The Danger of Old Studies and “A study of the population of Indian ancestry living in Manitoba” (1959)
In 1956, the Liberal-Progressive government of Douglas Lloyd Campbell commissioned a study of Manitoba’s Indigenous population. Undertaken by the province’s Social and Economic Research Office, under the supervision of Jean H. Lagasse, the project was far-reaching — for its time — and covered off-reserve populations in both urban and rural areas.
Results were published in 1959, filling three volumes and totalling almost five hundred pages. The first consisted of the study’s official “findings,” while the second and third — authored, respectively, by researchers “W. E. Boek and J. K. Boek” and “Walter M. Hlady and B. Ralph Poston” — were tacked on as “appendices,” and consist of researchers’ initial reports to the department, including process notes, copies of interview forms, and more detailed survey results.
While it would be a stretch to say the study is widely available — copies exist in some Canadian libraries, mostly universities, and one digital copy has been archived, oddly enough, by the genealogy service Ancestry — portions of the first volume have been cited moderately widely.
It is, of course, possible to use old sources responsibly, citing or quoting out the useful information and skipping past the questionable stuff. In this case, most academics and researchers reference a series of charts included in the first volume — essentially, a list of peripheral Metis communities, with population estimates. Unlike the rest of the report, which is sweeping and obvious in its racism, this section appears to be relatively safe. And I will stress in their defense, that as an incomplete, but fairly long, list of small, otherwise poorly-documented places, assembled by a team of researchers, however biased, the charts are genuinely useful. Digging through the rest of the study, however, it becomes clear that the population counts that accompany the place names, and, by extension, all of the “Metis” statistics found in the report, are deeply questionable.
The report divides Manitoba’s population “of Indian ancestry” into two sections: “Indians” and “Metis.” Most of the information about “Indians” is not unique. The reports’ authors didn’t do the work themselves. Instead, they drew heavily on government documents provided by the Department of Indian Affairs. These numbers have their own problems, but the definitions underlying them are fairly straightforward: “Indians” are those governed by the Indian Act — a fairly stable group we would now normally call “Status First Nations.” As these communities were, and are, subject to heavy government surveillance, the names and numbers the department provided can be assumed, at least in aggregate, to be basically accurate. As a general rule, the researchers counted subjects as “Indians” only if they had Status, their name had been included on a band list in the past, or they self-reported as having entirely “Indian” ancestry.
In the “Metis” sections, things get stickier. By the authors’ own admission, their working definition resulted in a significant undercount: “Only a small proportion, perhaps less than ten percent of the Canadians who have some Indian and White ancestry are known as Metis.” (Vol.1, pg. 50) and “It is estimated that 80 percent of the people of Metis ancestry in Manitoba are not included in the study population.” (Vol.1, pg. 77)
How did they arrive at their definition?
Well, they asked a random group of white people, and then, ignoring much of the information they received, came up with a definition that fit their own assimilationist views.
No, seriously. That is actually what they did. They began by asking twenty unnamed white people for their own personal definitions of “a Metis or Half-Breed.” Ten of these answers were included on pages 56-57 of the first volume:
I) Actually, in our company we evade the issue by referring to all treaty Indians, non-treaty Indians, breeds, etc. as natives. But I would consider that anyone who has a white grandfather is a Half-Breed.
II) Any person of mixed white and Indian blood having not less than one-quarter Indian blood but that does not include Indians as defined in the Indian Act nor non-treaty Indians.
III) Half-Breeds are persons of Indian descent living in poor houses similar to those on the reserve, one-eighth being as far as I would go in searching for people of Indian background.
IV) Any full-blooded or half-blooded Indian who is not living as a white person. In this connection, the attitude of the white neighbors may force certain families to remain Half-Breeds longer than they would otherwise.
V) The term ‘Half-Breed’ refers broadly to persons of sufficient Indian blood to be barred from the white class and with enough white blood to be distinguished from Indians.
VI) A person who when he has money lives like a white man and when he is broke lives like an Indian.
VII) A Half-Breed is a person who has some degree of Indian blood plus an upbringing which combines factors of primitive living usually in conjunction with a hunting and fishing economy. This applies even when these people have almost embraced the white way of life. A person with a similar degree of Indian blood is accepted into the Canadian way of life only when he conforms to all general requirements of this society; the degree of blood is not too important.
VIII) People with Indian background who do menial tasks or are generally employed on part-time jobs. They usually live in poorer homes and have poorer standards of living. For example, I would not consider Mr. X as a Half-Breed because he is an office manager and a respected citizen in our community.
IX) I would call Half-Breed any Indian not in treaty as well as any of those persons who have some white blood in their background.
X) Half-Breeds are individuals who possess half Indian and half white blood, and those who live like the Indians. Metis are different in that they have their own ways and live differently from the Indian. (Vol.1, pg.56-57)
It’s impossible to say what the remaining ten answers were, because the authors excluded them from the report. Notably, not a single person appears to have provided a definition that prioritized how Metis people saw themselves. Working from these answers, and their own opinions, the researchers crafted their own definition, one that downplayed ancestry, ignored personal identification, and instead focused on subjects’ level of assimilation into white society: “To be classified as a Metis a person must have some Indian ancestry. The presence of Indian physical characteristics is not important although it helps to identify as Metis some persons who might otherwise not be detected. A second condition to being classified as a Metis is living under poor circumstances.” (Vol.1, pg.57)
Whether or not a person identified as “Metis” or “Half-Breed” was not only seen as unimportant, it was actively ignored. Under the heading “The Metis not Included in the Study,” the authors discuss the exclusion of members of L'Union Nationale Metisse St. Joseph du Manitoba, the oldest Metis organization in the province, despite the fact that they were, “still very proud of their Metis background” (Vol.1, pg. 77), on the basis that their living standards were too comparable to those of their white neighbours.
Other biases also affected the data collection process. First, the researchers conducted all of their interviews in English and didn’t employ any translators. In Winnipeg, they surveyed only a small portion of the North End: specifically the city blocks stereotypically associated with Indigenous people, where they expected to find the largest concentration of subjects. In small rural communities, they used unnamed “informants,” local residents who walked with them down city streets and pointed out the dwellings with “Half-Breed or Indian” residents. In larger centers, including “Brandon, Dauphin, Flin Flon, Selkirk and Greater Winnipeg,” they consulted city officials, identified the areas with the largest visible concentrations of Indigenous people, and conducted their survey only in those locations. Of the resulting population count, the report notes: “more elaborate survey techniques would no doubt have revealed more names.” (Vol.1, pg.58)
Laying out the issue in explicit terms makes the conceptual collapse at the study’s core obvious. The team had been tasked with studying the living conditions, educational attainment, employment, and social attitudes of Manitoba’s Indigenous population. If they’d wanted the results of their study to be accurate, they’d have needed to use definitions that didn’t take these factors into account. In their own estimation, a basic ethnic or racial definition, unbiased by the consideration of living conditions, would have produced vastly different data. Instead, they chose to pre-determine the study’s results by limiting it to the poorest and most visible concentrations of Indigenous people they could find. The resulting report, predictably, reinforced harmful stereotypes by defining “Indian” or “Metis” as synonymous with social conditions of poverty and dysfunction. Worse, it explicitly encouraged Indigenous people who didn’t want to be seen as poor, “backwards,” and problematic to identify as “white.”
While the study’s findings about the marginal locations in Manitoba where some First Nations and Metis families lived can and should be used as a jumping off point for further research, both the population counts and the overall results of the study — handling work, education, schooling, living conditions, and social attitudes — should be viewed as dangerously unrepresentative. The document is, perhaps, more useful as a snapshot of the attitudes of government and researchers, than as a study of Indigenous communities.
In general, when dealing with older material, it’s important to take the time to slow down and fully understand the terms and procedures being used. A solid majority of old studies and reports wouldn’t meet modern academic or ethical standards. I’ve found “proceed with caution” to be a pretty universally useful rule of thumb. This is a great example. It’s far too easy to skim through a report like this one for useful statistics, miss the underlying bias, and end up perpetuating racist misrepresentations. We can, and should, do better.
Works Cited:
A study of the population of Indian ancestry living in Manitoba (1959)