James Murton

creating order: the liberals, the landowners, and the
DRAINING
OF SUMAS LAKE, BRITISH COLUMBIA

ABSTRACT

This essay considers the draining of Sumas Lake, British Columbia, Canada, by the provincial government in the 1920s. I argue that both the B.C. state and local landowners believed in the virtue of the ordered landscape of agrarian ideology. Both supported a quest to order nature that clashed with a disordered and inter-connected natural environment, leading to massive cost overruns on the building of drainage works and the creation of farmlands. The state battled local landowners over who would pay for the project. Landowners argued that the drainage had not made their farms more profitable. Yet both groups continued to agree that the drainage represented progress and improvement. The landowners' continued support of the project despite its cost and their meager gains suggests that the problems of the project lay less in the general logic of the state than in the discourse of liberal agrarianism that both supported. The financial and environmental problems of the project owed more to the limits of this shared vision than to inherent limitations in the state's relationship with nature or the local.

LATE IN 1919 a high-powered group of British Columbia (B.C.) politicians and bureaucrats traveled outside Vancouver and into the agricultural lands of the Fraser Valley. At the little town of Huntingdon the group–Premier John Oliver, Minister of Agriculture E. D. Barrow, and William Latta of the Land Settlement Board (LSB), the province's agricultural settlement and development agency—met with a group of local landowners. On the agenda was Sumas Lake, or, to be more specific, the LSB's plans to do away with it. With the approval of the landowners and of nearly everybody else who mattered in the province, the Land Settlement Board was planning to drain Sumas Lake.

Draining the lake was an old idea. Plans and schemes stretched back to the 1870s. For those who favored an orderly, settled countryside, Sumas Lake was an ongoing problem. For one thing, it was hardly orderly. Flowing into the mighty Fraser River and so subject both to its floods and to those of several other streams, it was unusually variable. Over the course of a year, it might vary from a winter low of nine feet deep to a spring high of thirty-six feet, and maps could agree only on its general outline. In 1894 what is still the largest flood on record on the Fraser caused the lake to double in area, covering both local Indian reserves in water.¹ The lake was also simply a barrier to Euro-Canadian settlement. Nestled in between Sumas Mountain on the north, and Vedder Mountain and the U.S. border on the south, it was skirted by railway tracks and roads, which held it back, most of the time, with ditches and dikes. In a province short on flat and clear, let alone arable, land, it occupied a large chunk of potentially prime farmland not sixty miles from Vancouver (see Map 1).

Map 1: Sumas Diking District.

Cartography by Eric Leinberger, © UBC Press 2007. Adapted from "Sumas Drainage, Dyking and Development District in the Lower Fraser Valley, Province of British Columbia. Choice Farm Lands for Sale," pamphlet, 1926, BC Archives, Library Collection, NWP 971.24 S949c.

Dashed lines show the outline of the former lake as indicated on the original 1926 map, though as noted in the text (and demonstrated in Figure 3), Sumas Lake had no definite shape. Note the Fraser River in the upper right corner, and Yale Road and the BC Electric Railway tracks across the bottom.

The LSB's plans aimed at "reclaiming" the lands under the lake water for farming and also imposing order on the lake and this part of the Fraser River. Premier John Oliver's purpose in visiting the area, however, was not to trumpet the good news to the owners, but to deliver a warning. He wanted to dissociate his government from any responsibility for the project's ultimate cost. Though the government was to front the funds for the scheme, it was ultimately to be paid for by the current landowners in the area and by the sale of the new lands that would be created. The secretary at the Huntingdon meeting recorded Oliver as stating that "the general impression was that the Government were backing the scheme; he was here to right the people's minds re this point and denied any such implication; stating plainly in the following language 'The Provincial Government is taking no responsibility whatever for this scheme; understand that clearly.'"²

Oliver's statement was both necessary and unsurprising. Necessary because his government was doing more than anyone in British Columbia had ever seen or probably expected of a provincial state that up to World War I had been mostly concerned with arranging for resource rights and railway contracts to fall into the hands of the businessmen that made up the province's political class.³ But it was unsurprising because of this same history of laissez-faire statecraft. Several historians have suggested recently that Canada (and British Columbia) can be best understood as an ongoing attempt to build a "liberal order," a society centered on the individual and the fostering of his or her prosperity.⁴ Oliver's government, I argue elsewhere, was part of what Barry Ferguson has identified as a remaking of that liberalism in this period, into a new liberalism with a greater concern for the social order and a recognition of the importance of society to the liberal individual.⁵ As Oliver's comments suggest, however, he and his government retained a strong sense of classical liberalism's focus on individual responsibility.

Oliver later claimed that he had issued a more elaborate warning than what the minutes of the Huntingdon meeting recorded. "I pointed out ... [to them] the absolute unreliability of engineers' estimates," he claimed; "I pointed out to them the past experience of diking indebtedness in the Fraser Valley. I went further than that, and I pointed out to them that as a result of the war, costs of labor, cost of material, had increased enormously."⁶ If Oliver really did say all this, then he showed great prescience, for the Sumas project quickly became an enormous financial disaster and an ongoing source of controversy. Though Sumas Lake was drained and new farmland was created out of the former lakebed, the project failed to create security for the farmers on the edges of the lake. It failed as well to create a landscape that both the government and local landowners could agree was ideal—an agrarian countryside of relatively small, individually owned farms. Across classes and in both the province's cities and its rural areas, many believed deeply in the agrarian ideal of the hardy individual farmer.⁷ This belief drove the province's attempts to foster agricultural settlement across the province, and was deeply ingrained in Oliver, himself a Fraser Valley farmer.⁸ Agrarianism's emphasis on the individual meshed nicely with the wider liberal order. Thus the state and landowners could agree on the future of the region. The landowners, despite Oliver's warnings, did eventually agree to finance the project. Further, they held on to the idea that the project represented progress, even after it became clear that in concrete terms it had helped them little, if at all.

This point of agreement is important in what it can show us about the historical relationship between the state and the environment. Environmental historians have not had enough to say about the state.⁹ A wide literature focuses on groups and forces—tourist promoters, sportsmen, conservation groups, and scientists—who attempted to capture and influence the state.¹⁰ Less common are works that focus squarely on the state itself and attempt to theorize, and work through the implications of, its interest in the environment. There are some notable exceptions. Donald Worster argued twenty years ago that the U.S. state was essentially a creature of capitalism, and was driven by an ideology of instrumental reason that saw no worth in a nature that was not being put to use increasing wealth.¹¹ More recently, James C. Scott has put emphasis on the state's own needs and interests, as determined by its very nature. Unable to manage (or even gather) the volume of data that would be necessary for full knowledge, the state instead historically has sought to make its subjects and terrain "legible" through a process of simplification, abstraction, and standardization. Detailed local knowledge and relationships, historical and ecological idiosyncrasies, all were defined out of the picture. Yet, Scott argues, such local knowledge was often key to making complex social and ecological systems work. Thus Soviet collective farms failed spectacularly while the small plots tended by stubborn Russian peasants thrived.¹²

Scott's critique of the logic of state planning is powerful and persuasive. Yet state attempts to organize society and environment may derive less from a general "logic" than from historically specific discursive and ideological formations, such as Worster's instrumental reason. Understanding the B.C. state's engagement with the environment in the interwar years means understanding the discourse of agrarian liberalism that shaped the relationship.¹³ We will find the problems of the Sumas project rooted in the contradictions and limitations of this countryside ideal, one that united both the Liberals and the landowners.

DISORDER

WE CAN MAKE TOO MUCH of this agreement between white male settler-farmers and the state, however. I will argue that it is an important and telling confluence of belief, suggesting that the sort of liberalism extant in British Columbia in the interwar period was as important as the structure of the state in understanding the engagement between the government and the environment at Sumas. The landowners were not the only inhabitants of the Sumas region, just the ones with the closest relationship to the levers of political power. Other stories about the lake could be, were, and are told, as has been vividly demonstrated by Laura Cameron's multinarrative of Sumas Lake's history, Openings: A Meditation on History, Method and Sumas Lake.¹⁴ And even the experience and understanding of the lake by white male settler-farmers was quite different from that of state planners and engineers. For while planners and engineers assumed the superiority, if not the naturalness, of an orderly nature, everyone in the Sumas area had learned to live with the fundamentally disorderly creature that was Sumas Lake.

The strongest factor in Sumas Lake's disorder was its connection to the Fraser River. The great river fed it tidal effects from the Pacific Ocean, and its regular floods swelled the lake as well (see Map 1). The Fraser rises in the Rocky Mountains near Jasper and the continental divide, on the eastern edge of British Columbia. Initially flowing northwest, it turns south near Prince George, crossing the rolling, arid expanse of the interior plateau until it meets the mountains. At Lytton the Thompson River, itself the result of the joining of the Thompson and the North Thompson at Kamloops, flows into the Fraser, and the resulting river cuts the spectacular Fraser Canyon through the mountains until it reaches Hope. Here the river leaves the mountains and turns west into the alluvial floodplain of the lower Fraser Valley. Near Vancouver it splits into two large arms and flows around the Vancouver suburb of Richmond and into Georgia Strait. Along the way the river picks up soil and debris, giving it a brown, muddy look.

But this description hides as well as reveals. It gives in to the temptation to view nature as timeless, describing the river as it is today, not as it was in the 1910s and 1920s. For nature has a history, as any number of environmental historians have pointedly reminded us. The history of Sumas Lake confirms the argument. Every summer the Fraser rises and, often, threatens to flood parts of the lower valley. But due to the distant sources of the Fraser, its large tributary (the Thompson), and the complications of weather in the mountain ranges of British Columbia, these floods are unpredictable. Flood waters result from the melting of snow in the Rocky and Cariboo Ranges through which the Fraser flows. These can be added to by snow melt at the headwaters of the Thompson River. The critical element is hot weather in the interior of the province, producing a large snow melt. Should hot weather in May be preceded by cold weather in the late winter in the mountains, resulting in a large snow pack being retained into May, the conditions for floods on the lower Fraser are set. The timing of the peak flood on the Thompson also is important. Generally the Thompson's flood passes through the lower Fraser before the flood from the Fraser's own headwaters, as it is farther south (and thus melts sooner) and is closer to the Fraser Valley. But occasionally high waters at the headwaters of these two rivers coincide, as happened in 1894 and in 1948. The result is a large flood in the Fraser Valley and the river delta south of Vancouver city.¹⁵ The 1894 flood rose to 25'9" on the high water gauge at Mission, causing millions of dollars of damage to homes and businesses and prompting the government to begin diking works.¹⁶

So the Fraser was not constant and unchanging. It changed in volume and character over the course of the year and from year to year. Its tendency to flood at irregular levels and unpredictable times, and its parallel tendency to meander across its floodplain and occasionally cut new channels for itself, were wholly natural parts of the way the river worked. Such was true of other rivers in the area as well. Thus in 1890 the Vedder River, which until now had flowed into the Fraser River to the east of Sumas Lake, cut a new channel for itself. It turned west into the lowlands near the Fraser and flowed into Sumas Lake on its southwestern side. Engineers planning the demise of the lake now had to contend with a much larger body of water.¹⁷

The changing conditions of the Chilliwack River, from which the Vedder flowed, also complicated the behavior of Sumas Lake. The Sumas River was comparatively regular. It drained mostly flat land, raised its level only due to rain and thus did what was expected of it—it swelled in the winter. Under hot summer conditions when the Fraser was likely to flood, the Sumas dried up. The Chilliwack was another matter, and flooding on this river presented a different problem. Compared to the source of the Fraser, the source of the Chilliwack was in a mountainous area lower in elevation, further south by four hundred miles, and affected by temperate winds from the Pacific. Although melting snow did produce some flood waters in the spring, the real problem came in the late fall or early winter when a combination of new snowfall was melted by warm winds at the same time as heavy rainfall swelled the river. Aside from flooding, the rainfall had two other problematic effects. The same rains could also raise the Fraser, while at the same time they choked the Chilliwack with debris such as gravel, sand, silt, and even trees. While not a problem for the lake, floods of debris were an obvious concern for anyone hoping to build a dike.¹⁸

Fed by floods and pulled about by tides, Sumas Lake was intermittently shallow and small, large and deep, and everything in between. Parts of it were swampy and its edges were ill-defined. Yet it was never the marshy wasteland that development boosters portrayed it as. When the lake waters receded, fields of blue grass sprang up in the silty soil. The grasslands of Sumas Prairie attracted osprey, bald eagles, flycatchers, and great blue herons, while the waters themselves were home to migratory salmon, round-fish, and sturgeon. At least by the time of the arrival of white settlers, hazelnut trees and wild strawberries grew. English naturalist John Keast Lord, visiting the area as part of the International Boundary Commission in the late 1850s, described a "'pretty'" and "'picturesque'" site, and Barbara Beldam, granddaughter of an early white settler, claimed that the area must have appeared to her grandfather as "the proverbial Garden of Eden."¹⁹

Those living in and around the lake dealt with its variability. The lake's floods, for instance, figured into the teachings of the native people of the area, the Sto:lo. One story told of how a group of people survived a severe flood by tying their canoes to Sumas Mountain. Gradually the waters receded, and when this happened Xexa:ls—the transformers—visited those who had survived the flood and taught them "the right way to live, work and pray."²⁰ Sto:lo fishermen caught sturgeon using a weir on Sumas River, gathered wapato (wild potato) on the prairie lands around the lake's edge, and killed ducks for food. They accommodated themselves as well to the clouds of mosquitoes that arrived in the summer, building a village on stilts in the middle of the lake where the mosquitoes did not go. Later, Sto:lo farmers grazed cattle on the borderlands, as did the white settlers who began to arrive in the 1860s. Because the lands adjacent to the lake were flooded regularly they did not pass into private hands, and so could function as a commons. Local customary arrangements governed their use. Local people also fished and hunted in and around the lake and used it for boating, swimming, and skating. Flooding could be taken in stride—Laura Cameron quotes one early white settler, Mrs. Fadden, who remembered of the flood of 1894 that "the water was 'spreading over garden, over orchard, quite high. Fine day'." Mrs. Fadden also recalled how "'Wild roses used to bloom just at the top of the water. And there was the very lovely perfume that came from them as the water came up to them'."²¹

For a time local people were allowed to continue with this negotiated and subtle interaction with Sumas Prairie (as the land surrounding the lake was known). This relationship, however, always existed within a set of larger structures that defined the relationship between this place and these people, and the wider world that encompassed them. Canadian historians are increasingly using the term "liberal order" to describe these structures and relationships. Canada, historian Ian McKay argues, can profitably be thought of as a "project of rule," implanting the "politico-economic logic" of liberalism both "across a large territory" and in the hearts and minds of people.²² Throughout the nineteenth century in Canada, liberal reformers challenged political and economic norms that emphasized such things as an established social order, the inherent right of particular men to rule, and endowed rights to land.²³ These reformers called on the state to guarantee an individualistic social order and a laissez-faire economic order, a "formal, rule-bound arena in which competition could occur without unfair advantage or interference."²⁴ Intercut with racialized and gendered discourses, the liberal order understood the liberal individual to be male and white. Combined with agrarianism's valorization of the independent farmer, the liberal order drove (and was supported by) changes in the land itself—"social ideology," in McKay's words, was "set down on the land and hence made part of everyday ... experience."²⁵ A pamphlet, put out by the LSB to promote sales of lands reclaimed from the lake, idealizes a transformed, liberal landscape, appropriate to an individualistic social order and a farming sector of independent commodity production (see Figure 1).²⁶

Figure 1: Sumas Reclaimed Lands.

Image GR929V48F1, courtesy of Royal BC Museum, BC Archives

From the cover of a pamphlet produced before the lake was drained, this image shows the LSB's (and the unknown artist's) vision of the ideal rural landscape for British Columbia. An individualistic, quasi-English landscape (note the hedges) sits underneath the soaring peaks of the Coast Mountains. 1923–1924.

This was a world that white settlers were at home in and that the Sto:lo, by definition, were not. The Sto:lo nevertheless continued. We can see that white farmers and native people often used the lake and its surrounding lands in similar ways—grazing their cattle on its grasses, hunting for deer and birds, fishing and vacationing. But the extension of the realm of private property to cover the lands under Sumas Lake could only marginalize the Sto:lo. A liberal landscape of individual properties, bounded by imaginary lines made real by law and patrolled by individual property owners, became a powerful tool of the colonial state for removing First Nations people from their land.²⁷

For the Sto:lo, the process of dispossession began in 1858, when the discovery of gold upstream of the Fraser Valley led to an influx of miners into their territory. As the miners moved upriver, the governor of British Columbia, James Douglas, had reserves of between 400 and 9,600 acres laid out for the Sto:lo in the Fraser Valley. As far as the colonial government was concerned, these lands were all that the Sto:lo now controlled out of all the lands that they had once lived in. Yet even this was too much for the governments that followed Douglas's. The Fraser Valley contained simply too much good farmland for much of it to be left to the control of native people. In 1867 Chief Commissioner of Lands and Works Joseph Trutch had the Sto:lo reserves reduced in area by 92 percent, leaving only three small reserves in the vicinity of Sumas Lake—the Upper Sumas, Aylechootlook, and Sumas Cemetery Reserves.²⁸ As well, native peoples were excluded from preempting property in British Columbia. Despite various efforts at revision, Trutch's reserves largely remain in place today.

With the land now cleared for white settlement there appeared, in the 1870s, the first of many schemes to control the fitful nature of the Fraser and the other, lesser rivers in the area, and to do away entirely with the lake itself. The first act of the provincial legislature to regulate diking at Sumas was passed in 1878. This act also allowed for reclamation work in neighboring Chilliwack and Matsqui. Gilbert Malcolm Sproat of the Joint Indian Reserve Commission, formed to settle the question of native reserves in British Columbia, protested on behalf of the Matsqui Band. Noting that local First Nations people had not been consulted, Sproat pointedly stated that they "must have winter and summer grazing land ... reasonable area and their fishing places ... Is this unreasonable for a law abiding people whose land title is inextinguished?"²⁹ Sproat's protest was cut short, the provincial government ordering him to desist and assuring the Dominion that native people would benefit from reclamation, as it would improve the value of their land.³⁰ This argument, a standard justification for seizing First Nations land, would later also be used to justify the expense of reclamation works. Yet it was largely meaningless, stripped of consideration of how the land had been and would be used. Whites as well as natives would lose their use of the prairie commons for grazing and the lake for hunting and swimming. Also, as we will see, some of the settlers would come to regret their support for the diking scheme, the benefits of privatizing the lakeside commons proving not to be worth the cost. In a larger sense, though, white male settler-farmers were far better positioned to take advantage of whatever benefits reclamation might bring than were the Sto:lo.

ORDER

THAT THE DRAINING OF SUMAS LAKE was considered to be a key part of the progress of liberal rural society in British Columbia is shown by the multiple efforts to achieve it over a forty-year period. That white male settler-farmers, despite their accommodations to the disorder of the Sumas region, agreed with this desire for order, is shown by their ongoing support for the various reclamation efforts.

The 1878 act was amended four times, then superseded by other acts in the 1890s. A government-sponsored scheme in 1894 saw further plans drawn up, tenders issued and bids accepted, only to have the work cancelled due to lack of funds.³¹ The twentieth century saw three serious attempts to drain the lake, not including the LSB's, all of which ultimately failed due to unwillingness or inability to commit the necessary funds. In 1905 the Sumas Development Company of Seattle proposed redirecting the Vedder River back into its old channel. However, this would have created a violent current needing large dikes to contain it. Local owners, afraid of a breach in the dikes, protested. A new plan that would have diverted the river into a different channel met their approval, but financing failed.

The most hopeful of the three schemes came along three years later, when the B.C. Electric company became interested in the Sumas problem. B.C. Electric was the province's largest corporation at the time, running the streetcar and interurban railway system in Vancouver and providing much of the electricity for the Lower Mainland. It had a corporate interest in being rid of Sumas Lake, in that the water almost entirely cut off the route for the corporation's planned extension of its interurban lines to Chilliwack. In the end, though, the corporation's London head office decided to abandon the attempt, in favor of an interurban line that sat on top of a dike snaking around the southern shore of the lake (see Map 1). In 1914 the L.M. Rice company, also of Seattle, went so far as to register plans and prepare a table of the assessments to be levied on local owners. However, the Canadian government held title to the lands under Sumas Lake, as part of the "Railway Belt" lands conveyed to the Dominion to pay for the Canadian Pacific Railway. It refused to grant title to the lands until the works were completed, but the Rice company's financing depended on having clear title before starting work.³²

Throughout local landowners maintained support for these various schemes. In 1893 legislation had created a board known as the Sumas Dyking Commissioners. This board, which reported to the landowners, officially oversaw all diking schemes. Thus the landowners were legally in charge of all diking plans and had to agree to any scheme before it could go forward. This situation was complicated during the Great War by the passage of the Land Settlement and Development Act (by which the Land Settlement Board was created). A section of this act authorized the LSB to act as diking commissioners. In 1917 Minister of Lands Duff Pattullo suggested to E. D. Barrow, the Minister of Agriculture, that the LSB could take over as Sumas Dyking Commissioners if local landowners were to petition the board to do so.³³ It is not clear what subsequent role the government might have played in mobilizing the landowners, and clearly there was a cadre of owners more excited about the project than others. The LSB questioned landowners who had not signed the petition, and one, Joseph Schmidt, stated his opinion that those leading the charge for reclamation were "educated men and talk like a streak and put things in a light to further their purposes." Schmidt's astuteness was further demonstrated by his great concern with how the project would be paid for and whether the government would assume financial responsibility. Even he, though, thought it would be an improvement to the area. "I myself have more lowland than highland and if the dyke is a success I have a good share in it," he explained, "but I am doing well as things are and will not venture too much to be harassed and embarrassed with extraordinary taxes."³⁴

Schmidt believed that there would be little opposition to the diking plan and, sure enough, in October a petition appeared requesting that the diking commissioners sever their contacts with L.M. Rice and company, and that the LSB take over. Thus in February, 1918, the LSB officially accepted the role of Sumas Dyking Commissioners, placing them in charge of the reclamation of the Sumas lands.³⁵ By taking on the role of Dyking Commissioners, technically a separate body from the LSB, the board ensured that it could not be held legally responsible for the cost of constructing the reclamation works. Instead, the landowners would be on the hook. Meanwhile, Barrow kept up interest in the scheme among the landowners by organizing a conference to discuss the mosquito problem, held in September following an especially bad mosquito season. Barrow declared that it would be "criminal neglect" not to deal with the problem now, and marshaled Canadian government Entomologist Gordon Hewitt and Hewitt's chief assistant in British Columbia as expert testimony in support of drainage as the only truly effective solution. Hewitt's assistant had the previous year suggested biological control measures as well, but now was fully behind drainage.³⁶

The new diking commissioners immediately hired engineers to review and revise the plans that had been drawn up for earlier attempts. The plan that they finally settled on was similar to the one that had been proposed by the B.C. Electric. At its heart was the diversion of the Vedder River into the Fraser (see Map 1). The Vedder, with its disturbing ability to change its own course and thus alter nature itself, was the most disorderly component of the Sumas Lake area. Ordering it was crucial. The river would be intercepted as it moved west and cut through the embankment on which the B.C. Electric line sat. From the embankment, two dikes would be built on either side of the river. These would then curve northward and become the East and West Vedder Dykes, diverting the Vedder into the Sumas river near where it joined the Fraser. At this point the East Vedder Dyke was to connect with a dike running along the southern edge of the Fraser. The western dike connected to the Sumas River High Level Dyke, which in turn connected with the main dam in the Sumas River. The dam was designed to keep the flood waters of the Fraser from backing up the Sumas River and into the lake.

The Vedder dikes had to be designed to contain a stream flow of up to 27,000 cubic feet per second. Further, as the newly diverted Vedder now ran into the Sumas and thus into the Fraser, the dikes had to be made large enough to withstand flood water backing up the Vedder Canal. Finally, as the canal ran through low ground for most of its length, the dikes had to be larger than any other on the project. The canal was to be 300 feet wide and 15 to 20 feet deep, 58 feet thick with 100-foot-wide protective earth walls (or berms) inside each dike. It was to carry a capacity of 60,000 cubic feet per second.³⁷

Designed to contain a flood as big as that of 1894, these dikes constituted the major protection works. The other half of the project was drainage. The entire Sumas Reclamation Scheme was designed to create farm lands by draining Sumas Lake and also to drain surrounding lands previously subject to inundation. The lands affected were divided into three areas: the lakelands and the east and west prairie sections. The east prairie constituted five thousand acres of land lying between the East Vedder Dyke and the town of Chilliwack, and the west prairie section fifteen thousand acres lying to the west of the lake. Reclaimed, the lake would yield ten thousand acres of land. The east section was to be drained via a network of ditches, which would flow into the Fraser when it was low and be pumped out when the Fraser was in flood.³⁸ The lakelands were to be kept dry by ditches that would drain into the Sumas Lake canal, built across the middle of the old lake bed. These waters, coming from land that was always below the level of the river, had to be pumped into the Fraser via a major system of pumps constructed near the main dam. The west prairie section also drained into the Sumas Lake canal as well as into another canal, but constituted a separate system that could be allowed to drain out by gravity for about nine months or so.³⁹

The east and west prairie sections were mostly lands already owned and cultivated. The landowners in these areas would be the immediate beneficiaries of the protection works. They would gain relief from the yearly floods, which threatened crops planted on lowlands or rendered such lands good only for pasture. It was on the basis of this benefit, and that the landowners had asked the government to carry out the work, that Oliver built his argument that these landowners should pay all of the costs that could not be covered by sales of the reclaimed lands. Though the government had a long interest in the reclamation of the Sumas area, officials had always assumed that their role would be largely hands-off. It would consist of organizing, through the office of the Sumas Dyking Commissioners, work that would be carried out by private concerns. When the LSB assumed the role of diking commissioners, it inherited this idea.

That the LSB was now taking on work previously assumed to be beyond the government's purview demonstrates the changes in the role of the state in the new liberal era. In the first third of the twentieth century governments across North America moved to address problems created by industrialization in the cities, farms, and wilderness areas of the United States and Canada. Progressive reformers argued that experts in the employ of the state could build and manage urban infrastructure, reform the poor, and efficiently marshal natural resources for the common good.⁴⁰ British Columbia came relatively late to this movement, and the Sumas project demonstrates the Oliver government's limited conception of this new role of the state. On the one hand, the project clearly was always something that would benefit the entire province. Oliver himself included a special provision in the Land Settlement and Development Act allowing the LSB to take on such work, and stated in 1917 that the Sumas drainage would be one of the first priorities of the new board.⁴¹ Powerful people agreed. In August the Vancouver Board of Trade sent Oliver a letter containing a petition from a Farmers' Convention, arguing that it was imperative that the province produce more food for itself and that therefore the government should decide immediately on the reclamation of the Sumas lands. They also wrote a letter to LSB Chairman Maxwell Smith noting that they had been urging the reclamation for three years.⁴² The newspaper Vancouver World declared that the project was of special interest to Vancouver and more generally would "greatly enrich [the] province."⁴³

And so Oliver's claim that the government was merely acting on the wishes of the area landowners was dubious. That he made the point so emphatically to the owners and to anyone else who would listen—stating to the press that "'the government guarantees nothing ... The property will have to carry the scheme'"—suggests that he knew this.⁴⁴ Oliver and his government had adopted a limited concept of the role of the new liberal state. In Oliver's mind, the state might lead development projects, but these were ultimately aimed at the prosperity of the classical liberal individual. That so many let Oliver get away with his argument that the owners were responsible for the costs of reclamation suggests British Columbia's citizens—including the landowners themselves—were not far from Oliver in their understanding of the relationships between the individual, society, the environment, and the state. The B.C. government was willing to consider leading the reclamation effort—a task it had previously left to private industry—and employing government experts to do so. It was not, however, willing to pay for the project. The owners signaled that their agreement with this philosophy in March, 1920. Despite some fears of cost overruns, in that month they approved a budget of $1,800,000 for the project, all of which was to come from sales of the former lakebed and from assessments on their property.⁴⁵

BUILDING THE WORKS

THUS THE LIBERALS and the landowners had an agreed-upon understanding of the proper relationships between individuals, society, the state, and the environment, one which under-girded the March 1920 agreement. This discourse, however, contained two major contradictory elements that proved to be central to the difficulties of the project.

James Scott has argued that the roots of the state's inability to embrace and manage complexity lies less in discourses such as liberalism than in the nature of the state itself. Scott argues that the state, by virtue of its goals and tasks, is incapable of perceiving and dealing with social and environmental complexity on the level that is necessary for social and environmental engineering. I do not wish to refute this argument so much as I want to demonstrate the value of further historicizing it. Following British historians Phillip Corrigan and Derek Sayer, I argue that the state is best thought of as not only a set of structures but also a "state-idea" that animates and guides those structures.⁴⁶ This state-idea develops over time in conjunction with the state structure. Scott might argue that the actors within the B.C. state did not and could not have the knowledge that could only be gained from living in Sumas—knowledge such as how the lakeside commons was used, for instance. In an attempt to close such gaps, Scott argues, states seek to create simplified, ordered environments and societies. In so doing, they blanch out the complex knowledge required to make their reforming projects work.⁴⁷ Yet, as I have argued, the state was shaped as well by a state-idea, a form of liberalism that also worked against complexity in its call for an ordered landscape of individual, capitalist property-owners. This is important, because it suggests that the way that the state deals with social and environmental change can shift over time. Further, when we see that landowners embraced the same liberal ideas that animated the state, we can better understand the process by which state-directed change occurred, and the power of this process.

In the remainder of this essay I will demonstrate the utility of understanding the failures of the Sumas project as deriving from a contradiction and a limitation within the liberal discourse that structured the state. By March of 1920 the transformation of Sumas depended on three capacities—of the project engineers to smoothly rework the river into a more ordered form, of agricultural scientists to develop the lakelands, and of the lands themselves to sustain an abundant agriculture. What these tasks ended up demonstrating was the inability and unwillingness of engineers and scientists fully to comprehend the environmental complexity and disorderly nature of Sumas. A need for an understanding of environmental complexity, created by the desire of the state to manage environmental change at Sumas, was undercut by a desire for order. This was the contradiction in the liberal discourse. Liberal discourse also limited the ability of the state to comprehend complex relationships between farmers and their land. This limitation becomes apparent in considering the fallout from construction cost overruns, and will be discussed later.

The contract to build the diking works was granted to the Marsh-Bourne Construction Company in April of 1920. Initial work quickly revealed problems, arising from the combination of the annual flood cycle, the shaky financial condition of the company, and the technical constraints of the dredging equipment. The flood, which usually arrives in May, almost immediately halted construction. When work began again in September, it was clear that Marsh-Bourne did not have the financial resources to do the job. The company supplied only one dredge for work on the dikes, and when it tried to get another it was seized in Vancouver because the company was unable to make payments on it. The Land Settlement Board was forced to step in, buy the dredge and rent it to Marsh-Bourne. One more dredge and additional money and expertise became available through a new partnership between Marsh-Bourne and a Tacoma, Washington, firm. However, most of the new equipment arrived only shortly before May of 1921, when the Fraser once again flooded the area and brought work to halt.⁴⁸ Thus by June of 1921 little work on the Vedder Canal, the key to the system, had been completed.

What was being revealed by the summer of 1921 was a troubling gap between state planning for the Sumas area and local complexities. Marsh-Bourne and the LSB had stumbled on (for there is little evidence that they anticipated) a central environmental constraint in the work. Simply put, work on the project could not be done gradually. The structures had to be substantial enough by May to withstand the arrival of the flood waters. A sufficient number of machines had to be working at once in order to complete work in the seven months of guaranteed low water. Further, due to the timing of the Fraser floods, the work had to be carried out in winter. The need for speed meant that the older, cheaper, more common steam-powered dredges were not adequate, as their great weight made them slow to move about. The most modern and expensive equipment was needed—generally electric-powered dredges, which then also required the running of electric power lines to the site (see Figure 2).

Figure 2. Dike Construction on the Sumas Project.

Image D-09690, courtesy of Royal BC Museum, BC Archives.

The BC Archives titles this photograph simply "Sumas Land Reclamation Project." This is almost certainly the Vedder Canal under construction, though, given its size. The original purpose of the photograph is unclear; the bland title and the fact that this is one of several similar photographs of equipment and work on the project suggests it is intended simply to illustrate the construction process, rather than promote it. 1922.

Despite Oliver's insistence that the property owners were ultimately responsible for the project, the LSB reacted with disdain to their increasing concerns with the progress of the work and the possibility of mounting costs. Demonstrating the contradictions in the government's understanding of the relationship between individuals and the state, LSB Director Davies replied to property owners' requests for financial information and progress reports by declaring that he found it "impossible to treat these communications as being anything but the inquiry of certain rate payers within the Sumas District." He then suggested that their inquiries sullied the good name of the minister of Agriculture, who had their best interests at heart.⁴⁹ Despite the landowners' financial liability, they were to leave the experts in charge.

The owners safely ignored, the LSB, now in direct control of the work since Marsh-Bourne had declared bankruptcy in June, tried to close the gap between planning and environmental constraints by adding another dredge. However, the rivers stubbornly refused to cooperate. In October a storm caused a flash flood on the Vedder that wiped out partially completed dikes and bulkheads on the canal, severely damaged the East and West Vedder Dykes, and spread lumber and pilings across Sumas Lake and prairie. Then winter began to pound the project. In November a ten-day ice storm halted construction, and in December storms, floods, and a hard frost caused further damage to existing structures. A log jam in the river forced the current into the East Vedder Dyke, causing severe damage. The construction camp was left in a foot and a half of water, water was a foot deep over the B.C. Electric tracks, and eight hundred feet of track were wiped out. Supplies to the work camp, which came in by river, were cut off when the rivers froze.⁵⁰

In an attempt to make up for these delays the board put two more dredges on the project in early 1922. The spring flood arrived, however, before it was able to protect the west prairie section and the lake area through the completion of the Sumas River Dam. The engineers hoped that a temporary earth dam might do the trick. As the flood approached, the dredges were still working on the dikes, and engineers were not able to move the dredge "Tobin" to work on the dam until just before the water hit. In the end, despite much time and effort, the dam collapsed under the waters of the Fraser. The east prairie section fared better. In April the Vedder Canal was completed, sealing off the east prairie from the lake, while the Fraser dike was finished in time to shut it out from the annual flood. With the completion of the pumping plant, consisting of two sixty horsepower pumps inside a concrete bunker, the east prairie protection and drainage system was officially complete. Costs continued to mount, however, as it became clear that several dikes would have to be built either higher or longer than had originally been planned.⁵¹

The next year saw the completion of the reclamation works with the building of the Sumas River Dam, and the continuation of mounting costs due to poor planning and the constraints of the flood cycle. Construction on the dam began with no final plan and no cost estimate. The engineer finally settled on a concrete-and-earth structure. Four sluices would drain water from the west prairie area and the lake when the Fraser was low. When the river was in flood, the sluices could be closed and water pumped into the river. By June, when the water rose, the structure was ready, though extra funds had been spent on an additional shift in order to make this deadline. As the level of the Fraser climbed to 92 feet at the measuring point, it became clear that the structures were holding. The system worked. The only task now was to pump out Sumas Lake.⁵²

The lake still had a short time to live. The pumping was put off until winter, when low water would make the job easier. Meanwhile the project's chief engineer offered his assessment of the work. "To the average layman," he declared, "it is hardly conceivable that Sumas Prairie, after all the years of discussion, is at last reclaimed from the Fraser floods." But it was true. Now, "the natural elements are being controlled to the lasting benefit of mankind," and "there seems to be very little doubt regarding the ultimate draining of Sumas Lake and the productivity of the submerged lands."⁵³ Of the draining of the lake, at least, he was right. By May of 1924 only fifteen hundred acres of the lakelands were still under water, and that only to a depth of less than one foot. Though stories circulated afterward of sturgeon left behind in marshy areas and of ducks that still tried to land on the former waters, by June 26, 1924, the lake was gone (see Figures 3 and 4).⁵⁴ Ordering the Fraser and the lake to the satisfaction of the LSB's planners, however, had required greater effort and costs than had originally been anticipated. The process of turning the drained lands into productive farms would be similarly problematic. Watching all this, with their hands on their wallets, were the landowners.

Figure 3: Sumas Lake in Flood, Looking West from Vedder Mountain, 1916.

Image G-04791 courtesy of Royal BC Museum, BC Archives

The lake has broken through the dike (under the B.C. Electric Railway tracks) designed to contain it on its southern edge, though it has not flooded Yale Road (on the far left of the image). Compare to Map 1 (where the lake edge is shown north of the dike) and Figure 4, opposite. The B.C. Archives lists this image as being from the 1930s. A local museum says 1916 instead, a much more plausible date given that the lake clearly exists (see Matsqui-Sumas-Abbotsford Museum Society Archives, Image P-5005).

Figure 4. Sumas Reclaimed Lands, 1924, Looking North.

Image G-03773 courtesy of Royal BC Museaum, BC Archives

This photograph shows the recently reclaimed lands in a sweeping vista. It looks north from roughly the same spot in which Figure 3 looked west.

THE VALUE OF THE LAND

THE PRODUCTIVITY OF THE LAKELANDS remained an open question. The province now owned the lands outright. Upon the completion of the work, the Canadian government had sold the 12,200 acres of Railway Belt land lying under the lake waters to the province for $1.⁵⁵ Now, like the larger biophysical environment around them, the ecosystem of the lakelands needed to be reordered. However, if the building of the reclamation works called into question the ability of the engineers easily to order and control the capricious nature of the Fraser River, then the development of the lakelands questioned the abilities of agricultural scientists to create productive land economically. It also called into question the faith that everyone concerned with the project had had in the value of the soil of the lakelands. The initial soil report indicated that the land contained good concentrations of important minerals but was short on nitrogen. It was also somewhat sandy and so prone to erosion and drifting. To solve both its lack of nutrients and its physical problems, the land needed more humus (soil rich in organic matter), which could be obtained by growing legumes and leafy crops.⁵⁶

The board, however, had already decided to plant a small cash crop of oats in order to demonstrate the value of the lands (see Figure 5).⁵⁷ In July it began subdividing the cropped lands into fifty-two lots of about forty acres apiece.⁵⁸ It soon appeared that it may have rushed matters. The crops could not be planted until the soil dried out, and the resulting lateness of the planting and the nitrogen deficiency of the soil meant that the crop did poorly.⁵⁹ The lakelands turned out to need a complex system of draining, and the heart of the drainage system, the Sumas Lake Canal, was not completed until November. The LSB proceeded to build ditches around every farm lot in the lakelands, sometimes on three sides of the property.⁶⁰

Figure 5: Display of Grains from Reclaimed Area.

Image C-05569 courtesy of Royal BC Museum, BC Archives, by Stuart Thomson

A display of the first crop of grains produced by the Land Settlement Board in 1924. Note the progression from raw plant to more processed foods.

A few months after the lake was drained a larger and more intractable problem arose when a forest of willows grew up, seemingly overnight, in the damp lakebed. The trees spread far faster than LSB employees could cut them down, choking out the timothy and clover grasses that they had planted in order to improve the soil. By March of 1925, 2,500 acres of the former lake bed was covered in the trees, and they were crowding out 2,000 of the 3,000 acres of timothy that had been planted. The board attempted to strike back, plowing under 2,000 acres of trees.⁶¹

By the end of 1925 the board had spent $86,742.63 on the development of the lake lands—"on general work, ditching, buildings, bridges, dredge work, fencing lake, horses, [and] implements."⁶² By early 1926 the land was ready to sell, though the LSB calculated that it would need to charge $200 per acre in order to cover costs. The willow-covered land the board hoped to get rid of through leasing, offering reduced taxes and a supply of clover seed in return for the lessee's agreement to plow down and destroy the trees.⁶³

The board's sales pitch mixed agrarian idealism with a celebration of the tools of modernity. The land was portrayed as the gift of modern engineering techniques, the outcome of a seamless unity of nature and the machine. Advertisements stressed the fertility of the soil, the mild climate, the area's proximity to the city, its "modern conveniences" in the home and on the farm, and "last but not least, a beautiful setting of mountain, river and pastoral scenery." The heavy machinery of reclamation was the means by which this ideal nature was brought into being. Advertising pamphlets were full of photos of heavy equipment, whether the dredges building the Vedder Canal or the great electric motors of the pumping plants. Next to these were line drawings of genial, beaming farm wives holding chickens under their arms and petting goats. Photos showed cherries, chickens, and a record Holstein. In one image of the unity of farm life and modern conveniences, fields of grain wave in front of a Model-T, obscuring everything but the car's roof.⁶⁴

The Sumas Reclaimed Lands offered "40 acres! a healthy life! a comfortable living!" They promised "health, happiness and good living!" Growing villages and "hospitable people" could be found nearby. All this was the result of the intelligent manipulation of the environmental legacy using modern techniques: "Nature provided Southern British Columbia with the alluvial delta of the great Fraser River and reclamation by man added to the fertile valley lands."⁶⁵ Nature had prepared this place for the Sumas project. Here would be built an economically productive, healthy, and virtuous community, a modern settlement in the midst of the mountains, rivers, and plains built by nature (see Figure 6).

Figure 6: The Land of Sunshine and Mild Winters.

Image NWP97124S954, courtesy of Royal BC Museum, BC Archives

A promotional pamphlet designed to sell lands at Sumas folds out to reveal this panorama, which glories in both the technical achievement of the project and the lush agricultural landscape that reclamation has produced. It also trades on British Columbia's mild weather, which was always thought to be an attraction for settlers. 1927 or 1928 (dates inferred from context).

It was, however, not to be, or at least not right away. Despite the issue of a lavish advertising pamphlet in 1927, the area's new administrator, Bruce Dixon, decided that the land needed more work: the drainage system had to be completed, the willows continued to be a problem, and the soil continued to show a "marked deficiency in nitrogen content." Dixon put up barns, planted demonstration crops, and tried to address the nitrogen problem by growing clover. But a crop of 1,300 acres of grains failed completely. Dixon's one sale—637 acres to the Canadian Hop Company in 1926 and 1927—would lead to large-scale agricultural production, not to the sort of individual family farms pictured in advertising material.⁶⁶ Hop-picking would, however, employ Sto:lo workers on land that had once been theirs. In 1928 a new Conservative Party government replaced the Liberals and moved to get rid of the remaining lands at Sumas and to settle some of the scheme's debts. The 8,700 remaining unsold acres (out of 10,000 acres reclaimed) were contracted out to a real estate firm to sell at the reduced price of $125 an acre, with the promise that the LSB would build a house for each settler at a cost of up to $2,500. By the standards of what the LSB had hoped to get for these lands, it was a fire sale, and the lots moved. What is more, many of the lots sold were the sort of small farms that the board had hoped to sell, though over 2,500 acres of them were sold to two owners. By 1930 the Sumas project was largely settled.⁶⁷

A later accounting showed 7,000 acres sold at around $100 an acre.⁶⁸ Though $100 was a respectable price, it was considerably lower than the price needed to break even on the construction costs. The LSB's balance sheet on the project in 1932 showed sales of the lakelands amounting to $1,134,580.54, although $908,143.68 of this was still outstanding. This was about 60 percent of the initial cost estimate that the owners of land in the area had voted on in 1920. By 1932, though, the cost of extra expenditures on building the works and developing the lakelands had pushed the total outlay on the project to almost $6,000,000. Taking into account the value of unsold land and other unrealized assets, the project was saddled with a deficit of just over $3.5 million. This money would have to be made up by taxing the land owners of the east and west prairie regions.⁶⁹ It was they who would pay for the difficulties with ordering Sumas Prairie.

THE LIBERALS AND THE LANDOWNERS

BY 1930 THE CONSERVATIVE government had faced up to the inevitability of paying a substantial portion of the costs of the project. Throughout the 1920s the Liberals, however, had consistently refused to consider the notion. During construction they refused the landowners' requests for information on the progress and costs of the work. After the lake was drained, evidence grew that the project had suffered from a crippling lack of planning, and that planners' confidence in the potential of the lakelands quickly to become productive farms was unwarranted. Critics suggested that the Marsh-Bourne company had got the contract through its Liberal connections, and had never been properly capitalized.⁷⁰ Nevertheless, Oliver's government never wavered from its insistence that the owners must ultimately pay the bill. The lakelands, donated to the province by the federal government, would be the province's only contribution to the costs of the project.

The owners got word of the increase in costs in 1923, and the next year their worries were confirmed when the assessments on their property arrived. By 1925, with the lakelands yet to go on the market, the landowners notified the government that they could not hope to pay off their debts and continue farming. Their plight quickly became controversial, so much so that in order to calm the clamor of the press and the opposition the government convened a special sitting of the legislature's agriculture committee to consider the question of the assessment. Running over four days in December, 1925, the hearings featured Premier Oliver, Minister of Agriculture E.D. Barrow, longtime opposition critic J. W. Jones, as well as a number of the area landholders, all offering their views on the worth of the scheme and who should pay for it.

These hearings provide insight into the limitation in the liberal discourse that structured the relationship between the state and the environment in this period. Added to the limits on state experts' ability to understand and deal with environmental complexity was the inability or unwillingness of politicians to understand the complex ways in which the settler-farmers related to the place in which they lived and farmed. We must be cautious here. The politicians discussed were testifying at a hearing where they were forced to mount a defense of their actions. Still, the sort of defense they mounted suggests the limitations created by liberal discourse for achieving a sufficient understanding of the deeply complex relations between farmers and the land.

We can see that the roots of the state's unwillingness to embrace social complexity lay more in liberal discourse and less in the nature of the state when we see the ongoing agreement between the state and landowners over the value of a progressive, ordered countryside. There was little argument at the hearings over whether the reclamation project was a good or a bad thing. Both sides embraced the idea that technology could improve nature, and both valued an ideal of the independent farmer making a comfortable living. Therefore, all generally agreed that the project was a success, a great improvement. This agreement came despite the fact that the success of reclamation had little practical value to the landowners. As they made clear, they could not both pay for the project and make a profit. Yet the government rested its defense on the marvelous improvement that the dikes had wrought on the area. They were practiced at this latter argument, having previously used it to justify building the Vedder Canal, without first asking permission, through the Aylechootlook band's reserve.⁷¹ The government also reminded landowners that they had supported the project from the start and had agreed to pay for it.

The hearings opened with Premier John Oliver, who lost no time in establishing his position that the government had merely acted on behalf of the owners. In its role as Sumas Dyking Commissioners, the board was merely the agent of the owners. The owners, Oliver claimed, should have taken into account the unreliability of engineer's estimates and that previous diking schemes had run over budget. The government, he argued, could not be held responsible for a scheme initiated and controlled by the owners, no matter how little control they might have actually had in practice. "There is no connection between the Government and the Land Settlement Board as Dyking Commissioners," he explained bluntly.⁷² This might have come as something of a surprise to readers of the previous year's promotional pamphlet, which featured a quote from Minister of Agriculture E. D. Barrow declaring that, "The Sumas Reclamation Project has been carried out entirely by the Government of British Columbia."⁷³

The government's other argument was that the project had improved the prospects for farming in the region. With proper attention to modern, scientific techniques, farmers could do very well. Barrow told the owners that all they had to do was to grow the proper crops. He cited the example of his own farm in the area, where he had made a healthy return growing crops of grasses (timothy, clover, and hay). Yet Blackman, the owners' representative at the hearings, was having none of it. Barrow's story of agricultural success ran counter to the experience of pretty much every other farmer he had met, he stated. He noted that "we take it from the standpoint, not from the man who has a comfortable income, that can do anything he likes in the way of farming, but from the standpoint of a man who has to work for his living and go on the land and property, pay the purchase price of it in addition to the taxation."⁷⁴ One landowner, a Mr. McCallister, used scientific agriculture against the government. His farm had been the subject of a study by Professor Hare of the Department of Animal Husbandry at the University of British Columbia, which had compared farm yields with costs of production. Drawing on his knowledge of local prices and on Hare's data and Hare's methods for calculating costs, McCallister concluded that it cost him $20.65 to farm one acre, leaving him with $4.25 in profit per acre. Any charge over $5 an hour, which was the standard demand of the owners, would clearly be excessive. Yet the LSB had estimated that his annual assessment would be $9.44 per acre.⁷⁵

The government's insistent challenges to witnesses as to whether the project had not in fact cleared the way for the growing of more profitable crops resulted mostly in showing that the project had offered little benefit to the existing landowners at all. First, testimony revealed details of how, previous to the diking project, farmers had found ways to work around the flooding. One, a Mr. Samson, explained that he and others had "run cattle over the prairie [lowland] part of it and we farmed the higher lands on our place with grain." Cattle could be placed on the land in mid-March or early April until some time in June, taken off for about six weeks as the land flooded, and then put back on the land in mid-July or August. Ultimately Samson did find that he could not make a profit on the land—though as he rented it out presumably someone else could.⁷⁶ Still he showed how farmers could work around the floods. For the farmers who used the unowned borderlands of the lake as common pasture land, the dike was actually a setback.⁷⁷ Further, dairying, many farmers insisted, was the most profitable form of farming available in the Fraser Valley. McCallister, again, stated that he believed that, "the old bossy cow is the most profitable operation there is in British Columbia today ... there is no other business in the farming line that will produce as much return, in my estimation, as the dairy cow." "You don't admit," challenged the government's Mr. Souter, "that the improvement of the land would enable you to grow more profitable crops than keeping dairy cows?" "I don't admit that," McCallister responded.⁷⁸

Farmers complained of their lack of success with sugar beets, potatoes, and clover. Grains such as oats and barley could be cultivated, but the problem was making money from them. The grain buyers, "just pay you pretty near what they can squeeze out of you. I have only sold one car of grain this year, and the rest is in a granary, and ... I cannot sell it to anybody."⁷⁹ Although no one here mentioned it, the most likely cause of the Sumas farmers' inability to sell grain at reasonable prices was competition from the Canadian prairies, which were just coming into full production at this time. Like small farmers in Washington State studied by Richard White, there was no way that the Sumas farmers could hope to compete with the prairie grain factories and the publicly funded railways designed to deliver their harvests nationwide.⁸⁰

The owners' central problem was that the dikes had not improved their lands to an extent commensurate with their elevated taxes. Reclamation had made it harder, not easier, for them to make a living. And yet the project remained a success, in some ways, in their eyes. At one point, the owners' representative, Mr. Blackman, asked farmer Alexander McIver, "We are not complaining at this particular juncture about the completion of the work, which after all is a great benefit to that district?" "Yes," McIver replied, "and there is no question but it is an outstanding benefit to the country at large." McIver did point to one concrete benefit, namely, that the diking had relieved the mosquito problem in the area.⁸¹ Dairy farmer John Conover, asked if there was any water on his land, commented "Gosh, no. Now the scheme is a success, there is no doubt about that."⁸²

Though the dikes had not helped these and other farmers to make a living, they could nevertheless judge the scheme to have been a success because it had done what they had wanted it to do: create a properly ordered and civilized countryside. Cononver had wanted the flood waters off his land and the mosquitoes gone, as "we did not want to live forever in a place that was not fit to live in." Though he had not particularly needed it, he was "public-spirited enough to vote for it [the dike] so that the community would come under it, so that we could live and not live like Indians." Similarly, another farmer's objection to the taxes was that the income he was left with was "not a sufficient amount for a farmer to raise his family on and be respectable.... If we are going to put a batch of Chinamen down there in that area, maybe they can make it go but we cannot."⁸³ From this point of view, to the extent that the high taxes worked against the creation of a properly white countryside, they defeated the purpose of the diking.

In the end, the government reduced the assessments to $5.50 per acre, about what the owners had asked for. But this did not mean that they were now willing to pay any of the costs. The total amount that each landowner owed remained the same, and interest charges on the money they owed the government would continue to pile up. As well, any future moneys spent by the LSB, acting as Sumas Dyking Commissioners, would eventually come out of the pockets of the landowners.⁸⁴ Obviously such a plan had the potential to generate massive debt, and by 1928 the size of these debts forced the government to reduce the total amounts owing. Landowners' debts for the cost of construction were cut by as much as a third and the terms of payment were lengthened.⁸⁵

The 1928 plan by no means brought this tepid drama to a close. The pattern of government concessions failing to prevent the further accumulation of debt, such debt leading in turn to more concessions, continued. In 1946, yet another commission reconfigured the payments. Full payment was now expected by 1981.⁸⁶

The 1946 commission also revealed more clearly some unexpected environmental legacies of the project. Farmers noted that the soil was drier than it had ever been, so dry that they had to irrigate. Another stated that the scheme had not been worth the cost for the local landowners. It had dried out the soil and taken away the free pasture lands on the former lakebed. In the end, he declared, "the cost was too great for the land to stand." The major beneficiary, as symbolized by the Trans-Canada Highway cutting through the old lakebed, was not the local landowners but the province as a whole.⁸⁷

The Liberal government's failure to recognize this fact speaks of the way that liberalism motivated and shaped the project, and of the limits of that vision. The owners of the east and west prairies were not the only beneficiaries of the Sumas project. Yet in order to justify the project within a liberal order framework, Oliver and others had to find a group of individuals closely associated with the project's benefits and costs. The Oliver government was never willing to put out the funds necessary to fund a project that was, in reality, larger than the neighboring landowners. The position of the landowners further illustrates the contradictions created by liberalism. Landowners agreed on the success of a project that brought them few clear benefits. This agreement was rooted in a common vision of a liberal countryside of individual, white male farmers existing within a capitalist order. We cannot understand the genesis, progress or failures of the Sumas project without taking into account the liberal order and the sort of relationship with nature it counseled, a relationship that both the Liberals and the landowners embraced.

CONCLUSION

THE LIBERALS AND THE LANDOWNERS were not the only players in the Sumas region, nor was their agreement on the desirability of a liberal countryside the only story that can be told about this place. As Laura Cameron has shown, there are many openings to the story of Sumas, many positions from which to view the drainage and disappearance of the lake. The lake was remembered fondly and wistfully by some area residents. One man, Cameron recounts, told her that the draining of the lake was not of great importance, while keeping a painting of the lake next to the window that overlooked the reclaimed land.⁸⁸ Another story is that of the Sto:lo, who lost over half of the area of one of their small reserves to the Vedder Canal, along with the right to graze their cattle on the border of the lake. They also lost a place that figured into their teachings and so their understanding of the world. Barred from purchasing land in British Columbia, their most notable gain from the project was jobs as agricultural laborers on hop farms. Finally, as the hearings demonstrated, for all their agreement with state representatives on the desirability of an ordered, liberal countryside, white male landowners found their accommodation to the place they lived in radically altered, generally not for the good.

Was the lake, Cameron asks, an Eden, a great spawning ground, or an impediment to traffic? It was all three, she answers, and more.⁸⁹ I emphasize the agreement of the landowners with the state on the issue of a liberal countryside not because it is the only story. I have done so, partly, because it is a story that Cameron's history of the lake does not really tell, a further opening onto Sumas Lake. Second, and for the most part, I have done so because of what this agreement suggests about how the B.C. state understood its relationship to the environment and the people that depended on it. The state's inability to comprehend and deal with the ecological and social complexity of this place was key to the failures of the Sumas project. Yet this was not just a case of the state imposing its vision of society and nature on local people. A sizable chunk of the local population agreed with what the state was trying to accomplish, an agreement that came out of a shared vision of a liberal countryside.

Yet the relationship with nature this shared vision created contained a contradiction and a limitation, both major. The contradiction lay in the understanding of nature it encouraged. Engineers and agricultural scientists assumed that nature should be ordered. But the lake and its rivers were at their most natural in a disordered state. Surging, unpredictable floods choked with debris, changes in the channels through which rivers flowed, and a lake that grew and shrank throughout the year, were not signs of a problem.

And so when the engineers tried to contain the Fraser between consistent banks and keep the flood from spreading out across Sumas prairie, they were not working with natural processes, but instead were trying to change them into new forms. Allowances were not made for the sudden floods that washed out dikes or for the bad weather which halted work. No one adequately took into account the implications of the Fraser flood. First, the spring flood meant that most work had to take place in winter. More seriously, the flood cycle put a box around their efforts. The need to have work finished in time to withstand that year's floods meant that a large and expensive amount of equipment and labor had to be employed in order to finish dikes or dams in the seven-month window that the floods allowed. Finally, the assumption that the lakelands would immediately become superior farmland proved unduly optimistic. Instead of selling the lands and paying off construction costs, the LSB found itself instead spending money in a battle against growths of weed-like willows, building up the soil through the growing of cover crops, and constructing barns, houses, and fences in order to make the land more saleable.

The cost overruns were also attributable to other factors. The Marsh-Bourne construction company was guilty of sheer incompetence, and bad planning abounded. The project was also simply a victim of the difficulties of estimating the cost of complex engineering works. But at their heart, the inflated costs of the construction phase of the project were the result of an attempt to work against nature rather than within it.

The attempt to pay off these costs showed the limitation in the B.C. state's understanding of the relationship between itself, on the one hand, and, on the other, individuals and their complex relationships with place. While John Oliver's government was willing to take on social and natural engineering, they were liberal enough not to want to pay for it. Yet ultimately the landowners could not handle the burden. Like the joke about how a liberal will try to save someone from drowning (by throwing a ladder halfway out, and expecting the drowning person to do the rest), Oliver and Barrow constantly stressed the wonderful things the scheme had accomplished without considering whether it actually helped the farmers to pay for the dike.⁹⁰ The lakelands were, in the end, not as good as claimed, and sold for less than anticipated. Owners were left to pay off capital costs. And scientific agriculture, expected by Barrow and others to work together with scientific land reclamation to produce a superior agricultural crop such as grains, could not overcome the geography and state-funded infrastructure that made wheat farming precarious and dairy farming profitable in the Fraser Valley. Thus dairying, already a dominant industry in the area, remained the most profitable way to use the land. Another established crop, hops, continued to be important as well.

In the end, the draining of Sumas Lake obliterated a local accommodation with the lake in favor of a state-sponsored vision of progress. A particular ideal of progress interpenetrated both local experiences and the state to motivate a particular type of development. Contradictions in the structure of this ideal and how it related to this place and these people lay at the heart of the problems of the project.

At Sumas Prairie, the lake refuses to go away entirely. A major freshet on the Fraser River flooded the area in 1948, and it was inundated again in 1990. Memories of the lake persist as well, though now, of course, mainly in recorded form. The efforts to order the fitful nature of Sumas Prairie continue.

James Murton is assistant professor of history at Nipissing University, Ontario. He is the author of Creating a Modern Countryside: Liberalism and Land Resettlement in British Columbia (UBC , 2007), and is currently researching the effects of global trade, pre-World War II, on local food systems in Canada.

NOTES

This excerpt is reprinted with permission of the Publisher from Creating a Modern Countryside: Liberalism and Land Resettlement in British Columbia by James Murton © UBC Press 2007. All rights reserved by the Publisher."

^1.� Laura Cameron, Openings: A Meditation on History, Method, and Sumas Lake (Montreal & Kingston: McGill-Queen's University Press, 1997), 46; F. N. Sinclair, "A History of the Sumas Drainage and Dyking District," typescript, n.d., Library Collection, British Columbia Archives [hereafter BCA].

^2.� W. L. Blatchford, Secretary, "Minutes of the Sumas Dyking Meeting," November 24, 1919, vol. 50, file 8, GR-929 Land Settlement Board, BCA.

^3.� Robert A. J. McDonald, "Politics Before Parties: Modernity and Province-Building, 1871–1903," unpublished paper presented at the Beyond Hope Conference, May 2001, Kamloops, BC.

^4.� Ian McKay, "The Liberal Order Framework: A Prospectus for a Reconnaissance of Canadian History," Canadian Historical Review 81 (2000): 617–45; Tina Loo, Making Law, Order and Authority in British Columbia, 1821–1871 (Toronto: University of Toronto Press, 1994).

^5.� Barry Ferguson, Remaking Liberalism (Montreal & Kingston: McGill-Queen's University Press, 1993); James Murton, Creating a Modern Countryside: Liberalism and Land Resettlement in British Columbia (Vancouver: UBC Press, 2007). On new liberalism more generally, see James T. Kloppenberg, Uncertain Victory: Social Democracy and Progressivism in European and American Thought, 1870–1920 (New York: Oxford University Press, 1986); and Peter Weiler, The New Liberalism: Liberal Social Theory in Great Britain, 1889–1914 (New York: Garland Publishing, 1982). In the United States new liberal ideas were part of the larger discourse of progressive reform, fuelling reformers concerned with "social consciousness," "social duty," and the "common good," having less direct influence with the better-known muckrakers, trust-busters, evangelical social reformers, prohibitionists, and promoters of technocratic government. See Kloppenberg, Uncertain, 311; and Daniel T. Rodgers, "In Search of Progressivism," Reviews in American History 10 (1982): 123–26.

^6.� "Proceedings of the Agriculture Committee of the Legislative Assembly of the Province of British Columbia, Held at the Parliament Buildings, Dec 3rd, 1925," 9, 48(10), GR-929.

^7.� David Demeritt, "Visions of Agriculture in British Columbia," B.C. Studies 108 (1995–96), 40–47.

^8.� Cole Harris, The Resettlement of British Columbia (Vancouver: UBC Press, 1997), 219–49; Margaret Ormsby, British Columbia: A History (Macmillan of Canada, 1971), 398–99.

^9.� See Paul Sutter, "Reflections: What Can U.S. Environmental Historians Learn from Non-U.S. Environmental Historiography?" Environmental History 8 (2003): 5–10. By the state I mean not just the government but the constellation of institutions associated with the government and involved in the administration and regulation of society. In the case of this essay that includes the linked institutions of the cabinet and legislature of British Columbia, the Ministry of Agriculture and the Ministry of Lands, and various boards and quasi-state bodies. I also do not mean a state in the sense of the junior level of government in the U.S. federal system. In Canada these are, of course, called provinces. For a brief discussion of the difficulties inherent in defining the state, see Allan Greer and Ian Radforth, "Introduction," in Colonial Leviathan: State Formation in Mid-Nineteenth Century Canada, ed. Allan Greer and Ian Radforth (Toronto: University of Toronto Press, 1992), 9–11.

^10.� Thomas Dunlap, Saving America's Wildlife: Ecology and the American Mind (Princeton: Princeton University Press, 1988); Thomas R. Dunlap "Ecology, Culture, and Canadian National Parks Policy: Wolves, Elk, and Bison as a Case Study," in To See Ourselves / to Save Ourselves: Ecology and Culture in Canada, ed. Rowland Lorimer et al. (Montréal: Association for Canadian Studies, 1991), 139–47; Alan MacEachern, Natural Selections: National Parks in Atlantic Canada, 1935–1970 (Montreal & Kingston: McGill-Queen's University Press, 2001); George Warecki, Protecting Ontario's Wilderness: A History of Changing Ideas and Preservation Politics, 1927–1973 (New York: Peter Lang, 2000); John F. Reiger, American Sportsmen and the Origins of Conservation, 3rd. ed. (1975; reprint, Corvallis: Oregon State University Press, 2001); Tina Loo, "Making a Modern Wilderness: Conserving Wildlife in Twentieth-Century Canada," Canadian Historical Review 82 (2001): 92–121; Peter Gillis, Lost Initiatives: Canada's Forest Industries, Forest Policy, and Forest Conservation (Toronto: Greenwood Press, 1986); Janet Foster, Working for Wildlife: The Beginning of Preservation in Canada, 2nd ed. (Toronto: University of Toronto Press, 1998); Matthew Evenden, Fish Versus Power: An Environmental History of the Fraser River (New York: Cambridge University Press, 2004); Michel F. Girard, L'écologisme retrouvé: Essor Et déclin de la Commission de la Conservation du Canada (Ottawa: Les Presses de L'Université d'Ottawa, 1994).

^11.� Donald Worster, Rivers of Empire: Water, Aridity and the Growth of the American West (New York: Oxford University Press, 1985), 51–60, 282–84. Work in the Canadian staples tradition has argued that the country's reliance on the export of lightly processed commodities such as fur and timber has encouraged the creation of an interventionist state by requiring nonmarket institutions to set the conditions for stable accumulation. For an earlier example that works particularly well as environmental history (and whose author has taken up environmental history more recently), see H.V. Nelles, The Politics of Development: Forests, Mines & Hydro-Electric Power in Ontario, 1849–1941 (Toronto: Macmillan of Canada, 1974). A more recent example is Trevor J. Barnes et al., "Stormy Weather: Cyclones, Harold Innis, and Port Alberni, BC," Environment and Planning A 33 (2001), 2127–47.

^12.� James C. Scott, Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed (New Haven: Yale University Press, 1998), 2–7, 201, 221, 304–06. Nancy Langston similarly stresses the unwillingness to address complexity in explaining failures in both reclamation efforts and restoration efforts in the Malheur National Wildlife Refuge in Oregon: Nancy Langston, Where Land and Water Meet: Transforming a Western Landscape (Seattle: University of Washington Press, 2006).

^13.� Colin Duncan has argued that environmental historians interested in ideas need to move beyond ideas specifically about nature to consider how more general logics, like liberalism, encourage a particular form of engagement with nature. See Colin A. M. Duncan, "On Identifying a Sound Environmental Ethic in History: Prolegomena to Any Future Environmental History," Environmental History Review 15 (1991): 10. Richard White demonstrates how shifting ideas of technology and its relationship to society and environment influenced development in Washington state in The Organic Machine: The Remaking of the Columbia River (New York: Hill and Wang, 1995), particularly 30–81. For the confluence between left-wing and reforming political ideas and conservation, see Robert Gottlieb, Forcing the Spring: the Transformation of the American Environmental Movement (Washington, DC: Island Press, 1993), particularly 15–105; and Mike Davis, City of Quartz: Excavating the Future in Los Angeles (New York: Vintage Books, 1992), 376–82. Mark Fiege argues that irrigators aimed at an "irrigated Eden" that celebrated the combination of nature and the machine in his Irrigated Eden: The Making of an Agricultural Landscape in the American West (Seattle & London: University of Washington Press, 1999).

^14.� Cameron, Openings.

^15.� C. Cartwright, "Report on Sumas Reclamation Project," October 18, 1919, 48(5), GR-929. On the 1948 flood, see W. R. Derrick Sewell, Water Management and Floods in the Fraser River Basin (Chicago: Department of Geography, University of Chicago, 1965), 15–16.

^16.� F. N. Sinclair, "A History of the Sumas Drainage and Dyking District," typescript, n.d., Library Collection, BCA. On the history of the Fraser River, North America's largest undammed stream, see Evenden, Fish Versus Power.

^17.� "The Sumas Reclamation Project," pamphlet, 2, Library Collection, BCA, reprinted from Agricultural Journal, November, 1922. On the changeableness of nature and our tendency to nevertheless see it as timeless, see Daniel Botkin, Our Natural History: The Lessons of Lewis and Clark (New York: Grosset/Putnam, 1995).

^18.� Cartwright, "Report," Oct 18, 1919, 48(5), GR-929.

^19.� Barbara Beldam, "Sumas Prairie—A Mosaic of Memories," in Millicent A. Lindo, ed., Making History: An Anthology of British Columbia (Private, 1974), 32–33; Cameron, Openings, 23, 44, 50–51.

^20.� From Jody R. Woods, "Sumas Lake Transformations," in Keith Thor Carlson, ed., A Sto:lo Coast Salish Historical Atlas (Vancouver: Douglas and McIntyre, 2001), 104.

^21.� Cameron, Openings, 16–40, 50–1; Imbert Orchard, Floodland and Forest. Sound Heritage Series no. 37 (Victoria, B

creating order: the liberals, the landowners, and theDRAININGOF SUMAS LAKE, BRITISH COLUMBIA

JAMES MURTON