to be rid of them'.55 A past that was evidently not yet gone, as a 1758 essay recommended that the orphans of Paris be shipped to Louisiana, even though 'this export [is] a quite horrible punishment, since one can't imagine any other fate for incorrigible and opinionated beggars'.56 At best, the colonists were portrayed as boring rather than as undesirable. Seeking to beg off the social demands associated with his rank, a man wrote to the Mercure to apologize for the temperament that he shared with a 'taciturn and savage Canadian, with whom […] you would not want to be forced to spend a simple evening head to head for anything in the world'.57 At worst, they were described as lazy and incompetent workers, according to the Mercure, which stated that those who crossed the Atlantic did so 'for lack of knowing some trade' that could let them succeed in France, 'while half of the Kingdom's land is left uncultivated'.58 Only an exceedingly generous reading of such material could have tempted anyone (except maybe a rascal) to make the trip.
If psychological inconvenience wasn't enough of a deterrent, America wasn't lacking in physical danger either. War, of course, was the most pressing. An investor who, by some miracle of selective reading, would have managed to avoid any mention of piracy and privateering in the Gazette, would undoubtedly have been shocked to read of a 1746 ordinance published in the Mercure that commanded the captains of every colony-bound ship to travel in convoys, under threat of 500 livres in fines and one year's (unpaid) service as a simple seaman in His Majesty's navy.59 When a writer accused the British King, in a 1756 issue of the Journal, of 'exposing his subjects […] to a loss of the fruit of their work' through his bellicose policy, how could the reader have failed to see that the same danger, or worse, awaited the French colonists who were the targets of this policy?60 And the account of a lawsuit brought forth by Bordeaux insurers who wanted to retroactively increase the premiums of merchant ships caught in transit by a declaration of war brought up the hazardous nature of transatlantic commerce, both for the merchants and for the colonists awaiting supplies from the metropole.61 Especially since, in America, the memory of a time when 'French companies had abandoned the colonies to dreadful extremities, from which only the Dutch had saved them' was still fresh and brought new resentment of the Exclusif and its high prices at every turn. 62 And of course, America's climates were a threat unto themselves, even in peacetime. The frigid temperatures, 'in Canada as in Hudson's Bay, must equal those of Siberia', according to a memoir presented at the French Academy of Science. 63 The sad tale of a British captain caught on the shores of Hudson's Bay during a winter when 'liquors all froze, even brandy' and when 'the interior of the bedchambers and the beds were covered with a layer of ice several inches thick that [the occupants] were forced to remove every day' certainly did nothing to attract colonists to a supposedly equally inhospitable Canada.64 Meanwhile, in the Caribbean, the threat came from unpredictably violent weather. A poem commemorating an August 1738 hurricane inspired its author to ask if God, repudiating his promise, had not decided to send a new Flood to 'exterminate all of the mortals'.65 Whether freezing in the tundra or cowering under gales and lightning, the colonist must have felt dangerously isolated from home, and the reader of these accounts would have been in no hurry to join him.

An Unchanging Land
When America made an appearance in the Mercure's cultural pages, it was often as an unchanging space unmarked by French presence. Coverage of the Parisian theatre scene provides several examples. Plays shown in 1740, 1743 and 1754 featured set pieces in which dancers and background actors dressed as Indigenous people were the sole representatives of America.66 A 1745 stage directing manual confirmed that plays set in America had to feature ersatz Indigenous people as the only marker of place, 'like Asia must have minarets, and Europe, church belltowers'.67 And when the commedia dell'arte character Arlequin mentioned a plan to run away to Canada to escape 'a very dignified wife [who] in malice is equal to twelve of them',68 it was presumably because the audience understood that even her very long claws could not reach him in such an untamed wilderness.
The word games published in the Mercure provide more evidence that the French presence in the New World was viewed as largely inconsequential. When clues asked the reader to think about an American prince, an American country, or an American city, the answer is the Inca, Peru, Boston or Lima more often than any French counterpart. Periodical Press, 1740-61 The Bostonian example is particularly telling, as the editor of the Mercure stated that it appeared in a puzzle written by a fourteen-year-old reader.69 As for the 'thing that joins France to America', it was not allegiance to the King or a common culture but simply 'the sea',70 while the explorer who had 'never set foot in America' even though he was 'always found in Mexico' was none other than the letter X.71 Inasmuch as the reader was expected to think about America (especially French America) at all, it was as an alien place, and only in the most superficial way.

Conclusion
The publishers of Ancien Régime periodicals had to work under the watchful eyes of the royal administration's censors. The censors' severity varied with time and ministries, but their meddling never truly went away. Thus, one should not be surprised by the periodicals' limited coverage of the French colonial empire: after all, the King's business was his own and the public had no right to an explanation of his policies. What is surprising, however, is the relatively large presence of the foreign Atlantic and the predominance of news from abroad even in coverage of French colonies. That news emanating from London would be readily available to the editors of French periodicals is obvious given the vibrancy of the London market and the multiple entanglements between France and Great Britain. Indeed, it is likely that the London market would have provided the lion's share of the content at any continental European editor's disposal regarding any topic related to foreign affairs. However, that such news would be allowed into the pages of periodicals so closely watched by the French authorities that the Gazette would be formally integrated into the government apparatus in 1762, and that they would be printed far more often than home-grown alternatives discussing the same topics, including the government's own propaganda, seems like the result of a deliberate messaging strategy. A similar argument can be made about reports of rich mineral finds in Spanish America, about the menacing climate and poor companionship to be found in French colonies, and to a lesser extent about the prevalence of non-French America in the Mercure's word games. What was the reader supposed to make of such asymmetric censorship? Maybe that America belonged to others, or at least that others paid it far more attention than his own country did. That continental New France, in particular, accounted for very little in the kingdom's grand strategy, especially in peacetime. That whomever ventured to the Caribbean had better pray for a lasting peace with Britain, in order to avoid the depredations of privateers and to receive help from the metropole in case of a natural disaster. That colonial ventures, in short, remained risky businesses.
Put another way, and with apologies to Fernand Braudel,72 it is possible to sketch a portrait in three layers, associated with three time scales, of the imagined America inspired by a reading of mid-eighteenth century periodicals. A base layer, painted by culture and common knowledge, shows an unchanging, lasting mental image of an alien land filled with riches and glory for the few, mortal threats for the many, and the best, perhaps, set aside for others. A second layer, made up of the periodicals' coverage of the slow production of knowledge through science and exploration, edulcorates this picture to some extent by showing that the New World is in the process of being domesticated, but that this process is very much still in its infancy. Buyer beware. Finally, the top layer, represented by the Gazette's news coverage, shows a French colonial world that is dominated by Britain, virtually invisible in peacetime, and fraught with chaos at every moment. This top layer is especially important since it was the only one visible to the majority of readers, as the Gazette reached an audience perhaps ten times larger than the Mercure and the Journal des Savants and was likely to be the only periodical that made its way into the hands of humble individuals considering indenture contracts.
Therefore, it seems that the overall picture of French America painted by the periodicals would have had little appeal for the reading public, except maybe for its boldest or most desperate segments. This is unlikely to have played a major role in mitigating enthusiasm for emigration to the colonies, as only a small fraction of the population would have had regular access to these periodicals. However, inasmuch as print would have influenced attitudes, it seems that it would have done so in the negative way proposed by Yves Landry's original hypothesis, at least during the last few decades before the Seven Years' War.
François Dominic Laramée recently defended his doctoral thesis in History at the Université de Montréal and will soon join the University of Ottawa as a Postdoctoral Scholar in the History of Medicine. His work has been published in Digital Studies/ Le champ numérique, Document numérique, the Revue d'histoire de l'Amérique française, and the Programming Historian. He has received the Fonds de recherche du Québec -Société et culture's Paul-Gérin-Lajoie outstanding emerging scholar award and has won numerous scholarships and fellowships. Before starting his PhD in his mid-forties, he occupied various creative positions in the video game, television, publishing, and interactive media industries.