to the school children of Missouri and Illinois, these are for all practical purposes, lost .... trade of house construction throughout the colonial period and well into the .... Broutin, and the letter of M. le Maire, S.J., from Pensacola, describing the ...... built in 1765 in St. Louis, and sketched in a reconstruction by F. L. Billon (Fig.
UPPER LOUISIANA’S FRENCH VERNACULAR ARCHITECTURE IN THE GREATER ATLANTIC WORLD Jay D. Edwards Dept. Geography & Anthropology Louisiana State University
2
About the Author Jay Edwards is Professor of Anthropology at Louisiana State University, and director of the Fred B. Kniffen Cultural Resources Lab. He specializes in the vernacular architecture and material culture of the Gulf South, the Mississippi Valley, and the greater Atlantic world. He is co-author and author of books and articles in his field.
3 Introduction Among that minority of Americans who have specific knowledge of the historic French settlements in the Upper Mississippi River valley, there is an almost universally shared perception that these tiny communities were isolates – cut off from all of the major events of American history. It is generally believed that these communities were populated by a few rough-hewn Canadian voyagers and coureurs-de-bois who contributed little beyond trading with the local Indians. In the popular imagination, the villages of eighteenth century Missouri and Illinois were frontier communities which were quickly absorbed into the vast tide of Anglo Manifest Destiny flowing inexorably westward. By the first third of the nineteenth century their cultures had been totally subsumed. Historians who have studied the area know better, but their deeper understanding of the vital role played by the people of the early French communities has never gotten much traction among the writers of the American histories and textbooks on American history [See, for example, the works of: John Francis McDermott, Henry Marie Brackenridge, Natalie Maree Belting, Margaret Kimball Brown, Charles E. Peterson, Charles van Ravensway, Louis Houck, Marc de Villiers]. Even to the school children of Missouri and Illinois, these are for all practical purposes, lost communities of the Middle Mississippi River Valley.
Initial Questions A question we wish to raise is what the historic French vernacular architecture of the pays des Illinois (Illinois Country) might have to teach us about the supposed problem of cultural isolation.1 Were these frontier communities really cultural cul-de-sacs, or did they have some greater and more cosmopolitan role to play in the history of Middle America? What part did they play in the wider development of American cultural traditions? A related consideration is the question is what can be learned from a close examination of the built environment which might not otherwise be obvious from a reading of the written history of the area.
4 The common perception of the “crudely built” frontier houses of the French settlers is that they were seventeenth century French Canadian derivatives in form, and that they were neither very numerous nor very long-lived. Nearly all had been replaced by better-built Anglo style brick and frame structures before the end of the nineteenth century. But if it could be demonstrated that this perception is erroneous, and that the influence of the vernacular architecture of Upper Louisiana extended far beyond the communities of early St. Louis, Cahokia, Kaskaskia, and Ste. Genevieve, then a strong argument might be made that their collective history ought to be more widely known and appreciated. In the latter case, what survives of these significant cultural assets would be deserving of first class recognition, protection and support from national and even international organizations. If the French settlements in Upper Louisiana are to function as a cornerstone of American frontier history, they should be deemed worthy of what the British term a “conservation area” – “areas of special architectural or historic interest, the character or appearance of which it is desirable to preserve or enhance.”2 The communities in question now reside in what is being called the “Creole Colonial District of the Mid-Mississippi River Valley.”3 As we explore the question of the importance of these historic communities and their structures, we may well wish to revisit the concepts of “Creole,” and creolization. They are increasingly the subjects of academic theorizing and debate, and they raise the potential of permitting us to view the heritage assets of the Illinois Country from a new and more elevated perspective.4 Why is the Illinois Country significant in the grand scheme of North American Culture History? There are several good answers to this question. One is that for about a century this area lay at the end of two great colonial era diffusion routes. Significantly different configurations of French culture, geographically separated over thousands of miles, flowed together here in this small region to establish a unique and fascinating cultural mélange. One diffusional limb arose directly out of late medieval western France. It crossed the north Atlantic and, for about a century,
5 perched tenaciously in the Maritime Provinces and along the banks of the St. Lawrence River. As the French readapted their culture to the harsh climates of the Canadian country, they gradually lurched westward, exploring the pays d’en haut – the Upper Country of the Great Lakes and beyond. Beginning in 1699 they began to establish permanent settlements along the Mississippi River. As Biloxi -- the first French settlement on the Mississippi Gulf Coast -- was being established by the Canadian brothers, Iberville and Bienville, Cahokia, Ill. was simultaneously founded by other Canadians coming from the north. Many other settlements were soon to follow on the banks of the Mississippi River. It did not take long for the upper Mississippi and Gulf Coast communities, about twelve hundred miles apart as the pirogue paddles, to begin the process of regular communication with one-another. Beginning in 1708, a decade before the founding of New Orleans, large amounts of wheat flour were being transported down the Mississippi and across Lake Pontchartrain to the Gulf Coast communities of Biloxi, Dauphine Island and Mobile. Lower Louisiana quickly became the principal trading partner for settlers in the Illinois Country.5 The downstream trip in the spring required about forty-five days (Vidal claims 15 -20), while the return trip required three to four months of hard rowing and poling. Yet, despite the constant threat of Indian attack, annually, over a hundred people made these return trips annually in the early colonial period.6 Enslaved Africans were employed for the difficult work of the return trip, bringing their skills and handicrafts with them into the Illinois country. The boats they muscled upstream brought many fine luxuries such as fabrics and household utensils to the settlers of the northern communities. They also brought nails, window glass and iron hardware for new houses. These would not otherwise have been enjoyed by the traders, farmers and lead miners of these remote communities without the arduous annual river voyages. Long before the Louisiana Purchase, the household inventories of certain wealthy northern families read like those of Philadelphia or Boston.
6 Margaret Brown and others have provided accounts of the travels of Illinois Country settlers in the years before they became permanent settlers. Many African slaves had, of course, come from as far as French West Africa (called, variously “Senegal” and “Guinea”), with stops in the West Indies, before arriving in Louisiana. The census of 1723-1725 combined with church records reveals that of a total of approximately six hundred individuals living in the Illinois country between Cahokia and Kaskaskia, nearly one-quarter were African slaves in 1726, and almost one third by 1752.7 Though local Indians were enslaved, most of bonded laborers were Africans who had come up the Mississippi River, working as rowers on the freight bateaux, and as farm hands following their arrival. African slaves were imported directly from Africa beginning in 1719, and were heavily used in river transport until 1731, when they became more valuable to those in Lower Louisiana.8 These population percentages reflect a very high proportion of Africans at the beginning of the settlement period. Since their labor was probably essential in the erection of many of the houses, their influence could not help but have been significant. Yet we know almost nothing about it. Several observations are relevant here. We do not have good records of the ethnic mix in the typical construction process. We know that by tradition, peasant Frenchmen and Canadians often raised their houses using the coup d’main process -- a house raising bee in which friends and relatives participated. In later years, a surprisingly large number of French craftsmen were listed in Kaskaskia and the neighboring villages, the most populous category being carpenters and cabinetmakers.9 People in the villages of Upper Louisiana depended on specialists for the erection of their houses. “Roofers,” for example, were also specialists in the building of the intricate Norman roof trusses. In Lower Louisiana, many Africans and Afro-Creoles had been specifically trained by French journeymen to become builders in the French tradition. It was they who dominated the trade of house construction throughout the colonial period and well into the nineteenth century. Were any of these specialists being sent to the north, or were their skills in too great demand in
7 Lower Louisiana? The question of exactly how these various competing approaches to the construction of houses worked out in the Illinois Country will need to be left to future research. The main point remains. In order to achieve a better understanding of the vernacular architecture of the Illinois Country, we must explore the confluence of these two great streams of culture through a comparative perspective. A better assessment of culture process must result. So let’s talk cultural process. It is the foundation upon which our evaluation of the significance of Illinois Country vernacular architecture rests. Some questions should be raised at the outset: 1) What survived the lengthy diffusion process, and why? 2) What didn’t survive and why? 3) How did these disparate traditions blend together, or did they? and, 4) What remained to be invented on the frontier due to the inadequacies of the traditional solutions of both traditions? This leads directly to the larger question of the degree to which the eighteenth century vernacular architecture of the Pays des Illinois may be considered to be domestic French, or the degree to which it was something else?” If, indeed, it was something else, exactly what was it?” Regional historians, tourist commissions, visitor’s centers, and the owners of house museums, might all discover useful information in definitive answers to such questions. But these questions set before us problems which extend far beyond the significance of the architectural adaptations which occurred in this specific region. They call upon us to seek out and explain the general processes of architectural adaptation in frontier and colonial environments. They impel us to cast our net broadly, to consider all of the other environments in which French pioneers formulated architectural traditions in the Atlantic World. In addition, they ask of us theoretical and conceptual formulations of clarity and depth. This amounts to an obligation which we cannot take lightly. Done well, it requires lifetimes of research and the thoughtful reflection of multiple specialists.
Foundational Studies
8 In the past, two great scholars wrestled with these questions. Both spent a significant portion of their active careers in the study of the traditional architecture of portions of greater colonial Louisiana. There is a certain irony in the results of their research. Charles E. Peterson, F.A.I.A., is most widely known for his 1933 proposal which lead to the creation of the Historic American Buildings Survey. It was the only one of a dozen or more WPA projects initiated under the Roosevelt administration to survive actively to the present, making it the most successful of all of those projects. It created the single most important archive of early American buildings in existence. But Peterson was also an architectural historian of unusual capacity. In the late 1830s through the 1950s he published a series of articles and a book on the French colonial architecture of Ste. Genevieve and St. Louis.10 It is clear from a reading of these materials that Peterson was intimately familiar with both the eighteenth century vernacular architecture of France, and that of French Canada. He adopted a broad comparative perspective in his descriptions of the origins of the traditions of the Illinois Country. He was familiar with the French builder’s vocabulary, and he understood their construction technologies. Most importantly, perhaps, his perspective extended beyond the bounds of France and Canada, into the West Indies. He traveled to all of these places, including islands in the French West Indies, “finding… architectural features which relate in some way to early St. Louis.”11 Although Peterson used the term “A Creole Capital” in the title of his book on French St. Louis, he never defines the term “Creole.” He is apparently using it as it is commonly employed in the Illinois Country to refer to a native of the country, but one of (French culture and) foreign parentage. However, it is clear from a number of passages that Peterson well understood the breadth of the process of amalgamation which we now refer to as creolization in architecture. “There is not a single structure standing which is purely French”12 He argued that poteaux en terre wall construction, the raising of houses high on a “basement” “in the style of Louisiana and the New World tropics,” and the idea of wrapping a wide gallery around one’s house, all derived from French
9 Louisiana and ultimately the West Indies. He correctly points to the north coast of Haiti as a particularly apt example of tropical vernacular architectural forms (Fig. __).13 By way of contrast, the architectural historian, Samual Wilson (also F.A.I.A.), conducted a wide range of architectural history studies in and around the city of New Orleans. Wilson began working for the Louisiana HABS in 1934, under architect Richard Koch, who would later become his life-long business partner. In essence, Charles Peterson was their boss – the CEO of the Historic American Buildings Survey. In 1938 Wilson received a grant from the American Institute of Architects to conduct architectural research on New Orleans in the archives of Paris. He returned with an enormous treasure trove of materials which formed the basis of his numerous articles published throughout the remainder of his life. Wilson was interested primarily in the history of individual buildings and in the accomplishments of specific French architects and engineers -- men such as Adrien de Pauger, Ignace François Broutin and Alexandre de Batz. His attitude towards the creolization process in New Orleans architecture was pretty much to ignore it. Wilson apparently felt that French colonial architecture was best explained as a series of logical adaptations of domestic French forms made to the heat and humidity of the tropics. On the topic of the broad gallery, so prominent in Louisiana domestic architecture, he appears conflicted. In a lecture given in his office for LSU students of vernacular architecture, Wilson stated that “French engineers were not drawing galleries on buildings prior to about 1750” (June 5, 1980). However, his own research turned up numerous examples of early galleries in engineering drawings. For example, I. L Callot and Alexandre de Batz were designing buildings galleried in the traditional Norman manner as guard houses and prisons for the Balise Post in 1731 and 1734, respectively.14 By1761, Bernard Deverges had begun to design guardhouses for “Fort Tombecbé” in the manner of Class II Creole houses (double pitched roofs).15 Even earlier galleries are apparent in his writings. The sketch of the de la Pointe concession on the Pascagoula river ca. 1726 which Dumont dit Montigny included on his “Chart of the Pascagoula River,” shows a raised plantation
10 house in the style of northern Saint Domingue, with a “balcon tout au tour” (see Figs. __).16 There was other evidence available to Wilson which indicated even earlier galleries were common along the Mississippi Gulf Coast, for example: the 1731 – 1734 plans of New Orleans by Gonichon and Broutin, and the letter of M. le Maire, S.J., from Pensacola, describing the houses of Dauphine Island ca. 1712 : “Most have a gallery all around…” Although Wilson was surrounded by examples of creolization in the architecture of the French settlers of New Orleans, he never referred to the concept, nor did he indicate that the West Indies played any significant role in the architectural traditions of Louisiana. For example, Claude Joseph Dubreuil de Villars, a native of Dijon, France, acquired a plantation in what is now Faubourg Marigny in the 1740s. Dubreuil was both a prominent builder and the first planter to successfully plant and crystallize sugar in Louisiana. He imported plants from the Jesuits and directly from Saint-Domingue. His plantation house must also have appeared much like those of the plaine du nord in that colony. It was described as “one hundred feet long by forty wide… with galleries on three sides.17 Yet the possibility that Dubreuil was influenced by the architectural symbols of sugar production on the north coast of Saint-Domingue is never considered by Wilson. Indeed, he eschewed the term “Creole” except to refer to several buildings “which have come to be known as ‘French Creole’ cottage[s], having a gallery across the front over which the slope of the gable-ended roof extends.…” 18
Approaches to Architectural History The work of these two great architectural historians illustrates that fundamentally different conceptions of architectural history in the Mississippi Valley may result from even the most experienced and knowledgeable of scholars. Perhaps, if we are to accurately assess the processes of culture formation which created these traditions, we will need to sharpen our understanding of the specifics of French vernacular architecture, and to broaden our knowledge of France’s outre-mer
11 colonial traditions. A appropriate diffusion study would require familiarization with those places settled by pioneers who eventually arrived in the Mississippi Valley in the seventeenth and eighteenth centuries by many diverse routes. We must not only familiarize ourselves with the rural cultural landscapes of the regions of western France, but also those of Acadie and Maritime Canada, Quebec and its hinterlands such as the Isle de Orleans, and the pays d’en haut or the western upper Canada. We must also consider the southern settlers -- those who came from the Caribbean and Lower Louisiana, and those who had arrived from further afield, particularly from the coasts of Upper Guinea between the Senegal River Valley and what is today Liberia and Sierra Leone. We begin by drawing a basic distinction between two significant forms of colonial architecture: French Colonial Architecture: All buildings constructed by the French, their acculturated subalterns, and their contractors in overseas colonies, in the traditions of France. These buildings would not be considered out of place if set in a domestic French context.
French Creole Architecture: Buildings built by or for Europeans in overseas colonies which were systematically changed from their European forms to better fit local conditions. Creolized forms are adapted to new environments. They become established traditions in their own right.
Creolized buildings may be clearly distinguished from French Colonial buildings on geometrical grounds. Though Creole houses and barns may include perfect French components, these are combined with elements which would be both out of place and misunderstood in the countryside of France.
12 With this basic distinction we may begin to explore our central question: To what extent was the eighteenth century vernacular house of the Illinois Country “French,” and to what extent was it something else completely? Which methodology should we adopt in order to attack this problem? Let’s begin with the proposition that every architectural tradition is composed of many distinct patterns. They are shared adaptive responses to the problems faced by the designerbuilder. That very practical and experienced specialist carries with him years of accumulated knowledge -- a virtual compendium of techniques and solutions for the problems which he will face in meeting the demands of a new homeowner (which may well be himself). Within his repertoire are alternate possibilities, and a conception of the advantages and disadvantages of each. Every solution has its own independent history. An individual historical background may be quite unrelated to those of alternate solutions. One way of envisioning the traditional house is to see it as a selection of solutions drawn from a much wider set of possibilities. Our job, then, is to get into the head of the designer-builder and to attempt to recreate the decision making process and the sources of his selections. We want to conceptualize the building at the same level as he (or they) did on the first morning of the job. This is a lofty goal which will never be fully achieved at such a profound temporal distance, but we can, perhaps, approximate it. Through fieldwork sampling and archival descriptions we establish a list of the range of solutions practiced in our culture area and our historical period. Some will be relatively popular, others rarely used. Each single solution is viewed as an integral component of an overall composition. One preliminary assumption is that each solution had functional interrelationships with other components and patterns of the house. In this assumption we will occasionally be wrong, but as a working hypothesis it is essential that we test it before discarding it.19 Ultimately, we may be able to comprehend the relative advantages and disadvantages of each alternate technique and pattern. For example, we know that close-studding was used for the most part in the vernacular houses of Upper Louisiana, even into the third decade
13 of the nineteenth century, while open-studding was used for walls in Lower Louisiana in the same time period. Yet, their buildings were constructed, for the most part, by the same kinds of people, including both Canadians and enslaved Africans. What accounts for the difference? Once we have developed a sense of the relationships between specific builders and the patterns of their creation, we should uncover answers to questions such as these.
An Historical Interpretation of Selected Components of Illinois Country Vernacular Architecture Because even simple architectural forms are complex and multi-layered, we have a wide variety of possible patterns to select from. These can be broken down into three basic classes: Geometrical Patterns, Decorative Features, and Construction Technology. For our immediate purposes I draw upon examples from traditional geometry and from techniques of construction.
1: British vs. French Timber Framing Traditions: Our first consideration is that in both Upper and Lower Louisiana, French settlers were in contact with British Americans from even prior to the end of the Seven Year’s War. Certainly, shortly after 1765 their influence was being felt in the French communities. For example, following the great fire of 1788, Manuel de Lanzos, a Spanish military officer, hired a British carpenter Robert Jones to reconstruct his heavily damaged house in New Orleans. The original had been built by Jean Pascal, a French immigrant, in the 1720s. It was rebuilt mostly on the foundations of the original, so the new house (oldest documented standing house in New Orleans) preserves some of its earlier geometry. Jones undoubtedly hired local French Creole and Afro-Creole brick masons, carpenters and roofers to work on the reconstruction of the house. The unusual construction features of the house – known as “Madame John’s Legacy in New Orleans – cannot help but be artifacts of the creolization process. 20
14 There is a basic philosophical difference between the French method of timber framing and that of the British. Both French and English timber frame houses began as one-room deep buildings in the post-medieval period. The British thought of the timber framed building in terms of vertical frame assemblies (bents), separated by social and functional “bay” spaces of different lengths. When the English began adding a second range of rear rooms, producing “double-pile” houses in the late sixteenth century, they typically roofed the rear range with an entirely separate set of roof trusses, producing curious “M” shaped roofs. This configuration is characteristic of British Jacobean houses and seventeenth century great houses in the British Caribbean, such as Drax Hall and St. Nicolas Abbey in Barbados. This double-pile roof system has continued down to the present in vernacular town houses and store houses of other British West Indian islands, and in the chattel houses of Barbados. When the French began expanding the depth of their timber frame houses, they typically widened them with more narrow “semi-pile” additions. To cover these one-and-one-half room deep houses, they employed a number of solutions. If the house was built from scratch with that plan, they simply covered it with a symmetrical “pavilion” roof, adjusted to the requisite size. If, as was often the case, the second range of narrow rooms was added after the house was originally constructed, they placed a new set of rafters from the ridge, or from a principal purlin, across the new rear wall of the house, producing a roof with an asymmetrical profile. In the former case, the rear roof was single pitch, but in the latter, it had a somewhat lower pitch towards the bottom. Historically, this double pitch solution was a carry-over of their earlier adaptation of through-purlin roof framing to structures with parapet walls.21 Through-purlins run continuously from one end of the roof to the other. They lie at a convenient position for attachment of a second set of lowerpitched outer rafters when the roof is widened. The purlins are, in turn, supported from below by even heavier “truss blades” (arbeletrièrs), which run parallel to the principal rafters (refer to Fig. __).
15 As soon as broad galleries began to be added to Creolized colonial houses in Lower Louisiana and the Mississippi River Valley in the early eighteenth century, this old French framing method for building expansion was resurrected. Almost all galleried houses had steep inner pitches and lower pitches over the galleries (Figs. __, __). In addition to all of this, the French thought of their timber frame buildings as being composed of a series of levels, each with its own independent inter-spatial geometry. There were no timber bents, but only principal cross-building assemblies at each of the different levels of the building. This encouraged French builders to provide each separate level of the building with its own independent spatial organization (Figs. __, __). The Old St. Gabriel church in Iberville Parish, Louisiana, provides an excellent example [Fig. __]. Its roof truss system is entirely separated from its rafter system and from the organizational structure of the walls which support them. The church was built in 1774-1775 for the newly arrived Cajuns. The frame of the church was probably designed by a French engineer, while the details of construction were left to the Cajun workers. The roof of the church is supported by six great trusses, stiffened by collar beams and mounted on arched tie beams which supported an “arch of heaven” ceiling, painted powder blue.22 One of the inter-truss bays is shown in Fig. __, with architectural historian, Sid Gray seated on the sousfaitage (under-ridge) beam. Examine the lightweight rafters behind Sid, and you will see that their spacing is entirely separated from that of the heavy truss system. In its original form, this church had four fenestration bays (window bays) in each of its long side walls. The window bays were distributed between thirteen framing bays (the spaces between the heavy vertical posts) [Fig. __]. Yet the roof truss system directly above them had only five bays between its six heavy trusses. Clearly, the designer was entitled to complete freedom of organization in the framing of each level of the structure, and each one was relatively independent of the others. This sort of organization is referred to as French platform construction. It is characteristic of both domestic French and Colonial French timber frame buildings. Unlike
16 anything found in France, though, the Old St. Gabriel church had surrounding galleries. The original galleries were removed at a later time when the church was moved back from the edge of the Mississippi River in order to protect it with levees. Something of the influence of English timber framing may be gauged by an examination of the framing system of the Pierre Menard house. This was one of the last large French style houses to be constructed in the Illinois Country. The house was constructed about 1814 according to tree ring dating.23 This date is confirmed, approximately, by nail dating.24 At the time the Pierre Menard house was constructed, the area had been under English and American control for about half a century. It would not be unreasonable to assume that British style would have influenced this house profoundly, but for the most part, the construction of the house is entirely within the French tradition. Some principal exceptions to this generalization are: 1) the use of sliding sash 6/6 windows, 2) the layout of the first floor of the house with a fourteen foot wide hallway-like room between two larger rooms, 3) the use of Georgian stylistic elements, and 4) the roof support system, which appears to be an Anglo style roof truss with principal rafters lapped at the peak and stiffened by collar beams, but with no roof ridge, truss blades or king posts. Thus, with the exception of details, the Pierre Menard house is in full conformity with the traditions of domestic French timber framing, as are all earlier houses which we have examined on both sides of the river.
2. Western French Floorplan Geometry and its Legacy: There is a tradition of room layout which is widely shared by late medieval and post-medieval vernacular farm houses both in Normandy, north of the Loire River. It is also found throughout the centre-ouest country south of the Loire, where most of the Acadians originated. Houses share a two-room “core module,” composed of a large central salle or “fire hall” (all-purpose living room), and a more narrow chamber or master bedroom (Fig. __). The entire module is generally about 16 – 20 feet deep and the salle is near-square, with the chamber being more than half the width of the sale and the same
17 depth. Additional rooms may be added to this base module, and generally are, but because it may stand alone, we must consider the sale-et-chambre to be the fundamental historic house plan of this area. Between the beginning of the seventeenth and the end of the eighteenth centuries it was the most popular basic plan among the vernacular houses of northwestern France.25 The salle-et-chambre plan was carried into America by the Norman settlers of the St. Lawrence River Valley. It was also transported by Creoles into southern Louisiana where it remained the most common plan for eighteenth century vernacular houses (Figs. __, __). Many of the oldest surviving country houses in Louisiana have this plan. It was adopted by those Acadians who came to Louisiana beginning in 1765, though many built single room cabins at the beginning. Even large plantation houses, such as Parlange Plantation house in Point Coupée Parish, have a salle-etchambre at their core. In the Illinois Country the story is a bit more complex. Many surviving late eighteenth century French houses have an alternate, two-cell or double-salle plan (Fig. __). Even expanded houses are centered around two near-square rooms, but there may be a problem in the degree to which these late colonial and early American period houses reflect the average settler’s house of the eighteenth century. Most of the surviving houses in Ste. Genevieve are those of somewhat more economically successful traders and businessmen. In addition, their houses were moved or rebuilt at the new Ste. Genevieve village after the flood of 1782.26 Earlier houses, and those of comparatively impoverished habitants on both sides of the river may well reflect a different geometric sensibility. In their invaluable survey of the early French houses of the original Village of Chartres, IL, Margaret Brown and Lawrie Dean recorded descriptions of numerous early settler’s cabins. Of nineteen houses in which dimensions were recorded between 1725 and 1760, the mean length and width of the group was 18.3 feet x 26.2 feet (L/W ratio: 1.43). A poteau en terre house dated ca. 1752-1765 and excavated at the New Chartres village measured 19.7 by 27.9 feet (L/W ratio: 1.42).27 These dimensions are incompatible with a double-salle (two near-square room) plan,
18 and also with a single room plan. Rather they would indicate a basic average plan of a near square salle about 18.3 feet wide, coupled with a narrow chamber about 8.1 feet in width. In reality, however, the average salle was probably about a foot more narrow and the average chamber width was something closer to nine or ten feet in width. No recorded houses in Chartres in this period possessed two near-square rooms. They appear to be a late development, probably dating to the second village of Ste. Genevieve. Surviving Illinois Country dwellings with classic sale-et-chambre plans include: the MartinBoismenue house in Prairie du Pont, IL , ca. 1790 (Figs. __, ___), and in Ste. Genevieve: The LasourceDurand house ca. 1818 (HABS MO-1281), the Boisleduc house ca. 1770; the Joseph Seraphin house (HABS MO-1282 – date uncertain); and the Rozier house (HABS MO-1280 – date uncertain; Fig. __).28 All of these asymmetrical plan buildings were constructed with the traditional close-studded walls. This leads to the conclusion that even in the last decades of the eighteenth century no less than a substantial minority, and probably a majority of the Illinois Country French houses conformed to age-old French asymmetrical geometrical preferences. It also means that in the late eighteenth century, important changes were transforming local French traditions for those residents who could afford to symbolize their newer identities and allegiances.
3. Tying the House Together – Parapet Walls and the Tie Beam-Wall Plate Connection: Of all the houses in France, those of Normandy had the steepest pitches, on average. Pitches of between forty-five and seventy degrees were not unknown, and end sheds on pavilion (hipped) roofs were often as steep as eighty degrees (Fig. __). In addition, Norman roofs were the most strongly supported. This resulted from a combination of environmental and engineering pressures. Normandy, which abuts the English Channel, receives considerable moisture in the winter and spring. It can drizzle for days at a time (Fig. __). In post-medieval times, farm house roofs were mostly thatched. Extended rainfalls will penetrate a thatched roof, so special measures must be
19 taken to prevent leaking. French builders handled the problem in two ways. You can thicken the bed of thatch, and you can steepen the pitch at which it lies. 29 A steeper pitch multiplies the area of the roof for the same house, and this places additional stresses upon the house itself, particularly at the position of the wall plates where spreading pressures are maximized. A steeply pitched roof with a two-foot thick thatch covering, soaked with rainwater, is trying to tear the house apart at the tops of its walls. As a result, Norman roof support systems were over-engineered. Exceptionally sturdy roof trusses were developed over the centuries to prevent the many tons of pressure from the roof-covering from sagging and splitting the supporting members and forcing the walls apart (Fig. __). Steeply pitched roofs offered benefits as well as liabilities to their inhabitants. A loft of considerable size was available even in relatively modest houses of only around six meters in depth (18-19 feet). This space was used, not simply as a traditional grenier or grain storage area, but also as work space. Dormers were added to provide natural light in the loft. In the sixteenth and seventeenth centuries, looms were set in the lofts for weaving, resulting in extra income. Other work was done there too, particularly in the winter time. However, because of the unusual spreading forces placed upon these roofs, the trusses which supported them were invariably spanned by collar beams set at about chin height. In order to permit full headroom under the collars, a system of parapet walls was developed. A short extra wall was added to the top of the wall plates – usually about one-half to two-thirds of a meter in height, but sometimes taller. A second wall plate was placed atop this parapet wall and the roof rafters (chevrons) were notched and lapped over this upper plate. The rafters continued below the plate, allowed the roof to expand out past the outer wall surfaces by as much as two feet or more, providing shade in summer and protection from rainwater in winter (Fig. __). Neither the idea of the steeply pitched roof, nor the concept of the parapet were forgotten by Illinois Country builders. When galleries were added to the Canadian style cabins, a secondary
20 set of light outer rafters was added to their roofs. In the Martin-Boismenue house in Prairie du Pont, and the Bequette-Ribault house in Ste. Genevieve, these gallery rafters were footed in a gallery plate supported by a peristyle of tall wooden colonnettes. The notch in the rafter which rests on the plate is referred to as a “bird’s mouth” in Louisiana. The upper ends of the gallery rafters were inserted over the principal purlins about half-way up the roof (Fig. __).30 The result was a roof with a sharply broken pitch in the middle, giving the upper Illinois Country house a distinctive “witch’s cap” appearance (Fig. __). Outer gallery rafters, unlike anything in France, were long and low-pitched and required support in their middles to prevent sagging. That support was supplied by a row of short posts mounted on directly atop the wall plates, in the same location as the posts of parapet walls had been added to the wall tops of steep-roofed Norman houses (Fig. ___). One surprising difference existed. In some of the Illinois Country houses at least, the wall plate mounted posts were not sheathed with an outer layer of clapboards to prevent air from flowing into the attic; rather, the loft space was left entirely open to the outside, though well protected from precipitation by the broad gallery roofs. Open lofts meant that everything above the ceiling of the house was open to the wind – a clear advantage in the heat of the summer but, perhaps, a considerable disadvantage in on cold winter nights! The parapet wall framing tradition appears to have been of considerable use in another Illinois Country case. Because there were so few stores, warehouses or anything else where boatloads of merchandise might be safely stored in Ste. Genevieve, merchants used the lofts of their houses for storage space. The Guibourd-Valle house provides an excellent example of this principal. So important was access to the large loft of this trader’s-style house that a flight of winder stairs to the attic was subsequently incorporated into one end of the back gallery, replacing an earlier ladder to the loft space.
21 Jacques Guibourd, from Angers, France, moved to the colony of Saint Domingue, where he worked as secretary to a wealthy planter. Legend has it that during the Haitian Revolution he was spirited out of the war-torn colony by a faithful servant, Moros. Guibourd arrived in Ste. Genevieve about 1799. The house is reported to have been constructed in 1806-07.31 It has been speculated that because of his earlier West Indian experience, it may have been Guibourd who introduced into Ste. Genevieve the idea of placing broad galleries on the front and back of a house, or surrounding it. Guibourd, though, did not arrive in Ste. Genevieve until the last years of the eighteenth century, while inventories from the Illinois Country reveal that as early as 1732 French houses were being constructed with galleries on the two long sides, and as early as 1758 houses were being constructed with surrounding galleries.32 In framing the loft of the Guibourd-Valle house, the roofer faced certain problems, well appreciated by the builders of roof trusses in Norman houses with parapet lofts. The traditional French roof is what R.A. Cordingly referred to as a through-purlin roof. It employs sets of inner and outer rafters separated by a space of only a few inches. The heavy inner rafters are referred to as “truss-blades” in English, and arbaletriers in French. They provide primary structural support for the entire roof. They generally spring from the very outer ends of a horizontal tie beams (which hold the walls of the house together), rather from the wall plates (which are reserved for the outer rafters (Fig. __). In parapeted lofts, however, the tie beam is set one to two feet below the wall plate. If one inserts the truss blades into the lowered tie beam, one has exceeded the distance required for the truss blade to perform its principal function. That is to transfer the weight of the roof from the rafters directly through the through-purlins into the truss blades, and hence to the tops of the walls and vertically to the ground. With the rafters set onto the tops of the parapet walls, the truss blades are separated too far to permit the purlins to be in simultaneous contact with both sets of rafterlike members. Thus, the vital support function of the truss blade is destroyed.
22 The Norman French, with their centuries of experience and experimentation, worked out a number of solutions to this problem. Here, I will mention only one. At some point in its development, a uniquely Gallic solution called the jamb-de-force, or “leg of force,” was integrated into the Norman roof truss. This separate inner strut for the support of the collar beam was added directly below and roughly parallel to the truss blade. Now, between the tie beam on the loft floor, and the collar beam at loft ceiling height there existed a system of no less than three structural members – rafters, truss blades, and jambs-de-force -- all more or less parallel, one above the other (Fig. __). Typically in France, the jamb-de-force was somewhat concave, providing for a wider usable loft space. It sprang from the exterior ends of the tie beam and was inserted into the outer ends of the collar beam, or it was mortised into the truss blade just below the level of the collar beam. This shifted the foot of the truss blade to a horizontal strut (blochet) at the level of the wall plate (Fig. __). In the Guibourd-Valle house with its generous loft, there is no jamb-de-force. Both the rafters and the truss blades are footed on the same narrow member – the elevated wall plate atop the parapet wall. The parapet wall rises about twenty inches above the floor of the loft. This provides six feet three inches of headroom between the floor and the collar beam, making the loft into usable social space, but it also introduces some awkward elements into the design of the roof framing system. The collar beam is placed too high to provide maximum structural strength for the truss. Ideally, it would be halfway between the loft floor and the peak of the roof in order to provide stiffening for the rafters and the roof surface. But there is an even more serious engineering problem, one shared with other Ste. Genevieve houses with so-called “Norman” roof trusses. The rafter and the truss blade should run parallel to one another about three to four inches apart, providing space for the horizontal principal through-purlin to be tightly sandwiched between them. A parapet loft, however, leaves no four-inch space for purlin sandwiching – the beams are too far apart. The local carpenter’s solution was to angle the two members so that they were both footed together on the same narrow wall plate, but so that they were spread over a foot apart at their tops.
23 The rafters are half-lapped over the roof ridge, while the truss blades are mortised into the sides of the king posts near their tops. While answering one problem, this solution presented the carpenter with another. No principal rafter could be placed directly over a truss blade because they both could not be footed at the same place on the narrow elevated wall plate. The roofer’s answer was to separate the truss blades and rafters horizontally at their feet (Fig. __). The elevation of the principal purlin within the roof had to be adjusted vertically to the gradually diverging space between then (no longer parallel) rafters and the truss blades. When the purlin was properly located and sandwiched tightly between the two members, the force of the entire weight of the roof could ideally be transferred directly from the rafters to the truss blades through the principal purlins, but, with the two members separated horizontally, the line of force from the roof burden and the rafters had to turn ninety degrees and run through the horizontal purlin before turning back, once again, to follow the truss blade down to the walls. The further the truss blade was separated from the rafter, the greater the tendency to split and fracture the principal purlin, which now bore all of the stress purely on the strength of its wooden material. [The angle of stress was 90° to the long dimension of the purlin – an illegitimate burden!] What is amazing is that these roofs have stood for over two centuries without noticeable splitting of their purlins. This miracle is attributable, both to the intelligence of the roofer who made sure that the truss blades and rafters were set close together, and also to the fact that no heavy roof covering such as thatch or slate multiplied the weight of the roof. Nevertheless, in forgetting the function of the jamb-de-force, the culture loss from Norman framing technology has rendered the Ste. Genevieve roof an imperfect solution to lofts with parapets. An assessment of this case leads to several conclusions: A well-rounded understanding of the range of solutions of Norman roof framing technology had been lost in the Illinois Country. We should perhaps no longer refer to these roofs as “Norman.” They are, in fact, relatively degraded examples of Norman technology. But if not Norman, how should they best be characterized? It
24 seems that historians of the French architecture of the area have forgotten the deeper history of these marvelous roof systems. They contain most of the elements of the Vitruvian system of Roman roofing which was developed in the first century before Christ. It was the Romans, or perhaps their Egyptian contemporaries, who developed the first true roof trusses and the through-purlin system of framing. 33 Once popularized and recorded by Vitruvius in the first century C.E., that complex tradition persisted in circum-Mediterranean architecture through the fall of Rome, through the Christian basilicas of the Middle Ages and Romanesque era churches, and through the Gothic churches of southern Europe. It was still in popular use in Renaissance period villas and chateaux of fourteenth through eighteenth century France. From these, it was adopted into the peasant architecture of northwestern France, where it underwent relatively high levels of elaboration based on Germanic and perhaps Celtic and Scandinavian patterns. The common vernacular French parapet roof is one example of such a syncretism. Several of the post-medieval innovations, including the jamb-de-force, were lost, once again, in the process of American settlement, but the Roman core of the through-purlin system survived marvelously intact. Early Roman roofs all contained: heavy hewn timber roof ridges, king posts, truss blades, through-purlins, tie beams, and either raking struts or collar beams – all members found in the French architecture of the Mississippi Valley. Rather than “Norman,” then, I believe we should refer to these roof framing systems as “Roman.” Hidden beneath the roofs of old Illinois and Ste. Genevieve houses stand one of the finest living examples of two thousand years of unbroken Roman vernacular framing tradition. This, alone, would render a European site worthy of special historical attention, but locate such a site in Middle America and it becomes all the more remarkable. The use of carpenter’s scribe marks and Roman numerals to mark the positions of various trusses and beams add yet, another surviving element of Roman tradition secreted within the walls and roofs the same houses.34
25 4. Timber Wall Construction in the Illinois Country – Close Studding: By the mid-seventeenth century, peasant builders in western France had worked out a series of standard techniques for wall construction in farm houses. Some of these were based upon earlier medieval techniques which, by the year 1700, were well advanced into obsolescence. One of these was close-studding, what the French referred to as pan-du-bois. A heavy sill timber (sablière basse, solle) was laid horizontally upon a foundation wall (solage) about a meter tall. Hewn posts were mounted vertically upon the sill beam. The posts usually measured about 0.2 meters deep by 0.2 to 0.4 m in width, and the space between them was approximately 0.1 to 0.2 m. Full height wind braces were set at the corners of the building and the posts there were segmented and mortised into the braces in a technique called houlice assemblage (Fig. __). Each post and wind brace was mortised into a heavy top plate (sablière haute). The joists and tie beams which spanned the house were mortised into or run entirely through this plate. When the French came to New France, their preferences for wall construction seem to have changed with the Canadian climate. True, at first they continued to build in the style of domestic French pan-du-bois with their posts mounted upon a stone solage or stone foundation sill.35 Rather quickly, though, the pioneers moved away from traditional French style houses, first to walls composed of vertical members buried in the ground. These included madriers debout en terre, upright planks in the ground, or pieux debout en terre, upright posts, hewn or partially round. Posts are described with three more or less synonymous terms: poteaux, pieux, and piquets. Unfortunately, the meanings of these three terms are seldom if ever more carefully defined in documents. Depending on the local parlance, notaries, travelers, and other non-carpenters seem to have used them more or less interchangeably. There is some indication that piquets could be round and smaller in diameter than the others, though the term is clearly used in reference to hewn posts in some places. One form of picket construction which was used in the seventeenth and into the first half of the eighteenth century was to form a continuous palisade of smallish, round or partially
26 hewn posts in a trench. This construction was then plastered with a heavy mud or lime plaster (crépis) on both sides to form a solid and weather-proof wall. In the same way, heavy hewn planks could be used as the foundation. Rather than using plaster, walls could be built, “with Spruce picketts very small in general covr’d in with boards and shingles.” 36 A second kind of construction was the pièces-sur-pièces, an abbreviation of, “piece of wood mounted on piece of wood.” This referred to a family of different techniques in which fully hewn members were laid horizontally. In one form -- probably borrowed directly from seventeenth century French fortification methods into folk architecture -- the ends of horizontal planks or logs were dovetailed together at the corners of a crib (Fig. __). This style became surprisingly popular in the east, and was carried far to the west and into Louisiana with diminishing frequency. It differs from the traditional log cabin construction so popular among Anglo-Americans in that the Canadian version the members are so carefully squared that it is almost unnecessary to use chinking material between them (Fig. __).37 A second variant of pièces-sur-pièces was à tenon en coulisse. Vertical posts were channeled out on those two sides in line with the wall. Hewn or sawn planks (madriers) were slid vertically down the channels to form a neat and solid wall with its verticals five to ten feet apart. If heaver logs were used, each piece was fitted with a salient tenon at its ends for the same purpose (Fig. __). En coulisse construction is clearly a variant of high-style carpentry technique used in France for paneling.38 These various forms of colombage and pièces-sur-pièces construction did not much outlive the seventeenth century in the St. Lawrence River Valley. The intense cold of the Canadian winters soon stimulated thicker and more insulating walls, and for this stone walls were widely adapted, particularly in Quebec. However, the earlier experience was clearly carried into the west by pioneers. We have taken note of the great difference between the dominant styles of wall construction in Upper and Lower Louisiana. In both cases the techniques were already obsolete in
27 France. The principal question is why French settlers in these places adopted such divergent methods. En terre buildings were almost ubiquitous among typical private settlers in both places in the early days, but curiously, very few open-studded buildings are recorded from upper French Louisiana, while very few close-studded buildings are known from Lower Louisiana. Perhaps one answer lies in the differences between the methods of early pioneering architecture firmly established in the periods of “initial occupance” of these two places. In Lower Louisiana, the open-studded bousillage-entre-poteaux method is precisely modeled on that found in country houses in western France in the eighteenth century (Fig. __). Since Louisiana was not settled until the first decades of that century, it would be natural for persons coming from France to employ the methods in vogue at the time of their departure. Coincidentally, open-studding was also used by enslaved Africans and Maroons in the construction of their small houses in Hispañola (Fig. __). Throughout Haiti and the Dominican Republic, open-studding with wattled panels between is employed for country bohios and ti kay. This tradition extends back, not only to France but to West Africa, and also to the houses of the Arawak Indians. Escaped African cimarrones (maroons) blended their architecture with that of the native Arawak to produce the ti kay of the early sixteenth century. It is still the dominant style of construction for peasant houses today. Thus, we have a perfect trifecta of reinforcement of the open studded method in Saint Domingue -- the jumping off place for almost all ships coming to Louisiana. The term piquet is essentially never used in these southern places. In my fieldwork on vernacular architecture in Haiti (1995-96), I never heard the term piquet used for even the flimsiest of posts. Similarly, it is not found in Louisiana documents describing house construction, poteaux en terre being the universal descriptor. Yet, in maritime Canada it was, perhaps, the most common term applied to vertical posts in the seventeenth century. The use of small piquets was abundant. It appears to have vied with pièces-sur-pièces as the most popular early form of house in the Maritime provinces into the middle of the eighteenth century. Occasionally, viewers distinguished between
28 piquets escarrés, squared posts in the ground, and piquets ronds, (or pieux en terre). [Richardson 1973:80]. In a study of twelve early eighteenth century piquet houses at Louisbourg, the diameter of the pickets varied between four and six French pouces, and the space between them was between one and two inches. It is clear from drawings made by French engineers that picket buildings were numerous in Louisbourg 1713-1758, and that the method was employed exclusively for smaller buildings.39 I could locate no detailed descriptions of the construction of seventeenth century houses in the Illinois country. Villages had not yet been established. Houses were scattered about incipient military posts or mission stations, but they were not yet numerous. Margaret K. Brown and Lawrie Dean have provided an invaluable service in their detailed descriptions of the Village of Chartres between 1720 and 1765, based on notarial and church records. Right from the beginning, the majority of houses in these villages are described as “built of upright pickets.” They were also small buildings of one or two rooms, ranging between sixteen and twenty feet deep, and twenty to thirty four feet in length. Some were built on sills, while most are clearly en terre structures, being described as “enclosed with [a palisade of]pickets.” At first their roofs were thatched. Some of the larger houses were also covered with straw, others with shingles. Thatching implies that the roofs were steeply pitched. Beginning in the 1730s, some houses are described as having “galleries on the two long sides.” For example, in 1741 Pierre Dirousse sold his small inn, “built of pickets” with double galleries for 1800 livres (p. 485). By 1755 houses are being described as “of upright posts,” and straw is still being used as thatch, but increasingly, houses are described as “built on sills.” By 1758, houses with “galleries all around” are beginning to appear. By this time the observation of houses, “falling into ruin” appears more often, providing some measure of the durability of the earlier en terre forms of construction. These descriptions raise important questions for understanding cultural process in the architecture of the Illinois Country. Why did the people of late eighteenth and even nineteenth
29 century continue to construct the majority of their houses with close studding? To answer that question we must ask who, exactly, constructed these houses? In Lower Louisiana it was African laborers and their progeny who built nearly all of the buildings of the colony after ca. 1724. Able African slaves were trained by French carpenters in the French colonial period, and they became the skilled builders of the area. “I have neither locksmiths, edge-tool men, nor sawyers. I cannot get along without hiring those whom you will send me to teach their trade to the Negroes.”40 For Louisiana carpenters, that training was based on eighteenth century techniques of charpente classique including open-studding methods, rather than the obsolute charpente medieval (close-studding) practiced in earlier centuries. After 1727, the population of enslaved Africans exceeded that of the free whites for the remainder of the French colonial period. “Most of the black slaves in New Orleans were apparently highly skilled. In 1743 Bienville reported that Joseph Dubreuil, the contractor of the king’s works, had trained blacks in all trades and employed very few French workmen.”41 Dubreuil employed about five hundred enslaved Africans who worked mostly building on the king’s properties, but also on his private commissions such as plantation houses. During the Spanish colonial period, the building professions were increasingly dominated by the growing population of gens de colour libres in lower Louisiana. Evidence of their skill at French framing technology is still to be seen in eighteenth century houses such as the Destrehan Plantation house with its magnificent roof support system worthy of a church in Normandy. It was constructed by Charles Paquet, a free mulatto carpenter in 1787 for Robert Antoine Robin de Logny, a St. Charles parish wealthy planter.42 In Upper Louisiana, the story is different. By the time African slaves began arriving upstream in the mid 1720s, the traditions of house construction had been firmly established by the Canadians for decades. A census of the Illinois Country in 1726 lists 188 black slaves and 83 Indian slaves. By 1752, these numbers had risen to 446 blacks and 150 Indians.43 In addition, the slaves which were released to travel upstream were almost certainly not the highly skilled and valued
30 Africans engaged in carpentry and construction. Thus, education in the building trades in Upper Louisiana was almost certainly on-the-job type training by Canadians already practiced in the more ancient techniques of piquet and close-studding. When we combine the nature of the dominant construction methods during the periods of “initial occupance” with the special skills of the two work forces, the reasons for the differences in wall construction preference are more clearly revealed.
5. Timber Wall Construction In the Illinois Country: the Wall Plate Lap Notch: One of the most important unanswered questions in the architectural culture history of the Illinois Country is the use of the Wall Plate Nap Notch on close-studded houses. Here, I will summarize; the issue has been dealt with elsewhere.44 I can find no architectural historian who has accounted for the origins of this popular local carpentry tradition. Houses built of piquets, poteaux and pieux in French Canada did not employ the well plate lap notch. Their members were either mortised into a wall plate, or, a heavy channeled wall plate was set atop the line of posts. In all buildings in the Illinois country with surviving wall plate lap notches at the heads of their posts, the beveled tenons which lap the exterior of the wall plates are fixed to it with nails, often two nails per tenon (Fig. __). Throughout most of the eighteenth century in the Illinois country, nails were very expensive. Brown and Dean list a pound of nails selling for 200 livres, or approximately $40.00 in 1724. At the same time a “master cloutier earned 150 livres per annum in New Orleans. This price would have been beyond the purchasing ability of most small settlers, but it began to drop following the Seven Years War. Nails were wrought by hand until the 1790s. The first cut nail factory west of the Alleghenies was established in 1795 at Brownsville, Pennsylvania.45 Because of the excessive expense of nails prior to the 1760s, it is highly probable that the wall plate lap notch was not in common use. Alternately, the lapped tenons might have been fixed with hardwood tree nails, but no evidence for such usage exists in surviving buildings. All surviving
31 buildings date from ca. 1790 or later. Although no proof exists, these facts lead me to believe that the lap notch is a relatively late local invention and that it became commonplace in the construction of houses immediately following the Seven Years War, when so many French Creoles moved to the western side of the Mississippi River.
6. The Creolization of Illinois Country Vernacular Architecture: Four Stages: Perhaps the most unusual aspect of Illinois Country vernacular architecture is the abundant use of broad galleries. To one with experience in the tropics, it appears that the oldest surviving houses of Ste. Genevieve were somehow miraculously set down there by West Indians. Before the adoption of the gallery onto houses of the Upper Mississippi Valley, all features of construction fall into one or another of the French based traditions, but by the 1730s, inventories of properties reveal an increasing use of full-length galleries on these otherwise purely Canadian style houses. What happened? It has been argued that the galleries of old Illinois Country houses derive from the St. Lawrence River valley.46 Elsewhere I have shown that the idea that Creole-style full façade width galleries were in use in Canada at the beginning of the eighteenth century is not supported by historical evidence. Canadian architectural historians uniformly agree that in the seventeenth century the term “gallery” referred to a completely different kind of addition to the house, mounted on the gable ends. The full Quebec verandah does not occur in the St. Lawrence River valley until the end of the eighteenth century:47
The verandah was not an original feature in Quebec and must have come to us from the south at the end of the eighteenth century or the beginning of the nineteenth century.
32 Many of our Quebec verandah houses are very like the Hudson valley “Dutch” houses. But such houses occur also, as early at least as 1750, in the French settlements on the Mississippi. Here they were probably derived from Louisiana, Mexico or the West Indies…. As early as the mid-eighteenth century French settlers, from Quebec, were building verandah houses in Missouri and they appear to have got their inspiration from further south.
Some historians believe that there was a natural progression in the development of the Illinois Country gallery, beginning with the belcast roofs of French and old Quebec houses (Fig. __), and expanding gradually into the Quebec larmier -- a kind of concave auvent or roof extension found on the front and rear long sides of houses along the St. Lawrence River (Fig. __). As the larmier was gradually extended up to four feet in width, posts were added to support its outer edges (Fig. __). Finally, a straight sloping roof four to six feet in width was attached at the tops of the walls, producing what we in Louisiana would refer to as a Class I gallery. This roof was supported by colonnettes. The floor beneath was elevated a foot or two above grade, producing a verandah (Fig. __). However, this seeming evolutionary enlargement of a decorative French roof is belied by history. As many Canadian architectural historians have noted, essentially all of this roof expansion process occurred in the last decades of the eighteenth century and the early decades of the nineteenth -- long after houses with galleries were already being reported in the Illinois Country. The earliest vernacular house with galleries in the Illinois Country of which I can find record was that of Lieutenant Pierre Melique, built in the summer of 1723 by François la Plume: “Posts in the ground, 30 ft. by 22 ft., floored, and with three doors and a galerie.48 It is not certain whether this gallery was supported by posts, or consisted simply of extended auvents in the manner of the Quebequois larmiers. By 1732, the idea of adding a gallery was catching on in the area. M. de Saint
33 Auge, Commandant of Fort de Chartres wrote a letter (May 25) in which he reports the planning of a 60 foot by 32 foot house with a gabled roof, two allonces (gabled end shed rooms) plus front and rear galleries “for the preservation of the soles” (sills).49 About the same year, Father Mercier in Cahokia built a house with two galleries running down the long sides of the house, “not for the beauty of the building, but for preservation of the sills, which will last half as long again.”50 Three years later, the same Father Mercier constructed the River L’Abbe chapel at Monk’s Mound. It was built along the lines of his earlier house (Fig. __).51
Archaeological reconstructions of the
galleries of early Illinois Country houses reveal that the galleries were narrow, such as those on the Beaugenou house, built in 1765 in St. Louis, and sketched in a reconstruction by F. L. Billon (Fig. __). This represents the first stage of creolization of the Upper Mississippi French vernacular house. I will refer to it as “proto-creolization.” During the first decades of the eighteenth century most of the houses were thatched with “straw” covering. This implies that the roofs were steeply pitched like those of Quebec. Roof pitches probably exceeded 45°. Houses constructed between 1699 and the 1730s are not described as having galleries of any kind. Galleries added to piquet houses in the 1730s and thereafter are clearly “protective galleries.” In other words, the idea of the gallery had diffused into the Illinois Country, but its principal tropical function as a living room had not. This kind of house is usually described in the inventories as something like “One house built of pickets, a gallery on the two long fronts…”52 Beginning in 1758 we find a new kind of descriptor for the houses of New Chartres: “One house with a salle-et-chamber, double stone chimney, two cabinets, a cellar, the whole covered with shingles, with a gallery all around… (11/10/1758).53 Here is the first evidence of an entirely new kind of Upper Mississippi Valley houseform – a fully creolized house with broad Caribbean style living galleries added to its peripheries. As we see from the surviving houses of Ste. Genevieve, these living galleries vary in width from roughly 7’4” on the rear of the Bequet-Ribault house, to a broad 10’10” on the Bolduc house. The extra width provided to these galleries clearly indicates
34 their new functions as living spaces as well as circulation zones. Environmental protection cannot account for the extra room-like width. By the time of the Seven Year’s War, the idea of a Louisiana model for houses surrounded by out-of-doors living spaces has finally penetrated the length of the Mississippi Valley (Fig. __). Yet the diffusion process still had a few tricks in store for the Illinois Country. In 1764, the trader Pierre Laclède founded the settlement of St. Louis as an extension of the New Orleans based trading firm of Maxent, Laclède & Co. Now, it was not simply architectural features which were being introduced into the new Missouri communities, but entire settlement systems with their patterns of street layout, their commercial principles and connections, and their symbols of commerce. One of these was the trader’s house – an elevated house surrounded by galleries which announced the principals of hospitality, temporary shelter and success in commerce to all passersby (Fig. __). The people of the Illinois Country had, no doubt, little idea of the antiquity of this particular symbol. It extended back through New Orleans to the Caribbean, and particularly the town of Cap François on the northern coast of Saint Domingue (Fig. __). From there it ran back across the Atlantic to the first creolized Afro-European trading posts in Senegal, and Portuguese Guinea. Beginning in the early sixteenth century, trading houses of the early Lançados and Tangomaos developed into a distinctive style of commercial architecture. Elevated rectangular houses fronted by broad reception galleries were known as maisons à la portugaise. They were adopted by Afro-Portuguese traders in coastal and riverine settlements of the coasts of Upper Guinea. This originally mulatto population of cultural brokers dominated the trade between Europeans and indigenous Africans between ca. 1500 and 1750. By the first decades of the seventeenth century, however, French and English traders had begun to establish their own trading posts, adopting a somewhat more stylish European form of the “Portuguese style” trader’s house. By the late seventeenth century, earlier in Brazil, these elevated houses had became the models for the maisons des maîtres -- that is, the successful sugar planters on the plantations of Brazil, the West
35 Indies, and Louisiana. They are recorded beginning in the 1720s in the French settlements of the Mississippi Gulf Coast and Louisiana (Fig. __). Now, four decades later, they were being imported as entire packages into the Illinois Country. Beginning with the Laclède house, the very first house in St. Louis, raised trading house architecture was now borrowing directly from the architectural symbols of commercial and social success in Lower Louisiana. The Laclède house in both its earlier and its post 1795 form, and the later Pierre Menard house in Kaskaskia, IL, are excellent examples this kind of architectural package. Even smaller houses whose owners had commercial or social aspirations imitated the larger trader’s houses. The Dodier-Sarpy house in St. Louis, and the Nicolas Janis and the Guibourd-Valle houses in Ste. Genevieve, draw inspiration from more ancient Atlantic World trader’s architecture. 54
7. Reverse Diffusion: There is an interesting coda to this discussion. Previous commentators have always described the influences on the French settlements of the Illinois Country as deriving from Quebec or Louisiana. There has been essentially no discussion of possible influences of the Illinois Country on Lower Louisiana. Yet, there is evidence, I believe, to support the notion of important reverse diffusion of architectural ideas downstream into lower Louisiana. My example is to be found in the form of the Class II Creole house (Fig. __). This houseform dominated French Mississippi Valley settlements during the middle and later decades of the eighteenth century, passing out of popularity in the 1790s. In Lower Louisiana it was primarily a rural form, but it was the symbol of success of the rural planter or farmer throughout this period. The Class II Louisiana Creole house is characterized by an asymmetrical core of rooms based on the traditional French salle-et-chamber plan. That modular core module was entirely surrounded by non-European style spaces including any combination of open galleries, loggias and enclosed narrow bedrooms (Fig. __). It was a fully creolized architectural form. Many dozens, perhaps hundreds, of these houses were constructed in the eighteenth century, though only a few
36 are known from urban areas. One of those is the ca. 1725 “House of the Engineer” (Adrien de Pauger) in the 400 block of Decatur Street in New Orleans. The essential feature of this house type is its very steep inner roof, coupled with a low-pitched roof brim which covers the surrounding gallery spaces. Now, there were many houses in New Orleans at the time which had surrounding galleries, but in most cases, these house had relatively low-pitched roofs of Mediterranean style, then popular in the French colonies of the West Indies. The Class II Creole house, rather, is characterized by an inner roof pitched at roughly 45°. In addition, many of these houses have extremely steep croupes or end-sheds pitched even more steeply. The basic roof of the Class II Creole house is modeled on the Norman/Quebecois “pavilion roof” with its steep end-sheds. Since prior to the settlement of New Orleans in 1718, permanent houses in Lower Louisiana had had shingled rather than thatched roofs. This is clearly shown in the Dec. 1720 Le Bouteau drawing of the Biloxi settlement, where workers are using the selle à tallier (shingle horse) to make shingles (Fig. __). Because shingles were ubiquitous, there was little need or requirement for a very steeply pitched roof surface. Such roofs are more expensive to build. Yet, dozens of surviving examples of the Class II Creole house retain “witches cap” profiles with steep inner roofs. Recall that during this period, a hundred or more traders and settlers were moving between upper and lower Louisiana annually, and a northern source of influence for the steeply pitched Class II Creole roof becomes more probable. In Upper Louisiana in this period, most roofs were still thatched, and therefore steeply pitched in the Canadian fashion. Upper Louisiana retains the largest number of Class II Creole roofs of any French settlement in the old Louisiana territory (Fig. __).
Conclusion Architectural historians generally do not consider the Atlantic World to extend into mid-continental regions. That construction is generally conceived of as a maritime complex based in the coastal zones, rivers, and islands of the continental rim-lands. The entire concept of “Atlantic-ness” invokes
37 a vision of the hardy mariner carrying scarce commodities, settlers, and new ideas in sturdy wooden ships across the unfathomed reaches of the vast oceans which separated the old world and the new. In addition, the Atlantic World to historians, has meant primarily the North Atlantic. Few historians have emphasized the essential culture-building role played by the Caribbean and the coasts of West Africa. Indeed, until very recently, the latter has been essentially ignored as exerting any influence whatsoever on places like Louisiana, to say nothing of the Illinois Country. Illinois Country history also suffers from another set of biases, well articulated by historian Bradley Bond.55 “Because Louisiana remains outside the dominant national story of English-speaking settlement and westward expansion into that heartland, the heirs of American exceptionalism and the assigns of patrician historians consider French colonial Louisiana, if not a phantom place, then a peculiar place that is tangential to the stories they tell.” It has been our presupposition that not only does the Illinois Country not stand outside of the Atlantic World in the eighteenth century, but that it has played an important part of the story of that vast arena of intense trading and cultural exchange. Perhaps more than most Atlantic places, the Illinois Country stood at the confluence of two great streams of flow of Atlantic World culture— the northern and the tropical. Although it appears isolated and land-locked in the American imagination, in fact it was no more isolated than the ports and trading stations on the St. Lawrence, the North Sea, the Seine River, the Thames River, the Senegal River, the Gambia River, the Casamance River, etc. It was said that in terms of actual travel time, New Orleans was about equidistant between Paris and Ste. Genevieve. That amazing lack of isolation is seen in the fact that in the 1730s, the weight of flour shipped down to Lower Louisiana annually averaged over a hundred thousand pounds, though with great variation from year to year.56 It alone, supported the equivalent of a population of several hundred people in the New Orleans area. Given the intensity of the Mississippi River commerce in the French Colonial period, it would be surprising, indeed, if that were not reflected in its culture, and particularly in its
38 vernacular architecture. As we have seen, the level of influence of Lower Louisiana intensified throughout the eighteenth century, until many of the houses of Cahokia, Prairie du Rocher, St. Louis and St. Genevieve were not unlike those of Pointe Coupee parish and St. Charles parish, in rural Louisiana. What is perhaps, more surprising is that this diffusion may well have represented bidirectional influences. Throughout the eighteenth century the architecture of the entire French Mississippi Valley had as many similarities as it had differences. This corridor must truly be considered an integral component of the Atlantic World. Finally, we find that the creation of Illinois Country vernacular, as we experience it in the surviving examples, was a relatively late phenomenon of the colonial period. Features such as the lap tenon assemblage, raised trader’s houses, and surrounding galleries all pretty much post-date the Seven Year’s War. They are mostly representative of the 1790-1820 period. As we consider the place of the Illinois Country in the history of America, we should not forget its significance in the larger Atlantic World and in the history of French North America.
