Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Prairie and Quebec Metis territoriality : "interstices territoriales" and the cartography of in-between… Rivard, Etienne 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2005-995402.pdf [ 49.96MB ]
JSON: 831-1.0092373.json
JSON-LD: 831-1.0092373-ld.json
RDF/XML (Pretty): 831-1.0092373-rdf.xml
RDF/JSON: 831-1.0092373-rdf.json
Turtle: 831-1.0092373-turtle.txt
N-Triples: 831-1.0092373-rdf-ntriples.txt
Original Record: 831-1.0092373-source.json
Full Text

Full Text

PRAIRIE A N D Q U E B E C M E T I S TERRITORIALITY: INTERSTICES  TERRITORIALES  A N D T H E C A R T O G R A P H Y OF I N - B E T W E E N IDENTITY  by  ETIENNE RrVARD  B . A . , Universite du Quebec a Montreal, 1996 M . S c , Universite du Quebec a Montreal, 1998  A THESIS' SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  in  T H E FACULTY OF GRADUATE STUDIES  (Department of Geography)  We accept this thesis as conforming to the required standard,  T^UNlVERSiTY^F BRITISH COLUMBIA  September 2004  © Etienne Rivard, 2004  Abstract This thesis is a historical and contemporary exploration of Prairie and Quebec Metis. The Metis, individuals of mixed Native and non-Native ancestry, have been constitutionally recognised as Aboriginal people(s) in Canada since 1982. They are the result of the numerous episodes of metissage that have occurred in the course of Canada's history. Metissage emerged early in the French Regime, as the intermingling of "Indian" and "white" blood was an inescapable outcome of the fur trade economy. In spite of this long history and recent official recognition, the mixed cultural origins of the Metis have challenged many aspects of Canadian society — its conception of aboriginality, its ethnic classifications and policies, and its conception of territorial integrity. On the other hand, the Metis also represent an opportunity for Canada to question its conceptions of aboriginality and to outline possible paths of reflection about the country's socio-political landscape. This thesis approaches these paths indirectly by exploring the historical importance of Metis geographies in the development of Canada. More specifically, it aims to identify the changing patterns of Metis territoriality — the changing Metis sense of identity and territory. M y historical exploration is largely based on an investigation of colonial maps, on which I have sought territorial markers (material, political, and symbolic) that identify the existence of metissage and the Metis. The visual nature of maps makes them influential territorial discourses and efficient means by which Metis geography and territoriality can be identified as well as the mental conceptions Canadians have of the country. This study of colonial maps is complemented by the analysis of Metis oral tradition as revealed by stories, individual accounts, songs, and place names. I also investigate the ways in which contemporary Metis conceive of history and the future, and how this affects (or supports) their self-identification. This contemporary inquiry is primarily based on Metis maps, Metis official web sites, and interviews I conducted with Metis living in different regions in the province of Quebec. Both historical and contemporary examinations reveal real regional distinctions between Prairie and Quebec Metis, although there have been significant social and familial connections between the two groups. Both Metis peoples also share common characteristics. The most important one is "in-betweenness," which appears to be a principal feature of Metis past and present territoriality.  Ill  Table of Contents Abstract  ii  Table of Contents  iii  List of Figures  vii  Acknowledgements Chapter 1: 1.1  ix  Introduction: Portraits of In-betweenness  1  Problematic, objectives and conceptual frame  3  1.1.1  Metis Geography(ies)  4  1.1.2  Metis territoriality(ies)  6  1.2  Methodology, complications and reaction  10  1.2.1  Maps: social representations and discourses about space  10  1.2.2  Metis "cartography."  12  1.2.3  The Canadian Metis today and methodological complement  13  1.3  Thesis outline  Chapter 2:  17  Le(s) metissage(s) en question: les reponses colon tales et cartographiques au  Quebec (1632-1857) et dans le Nord-Ouest (1785-1813) 2.1  Le metissage et la cartographie en Nouvelle-France (1632-1755)  2.1.1  20  Le metissage en Nouvelle-France: l'entre-deux d'un passage  21  Le Metis, une strategic d'implantation coloniale: la creation mentale d'un  peuple nouveau  22  23  2.1.2  2.2  19  Le Metis, une realite hors controle? L a cartographie du metissage?  27  Les ideologies socio-spatiales, l'identite mosai'que et les discours anti-metis..28  L'influence de 1'Autochtone et l'hybridite cartographique  Le metissage et la cartographie britannique du Nord-Ouest (1785-1814)  32 35  2.2.1  L a geographie metisse: Le Nord-Ouest, une terre de renaissance  36  2.2.2  L'ethnogenese et la formation des groupes «proto-metis»  38  2.2.3  L a cartographie du metissage: entre l'identite mosai'que et l'identite metisse  41  L'inscription cartographique des preconceptions ethniques: les transparences  ideologiques  42  L'inscription de l'hybridite cartographique: le Canadien, le proto-Metis et le  portage 2.3  43  Les effacements cartographiques dans the Province of Quebec (1764-1857) 2.3.1  Ambivalence (anti-)metisse  45  2.3.2  Les puretes et les clartes: la fonction de la carte et ses consequences  49  Le cloisonnement ethnique et la mosaique coloniale: la cartographie d'un  espace anti-metis  Conclusion  Chapter 3: 3.1  Cartography and the Vecu in the West —1819-1895  Cartographical inscription of Metis territoriality: from recognition to erasure 3.1.1  Material territorial marks: taking over the land  54 67 68 74  Transportation systems: Metis' tentacles in the Nord-Ouest  75  Land tenure: le paysage de la Riviere-Rouge  76  Socio-economic assimilation and the railroad: to "put on track" the Metis  3.1.2 3.1.3  77  Political territorial marks: delineating the Metis sense of common identity?  80  Surveys and battlefields: measures of a political and military seizure  81  Symbolic territorial marks: soul and sense of territory  82  Naming the land: making a "home."  82  The Queen sits on a pile of bones: a symbolic infiltration  84  Oral inscription of Metis territoriality: sketch of a primary writing  86  3.2.1  Metis songs and stories: the words that bound  89  Relating Metis experience of space: three areas of investigation  Shaping Metis experience of space  104  Overlapping spaces of in-betweenness  112  3.2.2  3.3  52  material marginality  3.2  49  L'occupation coloniale des marges: les «reductions» cartographiques de  l'Autochtone et de l'hybridite 2.4  45  Metis toponymic nomenclature: The land that knows a name  90  118  A n "oral history" of the Nord-Ouest: collective memory of place  119  A n "oral geography" of the Nord-Ouest: topography of mobility  121  A n in-between geography of the Nord-Ouest: naming distinctiveness  123  Concluding remarks  124  Chapter 4:  Metissage and Exclusion in the early 20 Century: The Metis, an "Official" th  Nameless Reality  132  4.1  Cartographical deletion: the Metis, a "past" people  133  4.2  Canada's Aboriginal policy: ambivalence, metissage, and spaceless Metis  138  4.3  Metis territoriality in the late 1960s: survival and renewal  143  4.3.1  The national decay and political (dis/re)organisation of the Metis  145  4.3.2  The "shack:" the material reality of a "Road Allowance people."  147  4.3.3  The shame, symbolic denial of pride, and identity modifications  148  4.3.4  The "no way through" mobility  150  4.4  Concluding remarks  Chapter 5:  (Re)Making of Ethnicity: Metis Territoriality & Cartography on the Prairies,  1960s - Present 5.1  156  From exclusion to (re)emergence: the Metis provincial reality  5.2  151  158  5.1.1  Preeminence over land: new Metis materiality  158  5.1.2  Spatial hierarchy of Metis organisations: political integration?  161  5.1.3  Pride, mobilisation and socio-spatial cohesion: redefining identity boundaries. 163  From (re)emergence to consolidation: the Metis land base  165  5.2.1  Material accentuations: mental conceptualisation  166  5.2.2  Self-(re)integration: delineating political reach  169  5.2.3  The historic Metis Nation homeland: mapping Metis geosymbols  172  5.2.4  In-between spaces of Metis distinctiveness: supremacy of territorial discourses. 175  5.2.5 5.3  Territory of the Metis Nation: social boundaries and exclusion  Concluding remarks  Chapter 6:  Les Metis au Quebec, une nouvelle perspective historique 6.1.1  199  199  Les perspectives historiques metisses: l'entre-deux spatio-temporel d'une identite clandestine  196  Un probleme historiographique et conceptuel: l'ethnicite, l'idenfite nationale et le nationalisme  6.1.2  182  La territorialite contemporaine des Metis au Quebec: la medianite comme  base territoriale d'une autochtonite repensee 6.1  176  De l'emergence du «pays metis»: une trame narrative originale  201 203  VI 6.2  Le baillon institutionnel anti-metis: les raisons du silence historique metis.. 206  Les Metis au Quebec, une perspective contemporaine  6.2.1  Les voies/voix individuelles: des cheminements identitaires  212  6.2.2  Les voies/voix collectives: la medianite, un carrefour identitaire  214  U n portrait de la presence materielle metisse: la carte, les effectifs et  1'organisation spatiale  214  «Qui representez-vous?»: la condition politique du fait metis  216  La medianite comme assise territoriale: la dimension symbolique de la  territorialite metisse 6.3  211  Conclusion  Chapter 7:  Conclusion: Aboriginality & the Genealogy of the Future.  220 222  226  vii L i s t of Figures  Figure 2.1  Sr de Champlain, Carte de la Nouvelle-France augmentee depuis la derniere, servant a la navigation, 1632  Figure 2.2  56  Guillaume Del'Isle, Carte du Canada ou de la Nouvelle France et des Decouvertes qui ont etefaites..., 1703  Figure 2.3  57  Philippe Buache, Carte physique des Terreins les plus eleves de la Partie Occiden du Canada..., 1742  58  le  Figure 2.4  S Robert de Vaugondy, Partie de I'Amerique septent? qui comprend la Nouvelle r  France ou le Canada, 1755 Figure 2.5  59  Peter Pond, Copy of a map presented to the Congres by Peter Pond, a native of Milford in the State of Connecticut..., 1785  Figure 2.6  60  Portion de Aaron Arrowsmith Map Exhibiting all the New Discoveries in the Interior Parts of North America..., 1802  Figure 2.7  61  Portion de David Thompson, Map of the North-West Territory of the Province of Canada..., 1813-1814  Figure 2.8  62  Thomas Kitchin, A New map of the Province of Quebec in North America..., 1764  Figure 2.9  63  Samuel Holland, A new map of the Province of Lower Canada describing all the Seigneuries, Townships..., 1829  Figure 2.10  Portion de Joseph Bouchette, Map of Lower Canada shewing the proposed land agencies and the townships district from the seigniories, 1857  Figure 2.11  65  Portion de Samuel Holland, A new map of the Province of Lower Canada describing all the Seigneuries, Townships..., 1829  Figure 3.1  64  66  Detail from S.J. Dawson's Plan Shewing the Region Explored by S.J Dawson and His Party..., 1859  126  Figure 3.2  Peter Fidler's Map of Red River District, 1819  127  Figure 3.3  A . L . Russell's Map of the Province of Manitoba, 1871  128  Figure 3.4  Detail from J.A.U. Baudry's Plan d'une partie de la Riviere Rouge..., 1871. .. 129  Figure 3.5  The Canada Bank Note co.'s Map of part ofNorth-West Territory shewing the Locality of the HALF-BREED  Rebellion..., 1895  130  Figure 3.6  W.H. Holland's Map of the Seat of Riel's Insurrection..., 1885  Figure 4.1  Detail from Canada — Department of the Interior's Carlton Sheet, Saskatchewan (West of Third Meridian), 1935  131  152  Vlll  Figure 4.2  C.S. Hammond & Co., Saskatchewan: Southern Part, 1946  Figure 4.3  Detail from Dent's Prairie Provinces, 1949  Figure 4.4  The Saskatchewan Golden Jubilee Committee's Historic Saskatchewan: A PictoralMap,  1955  153 154  155  Figure 5.1  Gabriel Dumont Institute's Manitoba Metis Federation, 1994  183  Figure 5.2  Gabriel Dumont Institute's Metis Nation of Alberta, 1994  184  Figure 5.3  Gabriel Dumont Institute's Metis Nation of Saskatchewan, 1994  185  Figure 5.4  Anonymous' Present MMF Regional Boundaries, 1978  186  Figure 5.5  Alberta - Intergovernmental and Aboriginal Affairs' Alberta Metis Settlements, August 1, 2002  187  Figure 5.6  Real Berard's Carte de I'Ouest au temps de Louis Goulet, 1976  188  Figure 5.7  Metis Association of Alberta's Cadotte Lake Settlement, 1980  189  Figure 5.8  Metis Association of Alberta's Trout Lake - Graham Lake Settlement, 1980. .. 190  Figure 5.9  Metis Nation - Saskatchewan's Metis Homeland, March 3, 2004  Figure 5.10  Real Berard's Selkirk's Grant and Successive Increases in the Area of Manitoba, 1975  Figure 5.11  191  192  Gabriel Dumont Institute's Canadian Atlas of Aboriginal Settlement (Cover Map), 1994.  193  Figure 5.12  K . Bigelow's Matrix's Map, 1994  194  Figure 5.13  Metis Association of Alberta's Metis Wintering Villages, 1980  195  Figure 6.1  Alliance Autochtone du Quebec, Communautes de VAlliance autochtone du Quebec inc., February 17, 2004  Figure 6.2  224  Nation Metis au Quebec, Carte geographique — communautes, June 12, 2003  225  Acknowledgements  Research is a territory of encounters, and each of them brings its own voice to the writing of a thesis. Aussi bien le dire, en raison de l'entremelement de toutes ces voix, cette these est profondement metisse.  There has been the voice of those who took an active part in the writing process. M y supervisor, Cole Harris, who provided generosity, patience, and, not the least, all the openmindness and intelligence that generate inspiration. M y committee — Matthew Evenden, Arthur J. Ray, and Graeme Wynn — whose comments, suggestions, and courage — they had to read one language that was not theirs (French) and another qui m'est second (anglais) — were accurate and extremely useful.  There has also been the voice of those who, although not directly or deeply involved in the writing process, have been of great support: the Fonds FCAR du Quebec and the University of British Columbia which provided financial assistance; Caroline Desbiens and Shaun Nolan who offered me their editing skills; the numerous scholars and Metis met on the Prairie and who generously offered their knowledge, advice, and time; ces Metis du Quebec qui ontbien voulu se plier au jeu et aux indiscretions de l'entrevue; mes parents et amis et leur appui indefectible; Real et Eva et leur chaleureuse hospitalite; Jean Morisset, parce qu'il m'a ouvert bien grandes les portes menant a l'univers metis.  Enfin, i l y a une voix plus douce que les autres et un soleil pour les jours sombres: par Belenos! tu l'as bien eu ta revanche Diane!  To these people, let me speak with one voice to say merci!  E. R.  1  Chapter 1:  Introduction: Portraits of In-betweenness.  la metis [est] cette ruse de Vintelligence au sens oil les anciens Grecs ientend Faite de roublardise, d'habilite, d'audace, d'observation, la metis est par defi V intelligence desfaibles devani les forts ou les puissants. Ses liens la rattach I'origine, a la dimensionprelogique, a I'intuition. — Fulvio Caccia, La Republique Metis.  Were metissage broadly defined, I would conceive of it as an ongoing process, both the result and the cause of the encounter and interaction of at least two worlds that mix with each other and that tend to create overlapping spaces, if not a "new order" marked by plurality. Generally defined in biological terms, metissage also involves social-cultural and intellectual dimensions, and is inscribed into discourse and ideology (Laplantine & Nouss, 1997, p. 7-11).  Complex versions of metissage had emerged in Canada as early as the French Regime, as the intermingling of "Indian" and "white" blood was a sine qua non of the fur trade economy. This dominant economy entailed many episodes of metissage, which were all similar in their motives (sexual, commercial, and political), while distinct because of the different people involved and the different contexts in which they took place. In la Nouvelle-France, colonial officials conceived of a "French" and "Catholic" Metis who would be the foundation of a new nation and would prevent officials from depopulating "la Vieille France" at the expense of a new one (Dickason, 1985; Perreault, 1982). Later, the Metis offered their own version of themselves in the Northwest in the late 19 century in Louis Riel's Republique metisse, which was intended th  to support a "supra-identity" that would recognise both Indian and Euro-Canadian constituents without choosing or rejecting either.  Thought about metissage has never ceased in Canada and has intensified over the last decades with the rise of post-colonial and, more broadly, post-modern studies (van Schendel, 1994, p. 103). Along with the concept of metissage culturel, post-modernists have been looking at concepts such as transculturalism, "mixed race," liminality, hybridity, and in-betweenness, all  2  of which have provided fresh ways of understanding identity within a contemporary context 1  marked by globalisation and the calling into question of national boundaries and identities.  This recent shift in the conception of metissage suggests the ambivalence about identity that faces any individual in a cross-cultural encounter, as well as the diversity, "unboundedness," and mobility of metis identity. Metissage is now understood to reject a purity of origin and to be a source of plurality. This plurality suggests that the roads linking metissage to metis people(s) are numerous, and that they do not necessarily converge. In this light, recent studies of the Metis in Canada have modified their investigations in two major ways. In the first place, many scholars have revisited the Metis past to offer revised versions of Metis identity, and to show, for instance, the heterogeneous nature of la Nation metisse at the Riviere Rouge colony or the instrumental and changing constitution of the identity of proto-Metis, who self-identified 2  sometimes as "Indian" and sometimes as "White" depending on the context or circumstances (Ens, 2001; St-Onge, 1985). Second, scholars have closely examined the evolution of the term "Metis" itself, revealing its shifting meaning through time and space and the use of parallel terms such as Bois-Brules and Half-breeds. From this evolution, some have pointed out social groups that have never been labelled "Metis" or have never defined themselves as such, but have developed separate senses of identity after experiencing a process of metissage, either biological or cultural. Martin Dunn provides many examples of Metis before the term was coined: Canayen, Coureur de Bois, Gens libres, Home Indian, Home Guard Cree, Labradorian, Malouidit, Pedlars, Rupertlander, Voyageur, and so forth (May 27, 2002a). Contemporary Metis also reflect such plurality and propose more than only one definition of who they are. I will further discuss that below.  It seems from this last enumeration that almost any human could claim metis identity. Obviously all do not do so. Metissage does not automatically lead to the genesis of Metis  The concept of hybridity is not new, as De Grandis and Bernd argue (2000), and appeared in various formulations as early as the 19 century. This recalls the argument of those who do not see an absolute epistemological rupture between post-modern and modern (Canclini, 1995, p. 9; Harvey, 1989), or even between the latter and traditional (Somers & Gibson, 1994). The term "proto-Metis" is problematic. The word "proto" may suggest that these Metis were only a "phase" in the creation of Red River Metis. This is a position that would be rejected by many contemporary Metis. On the contrary, I use this term — for lack of a better one — to note the existence (and persistence) of other groups of Metis outside Red River. One may consider the Red River Metis as proto-Metis as well. 1  th  2  3  peoples. In spite of the metissage that occurred in Quebec and of what van Schendel calls the 3  minority figure of the metis in Quebec identity formation — historically played by the coureur de bois and the voyageur — today Franco-Quebecois (the first Canadiens) do not define themselves in Aboriginal terms (1994). Similarly, the metissage processes that Canadian First Nations have faced for centuries, has not prevented Native people from defining themselves in opposition to the dominant society, while reducing metissage to an inconsequential phenomenon  Frederik Barth (1969) offers a meaningful theoretical framework for understanding the dynamic evolution of ethnicity, the concept of collective identity. For him, what defines ethnic groups has less to do with the development of a specific set of cultural traits or visible cultural symbols (language for example), than with ascription and/or self-ascription, with ethnic categorisation. Ascription alone — whatever the demographic, cultural and political changes — allows the persistence of social or ethnic boundaries. The perception of ethnic boundary maintenance, based on (self)identification, appears to be a relevant point of departure for exploring Metis identity.  1.1  Problematic, objectives and conceptual frame.  Canada's official definition of Metis people(s) is at odds with post-modern ambivalence, for it draws a specific social category for the Metis. The country recognises the Metis as 4  Aboriginal peoples, a recognition that is officially entrenched in the Constitutional Act of 1982. 5  As a result, people who define themselves today as Metis, and associate themselves with original Metis cultures, are spread all over the country, from Labrador to British Columbia, well beyond the prairie area traditionally recognised as Metis (Royal Commission on Aboriginal People, 1996, p. 199-270). One must consider Metis political activism as a major explanation for the persistence of a Metis category in Canada. For example, the Native Council of Canada — a  According to the Quebec botanist and ethnologist Jacques Rousseau, 40% of French people in North America could find at least one Native ancestor in their genealogy (Smith, 1979, p. 116, quoted by Perreault, 1982, p. 86). There is no equivalent in the United States as "Metis" is neither an official nor an academic category. There is some opposition to this general aggreement. For Thomas Flanagan (1985; 1991a), Metis did and do not qualify as Aboriginal people for they were treated differently from the Natives and received scrip to extinguish any Aboriginal rights they could had received from their Indian forebears. This confirms the fact, however, that the 5  4  national organisation representing Metis and non-status Indians, now the Congress of Aboriginal Peoples — had a great deal to do with the entrenchment of Metis in the 1982 Constitutional Act.  For my purposes, thus, I consider Metis any contemporary Canadian people of mixed Indian and non-Indian ancestry who self-identify as Metis. Ascription is central to this broad definition, which, I have to admit, is not everywhere accepted (Sawchuk, 1998, p. 14 & 19), an issue that I will address again in this chapter and will discuss at length from chapter 5 to the end.  1.1.1  M e t i s Geography(ies).  Why Metis are considered Aboriginal is a question that is embedded within the history and geography of Canada. Beginning in the late 17th century, the fur trade became a dominant socio-economical activity in Canada (Harris, 1987, p. 87), and largely took place at the margins of the reach of colonial settlements: this obliged Euro-Canadians to confront a land about which they knew little, and which belonged to Natives (Innis, 1999). Samuel de Champlain realised this early in the 17 century when he was stopped by the Algonkin chief Tessouat at lie des th  Allumettes on the Outaouais River, and prevented from heading further west. In the 18 and 19 th  th  centuries, factors and traders would experience similar constraints in Rupert's Land and on the prairie. As long as the fur trade lasted, so did the Native influence.  There is no need to review how power gradually shifted from Indian to Euro-Canadian hands. As Euro-Canadians settled in the lands largely abandoned by the fur trade, Native influence had declined proportionally. Even earlier, the trade had greatly modified Native populations and territorial patterns, as many authors, such as Denys Delage (1985) and Arthur J. Ray (1998), have shown. However, this shifting of control over land and resources (human and physical) was gradual and had been preceded by a relatively long period of interdependence — a middle ground to use an expression popularised by Richard White (1991) — perfect soil for metissage to create in-between spaces or, in other words, Metis geographies. A "Metis geography" would be, then, a spatial and social structure resulting from the encounter of distinct  Metis were recognised as a distinct "Native" people before 1982. Hence my general use of the term "Aboriginal" throughout the thesis to identify both pre-1982 and post-1982 Metis alike.  5  cultures (two or more) and the mutual and more or less symmetrical interpenetration of the basic components of these cultures: spatial behaviour, resource planning, or patterns of exchange. In Nouvelle-France as in Rupert's Land, such a Metis geography was the result of the overlap and interconnection of many geographies, which all had deeply structured spaces: Indian and colonial geographies, of course, but also those of the fur trade, and to a certain extent, of missionaries — primarily Jesuits for la Nouvelle-France and Oblates for Rupert's Land (Sawchuk, 1978, p. 18). What had made Metis geographies so influential is that they had become more than the result of such mixing; they had also become central elements of such interaction. The development of cart trails by the Prairie Metis is one example among other as these trails were essential to pemmican production (a principal source of food for the fur trade companies' employees in the West), which was mainly a Metis business in the mid-1800s, and "rivalled the rivers as transportation routes" (Sealey & Lussier, 1975, p. 21).  Despite the official recognition of the Metis as Aboriginal people(s) in Canada, it would be quite misleading to assume that such recognition settles the problems raised by the integration of the Metis within Canadian society. Actually, this recognition is quite equivocal, and leaves the question of Metis definition undetermined. This ambiguous situation — which stems primarily from the nature of Metis identity based as it is on the intermingling of two cultures — challenges many aspects of Canadian society as a whole: its conception of aboriginality; its ethnic classifications and policies; and its conception of territorial integrity. On the one hand, these challenges feed a certain fear in Canada that issues bearing on Aboriginal rights will never end, and that Metis claims will proliferate, for the metissage between Indians and mainstream society 6  seems to be an on-going process. On the other hand, Metis reality reminds Canadian mainstream society and First Nations of the importance of metissage in the making of the country and in the structuring of their relationship today and in the future. In other words, Metis also represent an opportunity for Canada to question its actual form, made of social(-racial) categories sealed under "purity," and to outline possible paths of reflection towards the renewal of its socialpolitical landscape.  My fundamental objective here is to approach these paths indirectly by exploring the historical importance of Metis geographies in the development of Canada.  6  From Arthur J. Ray, ad verbatim, 2001.  6  1.1.2  Metis territoriality(ies).  There are many ways of studying Metis geographies. Some scholars have focused on the role of the Metis within the fur trade economy in the Northwest during the 19 century and on th  the specific spatial patterns related to it. However, such studies do not necessarily place identity at the centre of investigation, as does the notion of territoriality. I define territoriality as a social and historical process, which tends to appropriate and delimit space using social markers (political, material and symbolic) that make sense for a specific group of people. Social markers and the spatial limits they draw are recognisable by each group member and, therefore, are also markers of identity (DiMeo, 1991, p. 129, 145 and 184; Sack, 1986, p. 26; Sbrlin, 1999, p. 109). As an on-going reality, territoriality is both the process by which a people appropriate space and create territory through their socio-cultural markers, and the process by which they redefine, at least partly, their identity and sense of belonging in relation to that territory. In other words, whether evenly or not, territory and identity can, simultaneously or asynchronously, affect each other (Raffestin, 1983).  Therefore, I define my objective in more focussed and concrete terms. I want to establish the changing patterns of Metis territoriality — the changing Metis sense of identity and territory. Those who call themselves Metis today are my starting point and I will try to establish links between these contemporary Canadian Metis groups and past Metis geographies. Given the importance of "shared experience" in collective identity definition (Marden, 1997, p. 55; Rokkan 6 Urwin, 1983, p. 67), it is relevant to consider how these contemporary identities are grounded (or how they "ground" themselves ) in the past.  This investigation of Metis territoriality fills a gap in the scholarly work on the Metis, mainly conducted by historians and ethnographers. For these scholars territoriality is at most partially expressed or defined, and often is unconsidered. If territoriality is an important notion in  According to the RCAP: "Individual identity is a matter of personal choice. One can identify with any people or nation, whether or not there is an objective reason for doing so and whether or not that people or nation agrees. For acceptance of that identification, however, it is necessary to win the approval of the people or nation with which one identifies. It would be inappropriate for anyone outside that nation to intervene. Therefore, when a government 7  7  geography (Sack, 1986), very few geographers have studied Metis people. Territoriality is best considered, albeit incompletely, through studies of the land and scrips question (Flanagan, 1991; Sprague, 1980a & 1980b; St-Onge, 1985) and through ethnohistorical studies. The debate encompassing the land and scrips controversy, which is still central for Prairie Metis land claims (especially in Saskatchewan and in Manitoba), focuses on a fundamental question: "Had the Metis been fairly compensated for the land they lost following the establishment of the Province of Manitoba and its annexation to Canada in 1870 through the land or money scrip system, or, rather, had they been deceived in this matter?" Whatever the answer to this question, territoriality would appear to be a key notion, since one can hardly talk about land reserved to "Metis" without referring to individuals who claim this identity and feel a sense of belonging to this land. Yet, territoriality, as situated within the debate, appears to be secondary and is reduced to its political dimension as the land base of the Metis Nation, with little regard to its material and symbolic dimensions. A similar comment can be made about ethnohistorical studies. In exploring the mechanisms of boundary making (social and geographical) that give birth to Metis ethnogenesis within the fur trade economy, ethnohistorians treat territoriality only in terms of its bounded materiality (Ens, 2001).  8  In order to facilitate the reading of territorial marks, I propose to group them into three different if interacting socio-cultural dimensions: political markers, which entail the collective awareness of space appropriation (its more or less formal and institutionalised organisation), a vouloir-vivre ensemble as Guillorel (1999, p. 64) would define it; material markers, which reveal the use of space, sometimes as a result of survival and adaptive strategies (land exploitation, economic activities, settlement patterns, and infrastructures); and symbolic markers, which organise space in terms of representations, ideologies, and norms.  9  whishes to know a nation's membership for the purpose of engaging in a nation-to-nation negotiations, it can legitimately consider only two criteria: self-identification and acceptance by the nation" (p. 202). 1 must admit that they could hardly do otherwise. Their primary source of information comes from the Hudson's Bay Company (HBC), which was, of course, economically focused. This selection is not intended to be exhaustive. Other important types of transactions between individuals, such as social class and gender transactions, were set aside as generally irrelevant to ethnicity; whatever the differences between male and female identities are, they are subordinate to what defines ethnic group identity. The only exception to that would be in an ethnogenesis or ethnic change context. Social classes within the fur trade economy could have played a major role in the rise of Metis identity as much as the importance of Native women (Brown, 1980; van Kirk, 1980). That said, my goal is not to describe Metis ethnogenesis or ethnic changes, but rather to seek indications of Metis existence on the maps. 8  9  8  A question remains: how can one recognise social markers that are exclusively Metis? Metis "in-betweenness" helps elucidate this question as it is a key component of Metis identity. The Royal Commission on Aboriginal Peoples (hereafter RCAP) describes the Metis as being neither "Indian" nor "European," but partly both (1996, p. 220). Earlier, Louis Riel proposed his own definition:  C'est vrai que notre origine sauvage est humble, mais il est juste que nous honorions nos meres aussi bien que nos peres. Pourquoi nous occuperions-nous a quel degre de melange nouspossedons le sang europeen et le sang indien? Pourpeu que nous ayons de I'un ou de Vautre, la reconnaissance et I'amour filial ne nous font-ils pas une loi de dire: "Nous  sommes Metis" (Ouimet,1889, p.78). An outsider observer, Henry Youle Hind, gave a similar, albeit Eurocentric, depiction: "The half-breeds of the north-west are a race endowed with some remarkable qualities, which" ' they derive in great part from their Indian descent, but softened and improved by the admixture of the European element" (1860, vol. 1, p. 178-179). Riel and Hind's definitions support the accepted ideas of that time about "blood" and race or about the dichotomy between "primitive" and "civilised" people, but once "translated" (blood is "culture" in Riel's definition) they reveal the plurality and ambivalence related to metis experience. In-betweenness imposes itself, thus, as the theoretical and practical base of Metis territoriality, and affects the character of Metis territorial markers.  Recent post-modern theory has mainly treated in-betweenness (along with hybridity, and liminal spaces) as a means of displacing hegemonic discourses about identity (Johnston et al., 2000, p. 364). The two Metis uprisings (1870 and 1885) in the North-West would be suitable examples of such resistance, for behind them was another version of the socio-cultural categorisation in the West, a more encompassing vision fitting both Indian and Euro-Canadian, and an alternative to the Canadian binary of primitive and civilised people (van Schendel, 1994, p. 113). However, one must wonder if the intention behind these uprisings was to offer alternatives, or if it was, more simply, to seek recognition for the people they thought they were, that is Metis. Whether the Metis saw themselves as "resistants" is a fundamental question here, since Metis views are my point of departure. Katharine Mitchell argues that the emphasis on resistance by cultural theorists is too abstract and excludes material and power relationships from the equation. For her, liminal spaces do not always challenge dominant hegemonies, and occasionally can accentuate them (1997). While the coureurs de bois did confront established  9  colonial conception of metissage and feed the fear that a metis power would arise at the 10  margins of colonial influence, their role as middlemen in Native lands had greatly benefited colonial mercantilist purposes in North America (Dickason, 1985; Perreault, 1982, p. 90 & 92; White, 1991, p. 69). On the other side, the Native also made a great use of the intermediate nature of the coureurs de bois, appointing a few of them Chiefs to promote their geopolitical interests (Hebert, 1984, p. 432; White, 1991, p. 213).  And yet, in-betweenness has been a site of resistance. Challenged in the core of their social ideology and in their geographical vision of control over land (resources and people), colonial authorities opposed the infiltration of the "other" , the Savage, into European social definitions. As Francois Laplantine and Alexis Nouss point out, metissage goes along with its opposite, antimetissage: "Pas de metissage sans antimetissage [...]: une culture metisse doit sans doute souffrir ses contradictions pour affirmer sa nature. [...] non pas resoudre ses contradictions mais vivre avec s'avere la meilleure maniere de les depasser" (1997, p. 57-58). These authors add:  Ainsi done, dans I'antimetissage — qui est la fascination de I'homogene, de Videntique et de la plenitude ontologique —, on rejette, on reprouve, on rompt. Mais que rejette-t-on au juste ? On rejette ce qui est vecu comme insupportable: I 'alterite, le doute sur sa propre identite comme sur la realite {idem, p. 86-87).  In this light, one would say that it was under the seal of antimetissage that the Dominion of Canada (notably the provinces of Ontario and Quebec) refused in the late 19 century to examine th  Louis Riel's Republique metisse, which proposed to exceed French-Canadim/Anglo-Canadiens and savage/civilised dichotomies and to affirm itself as the model of a new Canadian Nation (Caccia, 1997, p. 26; Morisset, 1983a, p. 280; van Schendel, 1994, p. 113). It seems that each 11  episode of metissage confronted a colonial reaction that aimed to diminish (or wipe out) its influence.  Again, whether or not these coureurs de bois bore Indian genes is not of great importance here. What matters is that they seemed to be culturally metisses. Their "indianised" behaviour — at least once they were out of the SaintLawrence valley's colonial influence — represented a high potential for the genesis of intercultural collectivities at the margins of the colonial settlement areas (Peterson, 1978 & 1985). It is obvious that Confederal authorities did not think they were resisting an established reality. The events of 1870 and 1885 were not called "uprisings" at thefirstplace, but rather "rebellions," supposing opposition to Confederation's established order. 1U  11  10  These episodes of resistance remain the best indicators of the importance of Metis geographies. Alone, Metis geographies hardly seem determinant in the course of Canadian history: after all, Metis peoples can still be seen as minorities within mainstream society (and even, as revealed by the report of the RCAP, within Aboriginal societies). However, placed within the context from which they emerged, these episodes of metissage, to which the colonial establishment reacted so strongly, expose themselves as keystones of Canadian history. "There can hardly be a more revealing way to decipher the basic nature of this state [Canada] than to examine its political and symbolic behaviour with respect to the issue of metissage, whether somatic or cultural" (Morisset, 1983a, p. 281).  1.2  Methodology, complications and reaction.  In order to understand the Canadian social-political landscape, to outline significant episodes of metissage-antimetissage,  and to explore Metis territorialities, I will focus my  analysis on historical and contemporary maps. I will scrutinise these maps for evidence of Metis social and territorial markers, be they political (boundaries), material (infrastructures, settlement patterns, and economic activities) or symbolic (religious symbols, or place names).  1.2.1  Maps: social representations and discourses about space.  Maps are spatial discourses that provide both the means (technical and cultural) and the motives (a territorial claim for instance) to articulate a geographical knowledge — a peculiar set of selected and hierarchically organised geographical facts. The discursive nature of maps indicates that knowledge is shaped by the "truth," representations, practices, codes, and rules recognised and authorised by the society within which the cartographer lives. Since they are socially and contextually constructed documents — the expressions (often unconscious) of a society's culture, norms, and ideologies — maps hold social markers that comprise territoriality (Harley, 1989a, p. 1-20).  11  Most of the maps I am working with were made by officials for colonial governments. Metis peoples did not produce their own maps before the 1970s. However, until the late 19  lh  century, these colonial maps depended heavily on the geographical information provided by Natives (Harley, 1992, p. 526; Ruggles, 1991, p.61-68) and Metis (Hind, 1860, vol. 1, p. 181): "During our residence in the settlement [at the Riviere Rouge] and on our exploratory excursions I employed many of the half-breeds, and was thrown of necessity so much among them [...]" (Dawson, 1859, p. 24). Because of the Aboriginal contribution to colonial maps, one can identify numerous social markers of Native and Metis peoples on these maps.  Despite their colonial bias, early maps o f Canada offer some advantages for my purpose. The first is visualisation. Maps, as opposed to other sources of colonial records (official documents, and correspondence letters), reveal their message more directly^ because they are made of images rather than words. Using visual perception studies, Jacques Bertin (1967) showed that an image is instantaneously envisioned as a whole. The visual nature of maps makes them influential territorial discourses and efficient means by which Metis geography and territoriality can affect the mental conception Canadians have o f the country (Harley, 1987, p. 4). Related to this is the second asset of these maps. Because maps convey images of Metis territoriality, although their function is to promote a Euro-Canadian geographical vision often at odds with Metis fact, they confirm metissage and Metis as inescapable realities of Canadian geography.  Maps offer another advantage for exploring metis territoriality. Produced at many spatial scales, they allow me to vary the angle and intensity by which I examine Metis territorialities. The relevance of scale differentiation has arisen recently in interdisciplinary studies related to Metis (Brown, 1980, p. xx; Brown, 2001, p. 60). Historians and ethnologists first tried to understand Metis identity in a broad sense and at a variety of scales — provincial, regional, national, and even continental (Brown, 1980; Howard, 1952; Giraud, 1945). Since the mid1980s, they have tended to work on a local scale and to study particular communities in detail (Payment, 1983, 157 p.; St-Onge, 1985, p. 149-173). More recently, some authors have come even closer to the subject by exploring Metis identity through family histories (Devine, 2001, p. 131; Podruchny, 2002; Thistle, 1997). However, scale interaction has never been a major  12 component of these studies, as scholars have limited themselves to explore one scale at a time. B y focusing as much on local, regional and national scales, my work is an attempt to fill this gap.  In selecting the maps analysed here, I did not focus on original and little known maps. A s the context of map production is so important for the analysis, it seemed better to look for maps that have already been discussed or analysed before. The orignality of my research is to identify the importance of maps in reflecting and shaping Metis geographies and territorialities. This importance may be easier to observe with maps published widely for they were more likely to reach contemporary readers and to affect their perception than were marginal maps. Another advantage is that these maps can be easily found today in various locations, including Internet for some of them. I have consulted maps in four archival institutions: the National Archives of Canada (NAC, Ottawa), Hudson's B a y Company Archives (HBCA, Winnipeg), la Bibliotheque nationale du Quebec (Montreal), and la Bibliotheque nationale de France (Paris). In spite of 12  the considerable number of maps held by each of these institutions, I am confident that my map selection is fairly representative of the periods and regions examined. The NAC were particularly useful in selecting maps. The archives have most of their map collection on microforms. These microforms make the map collection easier to consult — maps are visible, although not in detail — and help identify cartographical trends.  1.2.2  13  Metis "cartography."  One can hardly identify a Metis map produced before the 1970s. O f course, Metis had previously produced sketch maps, as Hind noted in the 1850s (1860, vol. 1, p. 143). However, these maps, like most orally communicated Native and Metis geographical information, were integrated into colonial maps. In the process, all their information was "translated" into colonial (intelligible) knowledge and cartographic conventions and became part of a colonial spatial  The latter institution was visited during my master's research in 1996-1997, which focused on la NouvelleFrance's cartography. In spite of the time I spent consulting maps in the HBCA, I ended up chosing only one map from there. The main reason for this was financial and practical. I found it easier and less expansive to order copies of maps from the NAC. I also considered that the maps found in the NAC were as representative as those in the HBCA. The Aaron Arrowsmith's map of 1802 (examined in chapter 2), for example, was found in NAC, although it was produced from the HBC's cartographic material kept in London, England. 12  13  13 discourse (Belyea, 1992, p. 270). This suggests both the care with which colonial maps have to be analysed, and the limits o f what map analysis can reveal.  Metis historical territoriality can only be partially made visible by official maps, and map analysis needs to be complemented by other investigations of the changing spatial-temporal context within which the Metis lived and understood themselves. This leads me to explore many oral components of Metis culture: narratives and oral histories, songs, and toponyms, all possible expressions o f Metis mental cartography/geography. They can be found in many sources: Oblates' letters and reports, scientific explorations (Palliser, H i n d , and Dawson), toponymic gazetteers and dictionaries, and Metis accounts.  1.2.3  The Canadian Metis today and methodological complement.  Although my description of the Metis i n Canada may appear homogeneous, I w i l l have to investigate very real regional distinctions. It is also clear that all Canadian Metis do not share the same social markers. These differences occurred at different scales: between the Red River Metis and the other northern and western communities for instance (Devine, 2001; Foster, 1985; Ray, 1982; Thistle, 1997). Henry Youle H i n d even observed such diversity among the Metis buffalo hunters: There are two distinct bands of buffalo hunters, one being those of Red River, the other of the White Horse Plain, on the Assiniboine. Formerly these bands were united, but, owing to a difference which sprung up between them, they now maintain a separate organisation, and proceed to different hunting-grounds (1860, vol. 1, p. 179). Contemporary Metis are diverse as well, a diversity that complicates to a great extent their integration within Canadian socio-cultural and spatial structures and that has barely been acknowledged by governments (Sawchuck, 2001). A question then naturally arises: "Who are the Metis"?  14  According to the Metis National Council (MNC), which mainly represents Prairie  Many authors tried to answer this question over the years: Bell (1989), Boisvert & Turnbull (1985), Brown & Peterson (1985), Foster (1985) RCAP (1996), etc.  14  Metis, "Metis means a person who self-identifies as Metis, is of historic Metis Nation 15  Ancestry, is distinct from other Aboriginal Peoples and is accepted by the Metis Nation" (November 11, 2002). This definition, adopted in Edmonton on September 27, 2002, clearly states that the only Metis who should qualify under section 35 of the Canadian Constitution are those related to the 19 century Metis Nation of the Red River area, and generally excludes other th  groups using the "Metis" label in Canada, often called the "Other Metis." This position is predominantly political, a form of advocacy for Prairie Metis rights and land title recognition, which are in an advanced stage of negotiation.  The "Other Metis" propose a broader definition de-centered from Red River: Metis people [...] are located in all of the provinces and territories of Canada.,These persons, who may or may not be entitled to status under the Indian Act, are Metis culturally, historically, and for purposes of constitutional recognition. Some of this group are a distinctive mixed blood population, others are closely identified with the history and culture of the French/Cree Metis in southern Manitoba and central Saskatchewan in the 1860s (Congress of Aboriginal Peoples, September 17, 2002). Although these Metis do not all refer to a common focal point at Red River, as does the Metis Nation, they are nevertheless nationally represented by the Congress of Aboriginal (CAP),  Peoples  which recognises one Metis organisation per province or territory. Right and land title  claims of these Metis are largely weakened by the literature. This literature argues, for example, that although metissage had occurred in the East, the Metis did not persist as a distinct people in this region, and that children of mixed marriages became members of either "White" or Indian society (Dickason,1985; Hebert, 1984, p. 432; Peterson, 1985). On the other hand, this literature also points to the problem of historical sources and to the fact that they are generally written — to which the largely illiterate Metis did not directly contibute. The Metis population in NouvelleFrance is a case in point. This population had been officially ignored in records because it was 16  closer to the "sauvages" than to the "civilised" and catholic French, and was not well accepted by colonial authorities (Dickason, 1985, p. 25; Perreault, 1982, p. 82; RCAP, 1996, p. 200). If, for  Five provincial Metis organizations are represented: the Metis Provincial Council of British Columbia, the Metis Nation of Alberta, the Metis Nation of Saskatchewan, the Manitoba Metis Federation, and the Metis Nation of Ontario (Metis National Council, April 29, 2002). Many Metis in British Columbia and in Ontario are not members of any of these organisations. Registering of mixed marriages by priests and missionaries when made under the Catholic Church would be a good example; marriages a la coutume du pays, made under Native "authority", were generally passed over in silence (Dickason, 1985, p. 22, 23; Hebert, 1984). 15  16  15  Jacqueline Peterson and Jennifer Brown, the enlargement of the use of Metis beyond the historical Red River area raises legal complications,  On the other hand, [it] recalls Marcel Giraud's suggestion that the processes and conditions which caused the metis to coalesce at Red River as a self-conscious ethnic group were rooted in both an historic past and a wider geographical frame, just as the processes of ethnic formation or "metisation" continued after 1885, often independent of the Red River metis (1985, p. 5). From this, it would seem inappropriate to invalidate Other Metis claims automatically, a conclusion reinforced by the place occupied by these Metis in the report of the Royal Commission on Aboriginal Peoples (1996, p. 255-264).  These different views among contemporary Metis stem to a great extent from flaws characterising the official recognition of Metis in the Constitutional Act of 1982 (Sawchuk, 1985). As pointed out by Catherine Bell, there are two principal and complementary issues with regard to the constitution and the Metis. First, the wording of the constitutional text does not make clear whether Metis are considered as many different peoples or as only one. Second, as mentioned above, it does not provide a clear definition of who qualifies as Metis, be they one people or many (1989, p. 1-59). However, the judgment of September 2003 of the Supreme Court of Canada in the Powley's case, which was ruled on favour of Metis Aboriginal rights in 17  Ontario, confirms that "Metis" includes more than people of Red River descent (Teillet, 2003).  It is not my intention to add anything to these legal and definitional issues. They are worth addressing only because they affect the nature of Metis claims and identity across the country. Moreover, since my point of departure is the peoples who self-identify as Metis, I consider the experience of Other Metis as relevant as those of the Metis Nation, and will study them accordingly regardless of legal issues.  This position cannot hide, however, the fact that the Metis Nation is the most documented and, consequently, easier to detect and explore (through maps of the prairie in the 19 century th  for instance) than the Other Metis, since written history does not present any direct evidence of  The Powleys are Metis from Sault-Sainte-Marie, Ontario, who were charged for illegal hunting after they shot a moose in 1993. They have always argued that they hunt only for subsistence and that, as Metis, hold Aboriginal 17  16  the latter. The nature of the historical records is certainly unsatisfactory (RCAP, 1996, vol. 4, p. 255) and I am obliged to find a way to fill historical gaps. A n exploration of oral history seems 1 ft  relevant; if official records (written source of collective memory) have revealed little, information may be found in community or family oral memory. Again, little has been done to gather and record the oral memory of these Metis peoples. Interviews become the inevitable methodological tool. It is unlikely that these interviews provide indisputable historical evidence, although they do lead us to a better understanding of the Other Metis' use of history for nationalistic discourse (mostly an unconscious phenomenon). I agree with the Finnish geographer Anssi Paasi that "the history of a given territorial unit is important for a population independent of any scholarly or academic historical standards" (1996, p. 47). Metis peoples may exist whether or not academics recognise them as such. It was obviously not possible to conduct interviews all over the country, not even in the Eastern part of it (where most of the Other Metis live), which, at least, would mean interviewing people in Labrador, New-Brunswick, Nova Scotia, Ontario, and Quebec. This was not feasible. Consequently, without changing any of the thesis objectives, I decided to limit my study to two Canadian Metis peoples, who seem fairly representative of the Canadian Metis as a whole: the Prairie Metis and the Quebec Metis. The selection of these two groups of Metis not only reduced the number of interviews, but also helped to focus map analyses. The Prairie Metis were an unavoidable choice, since they are the most numerous group in the country and the centre of the Metis Nation}  9  whose long-standing historical existence is well recognised (RCAP, p. 206 &  232). I chose the Quebec Metis for two main reasons: 1) they live in the space where some older ancestors of the earliest members of the Metis Nation originated {idem, p. 258); 2) and, the geopolitics of contemporary Quebec gives a very special flavour to the situation of Other Metis in Canada. They are "a minority within a minority within a minority" (idem, p. 264) for they live in the only Canadian territory where three forms of nationality are expressed: the Canadian, Quebecoise/Canadienne,  and Aboriginal identities. The Quebec Metis float among these  nationalisms.  rights under Section 35 of the Constitution. This was thefirsttime that Metis Aboriginal rights were de jure recognised by the highest court of the country. It is not impossible that these records have the potential to reveal more than they seem to do, but a huge reexamination of them would probably take years to achieve. The actual limits of the Metis Nation are not well settled and are a matter of discussion (RCAP, 1996). However, all Prairie Metis are part of it as opposed to Ontario and British Columbia Metis whose participation in the Metis Nation is partial and sometimes uncertain. 18  19  17  Thoughout the thesis the reader will find that my arguments are based on very few primary sources and on basic archival work. I mostly rely on secondary sources, be they contemporary (such as oblates, travellers or scientist writings) or recent academic studies on Metis issues. The broad scope of my thesis is the rationale for this. It seemed impossible to do in-depth archival research for a thesis that covers most of Canadian history and two different fields of exploration, that is the Prairie provinces and Quebec. It was appropriate, then, to take advantage of the considerable secondary sources. M y appoach, indeed, has the disadvantage of bringing little new evidence on the Metis for I mostly build my argumentation on others' work and conclusions. On the other hand, it has the advantage of offering a synthetic picture of what I consider to be an important geocultural layer of Canadian history and society, the Metis geography. The originality of my thesis lies with the historical exploration of this fundamental and recurrent geocultural layer of Canadian reality. Obviously, if my study had been limited to one specific time-space or one metissage!anti-metissage episode, it could hardly have emphasised the recurrence of Metis influence in Canadian society.  1.3  Thesis outline.  The maps and the cartographical shifts they reveal are the rationale behind my periodisation of the evolution of the Metis sense of territory and identity. This is the reason why, for example, I do not use the milestone date of 1821 — which is common for specialists of the fur trade — for starting my exploration of the 19 century Metis in the Northwest; I rather use 20  th  1819 as it was the year Peter Fidler's map of the Red River district was published. Yet, my periodisation is broadly in accordance with Canadian historiography, exposing maps as products of specific socio-cultural contexts.  See, for examples, Binnema et al. (2001), Brown (1980), Innis (1999), Judd & Ray (1980), and Ray (1998). This date corresponds to the merger of the HBC and its rival, the North West Company. This merger greatly affected the fur trade labour system, and following from that, the Metis who were an important part of the fur trade employees. According to Philip Goldring, the work force was cut in half in 1821 and was below 800 men, although it grew consistently thereafter (1980, p. 12).  18  This thesis comprises a historical part (chapters 2 to 4) and a contemporary section (chapters 5 and 6). Chapter 2 exposes cartographically the emergence of the Metis geographies in Quebec (during the French Regime) and in the Northwest (post-Conquest), and its subsequent disappearance in Quebec under British cartographical authorities. Maps prove to be central to the inscribing of the idea of metissage into European spatial discourse and ideology. Chapter 3 is devoted to the territoriality of the Prairie Metis during the "golden years" of the 19 century. The th  chapter is based on the juxtaposition of two methods of investigation: first, it explores Metis territorial markers as they are represented on colonial maps; second, it analyses Metis oral narratives in search of a more "internal" expression of the Metis sense of identity and territory. These two investigations establish the Metis reality as a dominant feature of the geography of the Northwest in the 19 century. Chapter 4 exposes the social and spatial marginality into which the th  Metis were confined during the first half of the 20 century; This period of Canadian Metis th  history is not explored in detail, the reasons for which are also given. Using Metis maps and the interviews I conducted with Quebec Metis, I will see in chapter 5 (Prairie Metis) and chapter 6 (Quebec Metis) how the contemporary Metis sense of territory has been affected by the rise of Canadian authorities' interest for the socio-economic fate of the Metis and non-status Indians in the 1960s. In turn, I will also examine how this sense of territory affects the way the Metis selfidentify today. If the Quebec Metis sense of territory and identity confirms "in-betweenness" as the main criterion of definition, it seems that Prairie Metis self-definition relies more and more on a historical belonging to a specific territory.  ***  The reader will find two chapters written in French (chapters 2 and 6). This is the result of a compromise between writing the thesis entirely in English — a language that still challenges me and often prevents the full expression of my intentions — or in French (my mother tongue), which proved to be difficult in the geographical and cultural context of this thesis. In spite of the difficulties of writing in two languages, this compromise turns out to be a "linguistic middle ground" that well suits a thesis on metissage.  19  Chapter 2:  Le(s) metissage(s) en question: les reponses coloniales et cartographiques au Quebec (1632-1857) et dans le Nord-Ouest (1785-1813).  // eft fans exemple qu'aucun [des Sauvages] ait jamais pris la moindre liberie avec les Francoifes, lors meme qu'elles ont ete leurs Prifonnieres. Ils n'en font pas meme tentes, & il feroit a fouhaiter que les Francois euffent le meme degout des Sauvageffes. — Pierre-Francois-Xavier  de Charlevoix, Journal d'un voyage fait par ordre du  roi dans l'Amerique Septentrionale. Au milieu des etablissements francais, les Indiens residents furent toujours le tres petit nombre. Et voila qui dispose, de ce pretendu metissage des premiers colons de la Nouvelle-France avec les Peaux-Rpuges du Canada, metissage dont la legende continue de courir en des milieux savants. — Chanoine Lionel Groulx, La naissance d'une race.  Le Canada des premiers temps fut marque par un decalage temporel fondamental entre 1'experience concrete d'un metissage biologique et culturel et 1'introduction abstraite et reflechie de cette experience dans l'univers mental colonial. Alors que la traite des fourrures et 1'interpenetration de l'univers europeen et du monde autochtone qui en decoule sont a l'origine d'un metissage de facto (Delage, 1991, p. 17; Perrault, 1982, p. 86), les preconceptions antimetisses freinerent considerablement l'infiltration de l'Autre dans l'identitaire europeen.  Par ailleurs, l'intensite et les formes prises par cet anti-metissage furent multiples, fluctuantes et ambigues (l'epigraphe en temoigne). S'il etait commodement admis que le Sauvage puisse se metisser, il n'etait pas dit qu'il en allait ainsi de l'etre «civilise», fut-il Fran5ais, Britannique ou Canadien. A l'evidence, l'ideologie des peuples primitifs, laquelle prone la civilisation de ceux-ci, fut au coeur de 1'anti-metissage en Amerique septentrionale.  La carte represente l'essence meme de l'ambiguite coloniale envers le metissage. D'une part, parce qu'elle se veut un compte-rendu objectif de la realite geographique, la carte ne peut eviter d'enregistrer la manifestation naissante de la geographie metisse. D'autre part, parce qu'elle est une production socio-culturelle, elle est sujette aux preconceptions anti-metisses.  20  Dans un tel contexte, on comprend pourquoi la realite metisse fait l'objet d'un manque flagrant de documentation, tant cartographique qu'historique, et qu'il soit d'ordinaire difficile d'identifier avec certitude les marqueurs territoriaux qui, sur la carte, devraient definir les frontieres sociales et geographiques de l'identite metisse . En consequence, l'objectif 1  fondamental de ce chapitre sera d'illustrer, avec le concours de la carte, l'emergence de la geographie metisse, source de l'intermediarite et de la territorialite des populations sang-melees du Canada. Nous verrons aussi que la representation cartographique de cette geographie est inegale dans l'espace-temps, car elle est parfois faite d'avancees appreciates et, d'autre fois et selon les cas, de reculs considerables. A u bout du compte, i l apparait que la carte eut un role essentiel a jouer dans 1'intellection du metissage.  Notre exploration cartographique de la geographie metisse se fera en trois temps. Elle s'attardera, d'abord, a la cartographie de la Nouvelle-France. Dans un deuxieme temps, notre analyse se concentrera sur la region des Prairies de la fin du xvm siecle et du debut du xrx e  e  siecle. Enfin, nous ferons un retour sur le territoire quebecois pour scruter l'attitude du colonisateur britannique a l'egard de la cartographie et du fait metis.  2.1  L e metissage et l a cartographie en Nouvelle-France (1632-1755).  La France entretient 1'ambivalence dans ses rapports au metissage et a l'etre metis en Amerique du Nord. Alors que le Metis remet dangereusement en question les conceptions socioculturelles coloniales, il represente aussi un rouage essentiel a la presence francaise en Amerique. Ce double sens fait de lui un etre a la fois redoute et un allie oblige, ce qui rendra les autorites coloniales incapables d'appliquer une politique arretee et coherente a son egard. Cette valse hesitation coloniale trouve sur la carte son expression geographique.  Les archeologues ont deja souleve ce type de probleme dans l'analyse de la culture materielle de certaines populations anciennes. II est en effet parfois difficile d'identifier l'origine ethnique de certains artefacts, car ceux-ci sont souvent marques d'influences culturelles multiples dues a des echanges interethniques. La theorie barthienne de l'ethnicite admet effectivement que les frontieres sociales puissent se maintenir meme si les criteres culturels definissant un groupe se modifient. 1  21  2.1.1  Le metissage en Nouvelle-France: l'entre-deux d'un passage.  Aux yeux d'une Europe en expansion, le debarquement de Christophe Colomb sur 1'ile d'Hispaniola fut davantage une deception qu'une decouverte. L ' Amerique etait une anomalie geographique, un obstacle sur la route des Indes et du Cathay — la Chine du Nord (Broc, 1980, p. 159; Litalien, 1993, p. 53; Morissonneau, 1996, pp. 222 & 225). Alors que les richesses auriferes incas consolerent quelque peu les Conquistadors, 1'Amerique septentrionale souffrait la comparaison, ce dont temoigne Jacques Cartier lui-meme en 1534 lorsqu'il considere la BasseCote-Nord comme «la terre que Dieu donna a Caen» . 2  En raison d'un tel constat, il fallait bien contourner cette Amerique-obstacle, ou mieux encore, la traverser de bord en bord de maniere a accoster enfin sur le littoral de cet Orient tant recherche. De cette pensee naitra l'utopie geographique fondatrice de la Nouvelle-France: le Passage vers l'Ouest (la fameuse Riviere de l'Ouest) . De continent-obstacle qu'elle fut d'abord 3  a l'esprit d'outre-mer, 1'Amerique devint, sans trop attendre, une terre de passage.  Le passage vers l'Ouest constitua l'impulsion premiere aux efforts de colonisation en Nouvelle-France (Morisonneau, 1996). H appartenait a la France d'etre la seule puissance coloniale a explorer, a prendre possession et a occuper les pourtours de ce passage dont le fleuve Saint-Laurent etait tenu pour porte d'entree. Cette consideration geopolitique dicta, plus que les imperatifs economiques, du moins au depart, 1'implication coloniale dans le commerce des fourrures (White, 1991, p. 127). L a traite fut une condition sine qua non aux alliances francoindiennes, et de ces dernieres dependait la penetration du fait francais en Amerique et 1'exploration du passage. Or, il ne saurait y avoir une penetration francaise du continent sans l'essor d'un espace social d'accomodation mutuelle entre Amerindiens et Francais, sans ce que Richard White a nomme le middle ground (idem).  On peut supposer que les pecheurs bretons et basques, lesquels exploitent les bancs de Terre-Neuve et du golfe Saint-Laurent au moment ou Cartier explore l'«inconnu», ne partagent pas les dires du marin de Saint-Malo. Les paroles de Cartier ne traduisent pas les pensees du pecheur de morue, mais celles de l'elite francaise — le roi en premier chef. Or, c'est cette elite qui produit les cartes de la Nouvelle-France. Les Anglais eurent leur Passage du Nord-Ouest aux confins de l'Arctique, parce que les Francais avaient opte avant eux pour le fleuve Saint-Laurent (Pastoureau, 1992, pp. 107 et 110). Le Metis, une strategic d'implantation coloniale: la creation mentale d'un peuple nouveau.  Explorer et occuper le territoire n'exclut toutefois pas la necessite de le peupler. Samuel de Champlain ne fut pas qu'un explorateur et un geographe, il fut aussi un administrateur colonial. Bien sur, force est de constater que 1'Amerique n'est pas tout a fait inhabitee. Cela dit, ceux qui la foulent depuis quelques millenaires ne sont que des pai'ens et sauvages qu'il faudra convertir. Ce faisant, ils deviendront Chretiens, civilises et sujets du roi de France.  La conversion du sauvage represents un double avantage. En plus d'assurer la suprematie du fait frangais, elle permettra de surmonter un probleme majeur a la colonisation, celui de la faiblesse des effectifs europeens. L a colonisation de la Nouvelle-France a un prix que les autorites coloniales se refusent a payer, soit celui de depeupler la vieille France en retour (Havard, 2003, p. 59). Afin d'eviter un tel ecueil, on aura tot fait d'entretenir une strategie d'implantation humaine sur le mariage mixte entre Autochtones et non-Autochtones. L'idee derriere une telle strategie de metissage est de fonder une nouvelle race (Dickason, 1985, p. 21; Trudel, 1960, p. 279). Le Metis, s'il est permis d'en parler ainsi, etait voue a devenir un rouage important de l'entreprise francaise en Amerique septentrionale, a devenir un «pion» tactique sur l'echiquier longitudinal du passage vers l'Ouest.  Neanmoins, la conception intellectuelle que se faisait la France du metissage et des Metis fut bien fragmentaire. II ne s'agissait pas de n'importe quel metissage ni de n'importe quel Metis. L'idee que le Metis puisse etre a l'image du «sauvage pai'en» peuplant cette Amerique «primitive» n'etait certes pas ce que les metropolitains avaient en tete. Le Metis officiel se devait d'etre socialement et religieusement conforme, c'est-a-dire, a la fois d'allegeance frangaise et catholique (Trudel, 1960, p. 278) . Le pere de Charlevoix rappelle ici le serieux des mesures 4  coloniales a l'egard de la religion:  lis sont pourtant nombreux ces protestants qui retrouvent refuge dans villes protestantes et cotieres de la France (La rochelle, Brest, Saint-Malo etc.) et qui souhaitent fuir a jamais la repression religieuse qui a cours en Europe. L'acces a la vallee du Saint-Laurent leur sera refuse. II ne saurait y avoir de Nouvelle-France digne de ce nom qui soit terre d'asile des «heretiques» huguenots (Jacquin, 1987, p. 101; Pastoureau, 1992, p. 116). Cette nouvelle contree sera catholique ou ne sera pas. 4  23  La Cour avoit donne des ordres tres-precis pour empecher qu'aucun Prote/tant ne pajjat dans la Nouvelle-France, & qu'on n'y permit l'exercice d'aucune autre Religion, que de la Catholique. [...] Les Mij^ionnaires de leur cote ye per/uadoient qu'en fixant le centre de leurs Mijyions dans un Pays, qui etoit en meme tems celui du Canada, il leur feroit ai/e de porter la lumiere de 1'Evangile dans toutes les parties de ce va/te Continent [...] (1744a, [tome I], pp. 180 et 186) Un tel metissage ne reposait pas sur une logique interculturelle. II se presentait davantage comme une manoeuvre d'implantation coloniale et comme une strategic d'assimilation culturelle de l'element indigene (Havard, 2003, p. 542). Paradoxalement, on peut qualifier cette conception coloniale de la relation autochtone/non-autochtone de profondement anti-metisse. L e M e t i s , une realite hors controle?  Une telle conception coloniale s'accorde mal a ce qu'on appellera la realite des marges coloniales. Ces marges sont les regions geographiques adjacentes aux principaux lieux de la colonisation francaise — la vallee du Saint-Laurent et 1'Acadie —, soit le Royaume du Saguenay, l'Outaouais, les Abitibis, la Mauricie, le Temiscouata etc. A ces marges plus immediates, il faut aussi compter celles comprises dans les pays d'en haut — la region des Grands-Lacs, les bassins de la Grande Riviere (Mississippi), de l'lllinois, de la Belle Riviere (Ohio) et de la Ouabache (Wabash River) — et celles qui gravitent autour de l'embranchement des rivieres Rouge et Assiniboine. L'Autochtone est, dans ces marges geoculturelles, une figure dominante et exerce une influence indubitable. Cette influence se traduit d'abord en termes materiel et politique. Comme le revele Conrad Heidenreich, la penetration europeenne du continent ne fut possible que grace a 1'adoption des methodes de transport existantes en Amerique du Nord, notamment le canot d'ecorce (2001, p. 239). L'influence materielle prend aussi une dimension demographique, laquelle est nettement a l'avantage de l'lndien. Contrairement aux treize colonies anglaises, la penetration coloniale du fait francais a l'interieur du continent se fait au compte-goutte, quelques Europeens a la fois, lesquels se retrouvent ainsi disperses sur un vaste espace autochtone (Delage, 1995). Cette realite demographique eut d'ailleurs une incidence sur l'influence autochtone sur le plan politique. Trop peu nombreux pour s'imposer par le nombre, il ne resta aux autorites coloniales qu'une seule alternative a l'obtention des droits de passage necessaires a la penetration continentale, soit celle de gagner la confiance  24  des nations autochtones par voie d'alliance et de s'impliquer ainsi dans les conflits intertribaux existants (Heidenreich, 2001, p. 241).  L'exploration du continent met a jour un autre type d'influence, informationnelle celle-la, et qui s'avere centrale a l'edifice cartographique europeen. A la fois orale et cartographique (Lewis, 1998; Harley, 1992), l'information geographique des populations indigenes est a la carte coloniale ce que les devotions sont aux missionnaires. L'experience directe qu'a l'Europe de l'interieur du continent est alors beaucoup trop modeste pour appuyer a elle seule la production cartographique. Ainsi, meme un explorateur de la trempe de Champlain dut recourir aux truchements — notamment Etienne Brule et Jean Nicolet — et a des informateurs autochtones. 5  Enfin, tout aussi crucial a notre propos est l'attraction qu'exerce l'univers symbolique autochtone sur 1'element eurogene. Malgre tous les efforts investis par les peres missionnaires a la conversion autochtone, i l semble y avoir davantage de Frangais qui s'ensauvagent dans les marges coloniales que d'lndiens qui se christianisent dans 1'ensemble de la Nouvelle-France (Jacquin, 1987; Trudel, 1960, p. 280). Les marges coloniales devinrent l'espace de predilection d'un «colon» bien singulier, soit le coureur de bois . Bien que le gros de ses activites soient 6  concentrees dans les pays d'en haut, le coureur de bois opere egalement dans les regions peripheriques du Quebec (Johnson & Martijn, 1994, p. 33). Cet individu qui prend souvent femme parmi les Premieres Nations, devient particulierement permeable aux manieres de faire et de penser autochtones, ce qu'affirme le pere de Charlevoix: «M. de Denonville revient en/uite aux Coureurs de Bois, dont il dit que le nombre eft tel, qu'il depeuple le Pays des meilleurs Hommes, les rend indociles, indi/ciplinables, debauches, & que leurs Enfants /eleves comme des Sauvages» (1744a, [tome I], p. 371 & 532-533).  . Le jesuite, comme tous les missionnaires et les commentateurs frangais de l'epoque, a possiblement exagere l'ensauvagement du coureur de bois (Beaudet, 1994; Delage, 1991, p. 21). Faut-il s'en etonner? Apres tout, l'univers mental europeen, du moins celui dont 1'elite se fait le  Francais qu'on envoie parmi les tribus autochtones pour qu'ils puissent en apprendre la langue et servir d'interpretes a la solde coloniale. De maniere specifique, le coureur de bois est un traiteur independant — lequel sera souvent associe a la traite illegale (Nute, 1955, p. 7) — qui commerce avec les Indiens des marges coloniales. Nous faisons toutefois un usage generique et incluons par ce terme, en plus du traiteur independant, tous les engages (employes de la traite) et les truchements ayant vecu parmi les populations autochtones (Beaudet, 1994, p. 2-3). C'est cette acceptation generale qui prone aux tous debuts chez les commentateurs europeens (Dickason, 1996, p. 84 ; Jacquin, 1987, p. 177). 5  6  25  porteur officiel, a du mal a concevoir l'existence d'une culture d'entre-deux comme celle qui se developpe dans les marges coloniales. Les missionnaires et les administrateurs — qu'ils se trouvent a Paris, a Quebec ou a Michilimackinac — pergoivent la rencontre franco-indienne comme le resultat d'une mise en opposition entre le monde civilise et l'etat primitif et sont convaincus de la superiorite materielle, culturelle et spirituelle du premier (Delage, 1992, p. 151). lis ne savent done pas trop quoi penser de cet individu deja trop indien pour etre considere civilise, mais encore suffisamment europeen pour ne pas etre confondu completement au «pur» sauvage americain.  Cela fait-il du Canadien un Metis? Encore faudrait-il que tous les Canadiens aient ete coureurs de bois. Plusieurs d'entre eux, la majorite en fait, furent d'abord et avant tout des «habitants». Si, comme l'affirme l'historienne Louise Dechene (1974) dans Habitants et marchands de Montreal au xviie siecle, plusieurs colons du Canada s'engagerent dans la traite des fourrures, ce ne fut generalement qu'une mesure temporaire visant a amasser le capital necessaire a l'implantation permanente et agraire dans la vallee du Saint-Laurent (idem, p. 487). C'est done que la course des bois, aussi importante fut-elle a l'exploitation de la ressource pelletiere, aux relations franco-indiennes et a la penetration continentale, n'en fut pas moins un fait marginal, et cela meme au debut alors que les effectifs coloniaux accusaient un net deficit feminin dans le ratio homme-femme (Dechene, 1974, p. 98-99; Dickason, 1985). Le metissage et l'influence autochtone n'affecterent pas tous et chacun de la meme fagon en Nouvelle-France. Si l'influence materielle autochtone penetra profondement les mceurs des colons du Saint-Laurent et de la baie Frangaise en Acadie (Delage, 1992, p. 110-136; Moussette, 2002) — comme en temoigne l'usage du canot d'ecorce, des raquettes et de la «traine sauvage» (Eccles, 1983, p. 24) — 1'assimilation symbolique de la culture autochtone eut un impact beaucoup plus limite. Comme l'indique le sociologue Denys Delage, la plupart des colons frangais au Canada n'abandonnerent point leur religion pour adopter et pratiquer une spiritualite autochtone. Les individus ayant emprunte, ne serait-ce que partiellement, les coutumes, les religions et les modes de pensee autochtones representeraient done l'exception plutot que la regie (1995, p. 60 & 62).  Aussi marginal fut-il, le metissage symbolique ne fut pourtant pas un phenomene sans importance. Pour Nicolas van Schendel, le coureur de bois a joue (et joue toujours!) un role aussi considerable dans l'imaginaire identitaire canadien que l'habitant. En fait, pour l'auteur, la canadianite repose des le depart sur une dialectique identitaire: d'une part, l'identite mosaique,  26  qui juxtapose et fait cohabiter des ethnicites distinctes et etanches (le Francais, le Canadien, ou rindien) et personnifiee par cet habitant enracine et isole dans les argiles peri-glaciaires du Saint-Laurent; et d'autre part, l'identite metisse du coureur de bois, figure marginale qui se construit sur les chevauchements culturels et l'ouverture a l'Autre ou au sauvage. Le coureur de bois decentralise done la territorialite canadienne en ramenant les marges autochtones dans l'univers territorial et identitaire du Canadien.  Metis ou pas, le coureur de bois pose bien des problemes, d'ordre conceptuel comme mentionne ci-dessus, et d'ordre politique. Non seulement le coureur de bois ensauvage est-il un etre imprevu et «illegitime» aux yeux des metropolitains, mais de plus, son intermediarite lui permet de s'imposer dans l'univers d'entre-deux de la geographie metisse. Essentiellement laisse a lui-meme — si ce n'est que de la presence spirituelle somme-toute modeste des missionnaires — le coureur de bois est pourvu d'un pouvoir croissant dans les marges coloniales qui n'est pas sans creer un sentiment de crainte en France (Perrault, 1982, p. 92 ; White, 1991, p. 69 & 214). Ainsi, apres l'avoir reconnu coupable de detournement d'intentions coloniales, les intendants de la Nouvelle-France tenterent periodiquement de mettre un terme a la domination du coureur de bois en limitant ses peregrinations dans les marges coloniales (par remission des conges, ces licences de traite, par exemple) et en sanctionnant les contrevenants (Dickason, 1985, p. 25; Innis, 1999, p. 67; Jacquin, 1987, p. 174; Perreault, 1982).  Manquant de coherence et de continuity, de telles mesures disciplinaires et coercitives n'atteignirent jamais l'effet dissuasif qu'on leur souhaitait et ne reussirent, au mieux, qu'a favoriser 1'emergence d'un commerce illegal avec les colonies britanniques au sud (Eccles, 1983, p. 137). En realite, la France se rendit compte, graduellement, qu'elle ne pouvait mettre un point final au coureur de bois et redefinir la geographie metisse selon les termes de ses propres ideologies spatiales et ethniques sans risquer de voir son empire americain s'ecrouler (Havard, 2003, p. 782; White, 1991, p. 318). Eliminer le coureur de bois se traduirait presque aussitot en pertes strategiques dans les contrees ou s'exercaient l'essentiel de la traite et les alliances francoindiennes. C'est ce contexte qui explique en grande partie l'incohesion des mesures etablies pour contrer les activites du coureur de bois et l'irregularite avec laquelle ces mesures furent appliquees (Perrault, 1982, p. 90-93).  27  En somme, le coureur de bois constitue une anomalie dans la conception que se fait la France de sa relation au monde autochtone. Qu'un individu (ou un groupe d'individus) puisse s'immiscer entre lTndien et l'Europeen et tirer profit de cet entre-deux n'etait pas prevu au programme. Les autorites coloniales reussissent bien un peu a ramener un tel individu sous controle colonial, a beneficier de son intermediarite et a faire ainsi des gains geopolitiques et geoculturels sur les territoires de traite (Delage, 1992, p.146; Havard, 2003, p. 407; White, 1991, p. 215 & 231). Mais a ce petit jeu, comme nous l'avons dit en introduction, l'lndien ne s'en laisse pas imposer, car lui aussi tire avantage de l'influence du coureur de bois qu'il va meme jusqu'a faire chef dans certains cas (Hebert, 1984, p. 432; White, 1991, p. 213) . C'est done dire 7  que l'etre metisse est une arme a double tranchant qui ne se laisse pas aisement manipuler (Delage, 1992, p. 146).  2.1.2  L a cartographie d u metissage?  La cartographie francaise du continent affiche une semblable anomalie. De maniere generale, elle est le reflet du projet geographique colonial et des perceptions spatiales et ethniques qui le supportent. Les cartes s'adressent a un public europeen et la plupart d'entre elles sont meme produites en France par des «geographes de cabinet» n'ayant jamais foule 1'Amerique. Les elements europeens et amerindiens sont tres bien differencies dans l'espace, mais aussi hierarchises de maniere a presenter la civilisation europeenne comme il se doit, e'esta-dire au dessus du monde primitif americain. En ce sens, la carte cree «an artificial image that gave America an European Identity. [...] cartography helped to invent America in the European  consciousness* (Harley, 1992, p. 530). En revanche, nous soutenons que la carte revele aussi, en filigrane, une image hybride de la realite geographique qu'elle represente et qui affiche, au mepris des conceptions coloniales, l'existence d'une geographie metisse.  Evidemment, les cartes neo-francaises repondent a une variete de fonctions et furent done produites a diverses echelles. II y a toutefois une dominance visuelle qui s'observe, soit celle des cartes a petite echelle, lesquelles permettent d'apprecier 1'expansion spatiale de l'entreprise  Cela s'entend, le commerce des fourrures ne depend pas uniquement du bon vouloir du coureur de bois, mais aussi des spheres d'influence que les postes de traite ont pour fonction d'exercer. Toutefois, la presence francaise dans les  28  coloniale francaise en Amerique ainsi que le developpement graduel du savoir geographique frangais. Ainsi, malgre un nombre non negligeable de cartes a grande echelle — plans de Quebec (Harris, 1987, planche 49) ou les cartes des seigneuries de Gedeon de Catalogne (1709) par exemple — la plupart des cartes s'attachent a la representation de superficies plus importantes du territoire. II en va ainsi des quatre cartes selectionnees pour cette analyse: la carte de Samuel de Champlain (1632, Figure 2.1), celle de Guillaume DelTsle (1703, Figure 2.2), celle de Philippe Buache (1742, Figure 2.3), et celle de Didier Robert de Vaugondy (1755, Figure 2.4). Les ideologies socio-spatiales, l'identite mosaique et les discours anti-metis.  A prime abord, on pourrait dire que 1'assimilation'del'Indien ne revet pas un souci particulier pour le cartographe. Au contraire, lTndien est fort present dans la cartographie neofrangaise. L a visibilite de la realite autochtone est d'ailleurs appreciable sur chacune des quatre cartes ici a l'etude. N'importe laquelle de ces cartes est truffee de toponymes autochtones et de noms de nations. S'ajoutent parfois a cela des commentaires revelateurs, comme celui qu'appose Champlain sur sa representation: Lieu ou les fauuagesfont  fecherie deframboife,  et blues tous  les ans. De plus, nombreuses sont les cartes qui presentent une iconographie assez fidele a la difference culturelle de lTndien (les images de tipis qui parsement la carte de Champlain et de DelTsle par exemple) et qui intensifie ainsi sa presence cartographique. La cartouche de DelTsle, laquelle expose des tableaux de la vie quotidienne de lTndien, est probablement le meilleur exemple de 1'importance graphique du fait autochtone sur la cartographie neofrangaise . 8  Le fait est que la carte repond a un objectif premier beaucoup plus terre a terre. La conversion de 1'Autochtone est un projet de longue haleine et dans l'immediat, ce qui importe, c'est d'en savoir plus long sur la geographie du continent. La carte de la Nouvelle-France repond d'abord a une fonction de synthese. Elle sert a rendre intelligible ce vaste continent (Boudreau, Courville & Seguin, 1997) et a etablir les strategies d'etablissement en Canada — a definir les  marges coloniales demeure ponctuelle et elle n'arrive que partiellement a controler les comportements des Indiens et des coureurs de bois. Le lecteur trouvera dans la carte du jesuite italien Francesco Bressani (1657) une iconographie «autochtone» encore plus abondante. 8  29  sites oil les forts seront construits par exemple (Havard, 2000, p. 17) — en vue d'imposer l'element colonial a la geopolitique du continent.  Et pourtant, 1'assimilation du fait autochtone est au coeur meme de la production cartographique. Non pas parce que la carte a pour but de faire disparaitre lTndien — meme s'il disparaissait sur la carte, cela ne l'eliminerait pas dans l'espace — mais parce qu'elle est tout simplement le reflet des representations socio-spatiales existantes . Si la carte constitue un 9  systeme de savoir, c'est qu'elle s'appuie sur les verites acceptees et legitimees par la societe d'origine du cartographe (Harley, 1992, p. 528). Elle permet done a la France d'assimiler mentalement le sauvage, avant meme qu'une telle conversion ne se produise. L a carte anticipe (Clayton, 2000; Harley, 1988). Cette capacite predictive de la carte est primordiale a une comprehension de l'univers colonial et cartographique a l'epoque de la Nouvelle-France.  Cette assimilation mentale pourrait d'ailleurs expliquer pourquoi plusieurs cartographes n'eprouvent aucune difficulte a avouer leurs sources autochtones, ce que fait Samuel de Champlain dans ses ecrits (Morissonneau, 1996, p. 225) et Guillaume DelTsle sur sa carte de 1701: «Sources du Misissipi suivant le rapport des Sauvages»  (Figure 2.2). Le fait qu'une telle  confession puisse mettre en relief l'influence du sauvage sur le Frangais ne semble pas affecter les cartographes. En fait, a l'instar de Barbara Belyea (1992a) et de Malcolm Lewis (1986), on peut dire que 1'information autochtone qui se retrouve sur la carte a ete, au prealable, traduite et assimilee, pour etre ensuite integree au savoir cartographique et geographique europeen. Une fois une telle operation mentale realisee, l'origine de rinformation importe peu, en autant qu'elle permette d'alimenter le savoir geographique dont la France a tant besoin pour etablir son hegemonie territoriale. Le cartographe ne s'interesse pas a rinformation autochtone pour ce qu'elle pourrait reveler sur l'occupation spatiale et sur la relation specifique que lTndien entretient avec l'espace, mais pour rendre intelligible et faire entrer dans l'univers mental europeen une realite qui lui etait jusqu'alors etrangere. Faisant appel a l'image, la carte connecte l'inconnu a 1'imaginaire europeen et est primordiale au processus d'intellection.  Cette assimilation informationnelle se confirme d'ailleurs avec la disparition visuelle et progressive de 1'Autochtone — un phenomene que Brian Harley a qualifie sous le terme de  30  silence cartographique  (1992, p. 531). A mesure que se concretised la penetration et  1'implantation francaises sur le continent, et a mesure que s'amenuise la dependance informationnelle envers les premiers occupants, la visibilite autochtone s'estompe. Ce fait est evident si Ton compare la carte de Champlain en 1632 et celle de Vaugondy en 1755 (Figure 2.1 & 2.4). Bien que lTndien occupe toujours une place considerable sur la carte de Vaugondy — mentions des nations autochtones et nombreux toponymes — cela n'a aucune commune mesure avec l'espace que lui accorde le fondateur de Quebec. Champlain represente avec moult details 1'emprise materielle du monde autochtone sur le continent, comme en temoignent ces icones representant des champs en culture et ceux rappelant les habitations amerindiennes (les maisons longues iroquoiennes et les tipis algonquiens).  Une attention plus particuliere a la carte de Champlain revele cependant un element etonnant qui met en relief la subtilite avec laquelle s'expriment parfois les ideologies coloniales. L'icdne representant les champs en culture rappelle davantage la maniere lineaire et geometrique d'organiser les terres cultivees en Europe qu'elle ne revele les techniques agraires autochtones. L'organisation spatiale des cultures iroquoiennes se faisaient generalement sous la forme de buttes ou etaient regroupes l'ensemble des produits cultives, a savoir les trois soeurs: la courge, le mais et le haricot. De plus, 1'agriculture autochtone etait generalement accompagnee d'activites complementaires telles la cueillette, la chasse et la peche, une realite que l'icone de Champlain garde sous silence (Rousseau, 1960, p. 45). JJ est bien possible que le cartographe ait choisi cet icone parce qu'il le savait intelligible a un esprit europeen. Ce faisant, il s'assurait que son lectorat ait une comprehension evidente de 1'usage des terres dans ces regions de 1'Amerique. En contrepartie, cette image peut aussi vouloir suggerer qu'un Indien cultivateur etait en quelque sorte un Indien en phase d'assimilation. Chose certaine, la carte de Champlain etablit une hierarchie visuelle parmi les differentes tribus autochtones et accorde une attention plus approfondie aux populations pratiquant 1'agriculture. S'il est difficile d'interpreter avec certitude les intentions du Saint-Ongeais, l'effet visuel reste sans conteste.  La cartouche de Del'Isle (Figure 2.2) peut etre re-analysee sur les bases du meme constat. La presence de lTndien ne se veut ni fortuite et ni la preuve exclusive de l'influence autochtone. Dans une perspective coloniale, elle est 1'expression de la suprematie de la civilisation  9  La repose l'essentiel du «pouvoir de la carte», soit sa faculte de jeter de la lumiere sur les conceptions territoriales  31  europeenne. Ce qui se retrouve au-dessus de 1'Autochtone represente le pouvoir supreme, a la fois temporel (comme en temoignent les armes du Roi) et a la fois spirituel, comme l'illustre ces scenes de conversion et de bapteme. Ces scenes exposent le passage ascendant du sauvage a la civilisation.  L'icone de Champlain et la cartouche de DelTsle etablissent la hierarchie coloniale et la suprematie du fait francais sur le fait autochtone. La representation des forts, ces ambassadeurs du fait colonial au cceur de l'espace autochtone, en offre une autre illustration. II s'agit d'observer les icones que Buache utilise pour saisir l'emphase que celui-ci met sur la presence europeenne dans l'espace (Figure 2.3); l'icone representant les forts Bourbon, la Reine, Maurepas ou S Charles ne fait pas la part belle a celui identifiant le village sauvage de l  Poskoyac, situe tout juste a l'ouest du lac Bourbon (pres de Le Pas au Manitoba). L'edification des postes de traite avait pour but de surimposer les imperatifs coloniaux a la geographie autochtone en tentant d'influencer les comportements spatiaux et commerciaux dans les marges coloniales; les autorites coloniales esperaient ainsi convaincre les nations autochtones de se reetablir a proximite de ces postes (Havard, 2000, p. 13-14; Havard, 2003, p. 283, 295 & 298). Or, si l'influence des forts tend a se consolider dans le temps — notamment au x v n i siecle (Havard, e  2000 p. 15 & 17) — il est clair que la carte anticipe et amplifie le succes tangible de l'emprise franchise des marges coloniales. Contrairement a ce que laissent supposer ces icones, rares sont les postes de traite qui s'apparentent a des forteresses imprenables, la plupart n'etaht que de modestes constructions parfois meme inocuppees (Eccles, 1983, p. 7; Havard, 2000, p. 14). Une telle representation laisse aussi sous silence le fait que ces postes de traite doivent une bonne part de leur existence aux Autochtones qui les acceptent sur leurs territoires (Delage, 1995, p. 59) et le fait qu'ils sont volontairement situes dans des endroits ou les Autochtones sont dominants, c'est-a-dire sur des points de passage, des lieux d'habitation ou de campement et des sites sacres (Harvard, 2003, p. 230-234). En bref, la carte de Buache evoque la suprematie visuelle des forts, et avec elle, celle du controle francais de l'espace americain; par surcroit, la carte met en scene l'effacement relatif du fait autochtone.  Malgre tout, l'effacement de lTndien demeure toujours incomplet et sa presence toujours aussi considerable, un fait qui met en relief la vision mosai'que de l'identite, pierre angulaire des  et culturelles deja en place.  32  conceptions coloniales en Amerique. En faisant entrer l'Autre, ce sauvage, dans l'imaginaire europeen, la carte expose, par la meme occasion, des categories spatiales et ethniques distinctes qui dessinent dans l'espace cartographique des spheres culturelles opposees ou l'ouverture a l'Autre semble symboliquement impossible. D devient ainsi de visu difficile de confondre l'espace culturel decoulant de la mention des Chekoutimiens, des Piekouagamiens ou des Anciens Algonquins sur la carte de Vaugondy (Figure 2.4), et la sphere symbolique dessinee par la croix (Quebec) marquant la presence catholique et coloniale dans la vallee du Saint-Laurent. La toponymie offre un autre exemple de ce cloisonnement ethnique. Alors que les toponymes euro genes se concentrent surtout dans la vallee du Saint-Laurent et en Acadie, les noms de lieux aborigenes sont en dominance sur le reste de 1'image cartographique.  L'image mosaique laisse bien peu d'espace pour autre chose, pour la representation de cet espace d'entre-deux d'ou emerge le fait metis. Cet impact visuel est d'autant plus considerable du fait que le vide est pratiquement inexistant sur ces cartes: i l est parfois rempli par des toponymes autochtones, des icones representant la vegetation (carte de DelTsle et celle de Champlain), etc.; il est d'autres fois camoufie par des elements satellites de la representation cartographique tels que la cartouche de DelTsle ou le carton Supplement Pour les Lacs du Canada de Vaugondy. L'espace residuel, celui-la meme ou les groupes Metis risqueraient le plus de s'affirmer, est pour ainsi dire evince de la carte. L'influence de l'Autochtone et l'hybridite cartographique.  Un second regard sur ces spheres culturelles permet toutefois de saisir une realite concomitante a l'image dominante de la mosaique, soit une image hybride. Si ces spheres sont presentees, comme nous venons de la voir, a la mariiere de faits cloisonnes et independants, force est d'admettre que les limites respectives de ces spheres ne sont pas toujours bien definies et qu'elles ouvrent la porte a un enchevetrement des faits autochtone et francais. Faut-il s'etonner de ce lien a premiere vue paradoxal entre l'image mosaique, expression de la difference et des categories ethniques, et l'image hybride, indice visuel de l'interpenetration culturelle, de l'ouverture a 1'autre et de la realite metisse? Pas du tout selon Sylvie Kande, un professeur de litterature, pour qui l'idee du metissage des cultures ne peut naitre que d'un monde ou la difference entre les identitaires est d'abord mise en relief (1999, p. 31). Pour Kande,  33  1'identification des cultures doit preceder leur melange, car l'ouverture a l'Autre suppose d'emblee que Ton s'en distingue. Pareillement, 1'intrication visuelle des realites autochtone et francaise n'est possible qu'apres les avoir differenciees et compartimentees dans leur espace respectif.  Une relecture de la nomenclature toponymique permet de voir en quoi celle-ci est la marque d'une telle hybridite cartographique. Au-dela de la tendance generate, qui fait de la toponymie un element a la source de l'image mosaique, on se rend bien compte qu'il existe plusieurs breches dans les limites spatiales que dessinent les differents univers toponymiques. Ces breches devoilent un espace de transition ou se melangent les noms de lieux d'origine differente et qui permet la cohabitation des saints catholiques («Saint-Laurent» en l'occurrence) et «des seins» sauvages — Tadoussac signifiant «mamelles» dans la langue algonquine. Du reste, une telle intrication symbolique culmine dans les marges coloniales les plus immediates tels le royaume du Saguenay, le Temiscouata, le Richelieu, l'Outaouais, la Mauricie et ainsi de suite.  Le deuxieme indice de l'image hybride de la carte est de nature graphique. L'information cartographique autochtone, bien qu'assimilee au savoir europeen, demeure identifiable sur la carte coloniale. Selon le specialiste de la cartographie autochtone, Malcolm Lewis, les endroits de la carte ou les details physiques apparaissent schematiques, ou les lacs semblent geometriques et ou les reseaux hydrographiques montrent une facture simple et hierarchique (1986, pp. 9 et 2327) represented des preuves indirectes de 1'assimilation europeenne du savoir cartographique des premiers Americains. Ce que traduisent ces formes geometriques — ces «Beads on a String» pour faire usage de l'expression consacree (Ruggles, 1991, p. 66) — c'est la nature topologique du savoir cartographique autochtone (Lewis, 1980, p. 15; Lewis, 1987, p. 559). Contrairement au modele topographique de la carte europeenne, laquelle s'attache au respect des distances et aux specificites physiques (les meandres d'une riviere par exemple), la carte amerindienne s'interesse davantage a 1'arrangement relatif des differents elements physiques qu'elle represente. Ce qui importe a lTndien, ce n'est pas la distance precise entre deux lacs, mais le positionnement de chacun de ces lacs l'un par rapport a l'autre. La nature topologique des cartes autochtones favorise 1'inscription mentale de la carte, une condition essentielle aux cultures orales de 1'Amerique. Or, si 1'identification de la representation autochtone est parfois evidente 0'extreme ouest de la carte de DelTsle ou la partie nord de la carte de Champlain par exemple), elle est  34  dans d'autres cas incertaine. Que penser, en effet, du Saguenay de Vaugondy ou de l'Outaouais de DelTsle qui, dans un cas comme dans 1'autre, affichent une representation somme toute incomplete et schematique de regions qui, pourtant, accusent une activite coloniale non negligeable? Que dire encore du reseau hydrographique autour du Lac Ouinipigon [lac Winnipeg] sur la carte de Buache, une region dont la connaissance resulte a la fois de l'exploration des Laverendryes et de l'information autochtone ? On voit bien que l'image 10  cartographique se situe a mi-chemin entre le caractere geodesique du savoir cartographique europeen et la nature topologique des cartes autochtones. C'est justement ce «clair-obscur», cette ambivalence visuelle a l'egard de l'origine de l'information, qui ajoute a l'hybridite cartographique.  Ce n'est pas un hasard si- l'intrication visuelle entre l'element autochtone et le fait francais se concentre dans les marges coloniales les plus immediates — celles qui cotoient le Saint-Laurent (le Saguenay, l'Outaouais, la Mauricie ou le Temiscouata) et celles qui avoisinent les postes de traite des pays d'en-haut et du Nord-Ouest — parce que c'est justement la ou s'opere le plus gros de l'interaction entre Autochtones et Europeens et que s'exprime en premier lieu la geographie metisse en Nouvelle-France. L'image hybride est a la carte ce que l'identite metisse est au discours colonial, une realite marginale mais incontournable.  En depit de sa presence timide, l'image hybride temoigne a elle seule du role primordial que jouent les cartes de la Nouvelle-France dans l'avenement du metissage franco-indien et dans l'emergence de la geographie metisse. On doit le metissage a la penetration francaise de l'espace autochtone et a un certain equilibre du pouvoir d'influence dans les marges coloniales. Si les postes de traite constituerent la charpente materielle et strategique de cette penetration, on sait que la carte fut essentielle a la mise en place d'une telle charpente. C'est grace a l'information geographique que la carte contient que les autorites coloniales sont en mesure d'evaluer le progres de la marche francaise dans les marges coloniales et d'identifier les lieux les plus propices (et necessaires) a l'etablissement de nouveaux postes. Des lors, tous les yeux de la France sont rives sur l'hybridite cartographique, car elle seule fait etat a la fois de la prise de controle coloniale et de la subsistance de l'influence autochtone dans ces marges. L'image  On sait que 1'information geographique des routes reliant les lacs Superieur et Winnipeg repose pour beaucoup sur une carte produite par le Cri Ochagach. La Verendrye a reproduit cette carte en 1728-1729. II existe plusieurs autres  35  hybride n'est done pas seulement en aval des processus de metissage, e'est-a-dire le reflet cartographique d'une geographie metisse preexistante; elle est aussi, en tant qu'instrument principal de la strategie coloniale, en amont de ces mecanismes d'interaction interculturelle et a sa part de responsabilite dans l'emergence et le developpement du middle ground franco-indien.  2.2  Le metissage et la cartographie britannique du Nord-Ouest (1785-1814).  Pour l'Euro-Canadien de la fin du xvme-debut xrxe siecle, le Nord-Ouest est une vaste contree riche en fourrures pouvant engendrer une part reguliere de benefices. A l'instar du Regime francais, 1'interaction avec le monde autochtone est necessaire — celui-ci demeure le premier fournisseur de fourrures - — et le metissage est toujours une resultante incontournable de ; :  cette interaction. A la difference de la Nouvelle-France, toutefois, les processus anti-metis ne resultent pas de positions officielles imposees par un ordre politique superieur, mais plutot d'un amalgame bigarre de perceptions personnelles et communes colorees par l'origine sociale de l'individu ou de la collectivite (Van Kirk, 1980).  Le contexte dans lequel s'inscrit la relation Autochtones/non-Autochtones dans le NordOuest confere aussi au metissage une portee symbolique beaucoup moins chargee en opposition. Dans un premier temps, l'occupation du territoire n'est pas epaulee par un projet civilisateur. Les premieres congregations religieuses, catholique et anglicane, n'arrivent qu'a partir de 1818 (pour l'Eglise catholique) dans la jeune colonie de la Riviere-Rouge. D faudra meme quelques decennies avant qu'elles n'etendent leurs tentacules spirituelles en dehors de la colonie et mettent sur pied des missions volantes et permanentes dans tout le Nord-Ouest. La relation euroautochtone ne fut done pas, du moins officiellement, embrouillee par un bagage ideologique imposant 1'assimilation du primitif comme une condition sine qua non de l'occupation europeenne. Dans un deuxieme temps, la presence euro-canadienne dans le Nord-Ouest ne se fonde pas sur une entreprise colonisatrice; les colons ne sont pas invites, comme e'est le cas dans l'Est du pays, a venir s'installer et a cultiver les terres disponibles . En fait, l'arrivee massive de 11  cartes semblables a celle de Buache et produites a la meme epoque dont celle de de Bellin (1752), de La Verendrye (1737) et celle de son neveu la Jemeraye (1733). II faut preciser que les conditions se pretent difficilement a une telle entreprise. Si les terres sont fertiles et deja depouillees de vegetation arborescente, l'isolement et la difficulte que pose le transport limitent les ouvertures aux 11  36  colons serait nefaste a une entreprise qui se veut d'abord et avant tout commerciale. En effet, si la traite des fourrures exerce une pression enorme sur les territoires de chasse autochtones et leurs ressources, elle a le merite de s'appuyer sur les activites et l'occupation territoriale traditionnelles autochtones . En revanche, la mise en culture des territoires representait un 12  risque a l'emergence de luttes d'espace entre eurogenes et indigenes, comme cela fut d'ailleurs le cas aux Etats-Unis. Dans l'ensemble, la domination de la geographie de la traite des fourrures et 1'absence de politique coloniale a proprement dit ont favorise la concordance et la complementarite des interets autochtones et euro-canadiens dans le Nord-Ouest et ainsi reuni des conditions presque ideales en faveur du metissage.  2.2.1  La geographie metisse: Le Nord-Ouest, une terre de renaissance.  C'est dans ce contexte et sur les cendres du vieux metissage franco-indien — qui s'est etendu a la suite des poussees exploratoires des Laverendryes, de par les lacs de Bois et de Pluie, jusqu'aux contreforts des Montagues de Roches Brillantes — qu'une nouvelle geographie metisse prendra forme dans le Nord-Ouest. U faut dire que les cendres laissees par 1'abandon des alliances franco-indiennes dans les annees post-conquetes sont toujours chaudes au moment ou, au debut des annees 1760, l'interet britannique pour cette region se manifeste et la reprise des activites de traite s'amorce (Giraud, 1945, p. 210). Le pere Morice recense quelques-uns de ces Canadiens deja presents dans l'Ouest avant que n'arrivent les premiers contingents de la Compagnie de la Baie-d'Hudson  (CBH) et de la Compagnie du Nord-Ouest (CNO). Le frere oblat  mentionne ainsi Fleurimond, probablement ne en 1735 d'une mere siousse, et Francois Beaulieu, ne en 1771 d'une femme montagnaise (1938, p. 13-14): «Francois avait douze ans lorsqu'il vit arriver le premier blanc du Grand Lac des Esclaves, et, vers la fin de 1792, il dut pour la premiere fois se separer de son pere, qui allait accompagner Alexander Mackenzie...» (idem, p. 14).  marches ou pourraient etre 6coules d'eventuels surplus agricoles. Pour l'essentiel, ces colonies se devraient d'etre autosuffisantes. Bien sur, comme le montre Arthur Ray dans Indians in the Fur trade (1998), la traite des fourrures apporta aussi de nombreuses modifications dans les geographies autochtones (migrations, nouvelles activites economiques, etc.). Neanmoins, la traite reposait largement sur le maintien des activites economiques ancestrales autochtones (notamment la chasse et la fabrication du pemmican) et 6tait assujettie a plusieurs autres pratiques d'ordre socioculturel telles que les ceremonies d'echange de cadeaux. 12  37  L'historienne Jennifer Brown a retrace, quant a elle, l'existence avant 1770 d'une quinzaine d'alliances maritales interethniques au sein de la CBH impliquant generalement des officiers. Si ce type d'alliance fut d'abord condamne, i l se normalisa vers la fin du xvuf siecle, epoque a laquelle il se mit a jouer un role plus considerable dans la strategie commerciale de la compagnie (Brown, 1980. p. 52-80). Ce changement d'attitude etait commande par 1'amplification, dans le Nord-Ouest, de la competition commerciale que la France avait amorcee au milieu du x v i n siecle. Les traiteurs anglo-ecossais, lesquels operaient a partir de Montreal e  une variete de compagnies dont plusieurs se regrouperent en 1783 sous l'enseigne unique de la CNO, etaient a l'origine de cette reprise appuyee de la concurrence. Reprenant les anciennes routes de traite francaises, la CNO fit des manages interethniques la pierre d'assise de son etablissement commercial dans le Nord-Ouest. II etait en effet coutume parmi les bourgeois et parmi les engages cariadiens (qui composaient l'essentiel de la main-d'ceuvre) de prendre femme au pays et d'ainsi s'allier aux bandes autochtones (Van Kirk, 1980, p. 4).  Pourtant, ces alliances maritales ne furent pas le gage de 1'acceptation sans condition du metissage par l'element europeen. L a vie sociale dans l'espace commercial du Nord-Ouest reposait essentiellement sur les memes conceptions ethniques et eurocentriques observees en Nouvelle-France, lesquelles placaient la civilisation outre-Atlantique au-dessus de 1'Amerique primitive. II suffit de constater avec quelle violence les bourgeois de la CNO traitaient parfois les Indiens pour se convaincre que la relation qu'ils entretenaient avec le monde autochtone ne reposait pas sur des considerations d'egalite (idem, p. 90-91). Pareillement, quand venait le temps d'assumer ses responsabilites parentales dans le Nord-Ouest avant les annees 1790, le bourgeois de la CBH ne voyait que deux options possibles et opposees: soit qu'il laissait l'enfant a lui-meme, c'est-a-dire aux soins exclusifs de la mere et de la bande indienne, ou soit qu'il le prenait en charge et l'arrachait ainsi a l'influence autochtone pour lui fournir une education de nature europeenne, pour le «civiliser» en quelque sorte (Brown, 1980, p. 156). Quelle que fut sa decision, la dualite culturelle, base meme de l'identite metisse, courait moins de chance de s'imposer comme un element central a l'enculturation de l'enfant. Si la dualite culturelle etait davantage le lot des enfants issus du systeme etabli par la CNO, il n'est pas dit qu'elle etait valorisee parmi les partenaires de la compagnie. Au contraire, la plupart des enfants metis  38  avaient peu d'espoir de promotion au sein meme de la compagnie (idem, p. 45 & 158) . Les 13  rares parmi eux qui parvinrent a monter dans l'ordre hierarchique furent generalement des fils de bourgeois ayant profite d'une certaine education euro-canadienne a l'exterieur du Nord-Ouest . 14  Une chose demeure, les processus interculturels ayant eu cours dans le Nord-Ouest de l'epoque ont sans contredit donne forme a une geographie metisse. Par ailleurs, a l'equation socio-culturelle «Europeens\Indiens» sur laquelle reposait la geographie metisse en NouvelleFrance, le Nord-Ouest propose un nouvel element ethnique supplemental: le Canadien. Et cet ajout est de taille, au point ou le frangais, de prime abord etranger aux bourgeois anglo-ecossais et aux peuples autochtones, deviendra la langue premiere dans les operations de traite dans tout le Nord-Ouest (Nute, 1955, p. 5).  2.2.2  L'ethnogenese et la formation des groupes «proto-metis».  L'un des resultats les plus tangibles de cette geographie metisse est la naissance, au debut du XLX siecle, de la Nation metisse, sise principalement au carrefour naturel que forment les e  rivieres Assiniboine et Rouge au Manitoba. L a Nation metisse est bien documented et on reconnait depuis longtemps son importance historique (positive ou negative, selon l'historiographie a la mode) au Canada. Seulement, i l s'avere beaucoup plus ardu de trouver des sources historiques decrivant 1'existence des communautes de sang-meles a l'exterieur de la Riviere-Rouge, soit celles des regions boisees du Nord-Ouest, qu'on trouve generalement le 15  long des routes de traite et a proximite des forts, et que Ton qualifie de proto-metisses.  Depuis  une vingtaine d'annees, plusieurs historiens et ethnologues ont tire profit des theories sociales et du developpement de nouvelles methodes d'enquete (entre autre quantitatives) pour re-analyser les fonds d'archives, particulierement ceux de la CBH, et tenter de degager les mecanismes de delimitation sociale et geographique a la base de l'ethnogenese de ces groupes proto-metis.  L'historien Philip Goldring (1980, p. 57) precise toutefois que le rang socio-economique etait, plutot que l'origine ethnique ou raciale, le critere principal donnant lieu a cet ordre hierarchique. Pour plus de detail sur le role socio-economique des Indiens et des Metis dans l'organisation du travail au sein des compagnies de la fourrure, voir: Goldring, 1980; Grabowski & St-Onge, 2001; Judd, 1980; Judd, 1982; Nicks, 1980. Par «zones boisees» nous entendons a la fois la foret boreale (Woodland), surtout dans ses parties plus meridionales, et l'espace transitoire (ou ecotone) entre la prairie et la foret boreale — le Parkland (Ray, 1998, p. 28; Friesen, 1987, p. 91). u  14  15  39  Pionnier en ce domaine, John Foster propose un processus d'ethnogenese s'articulant en deux phases. Pour la premiere phase, il fait appel a la theorie sur l'ethnicite de l'ethnologue scandinave Frederick Barth, lequel souligne 1'importance des experiences et des comportements communs dans l'emergence d'un sentiment identitaire. Selon Foster, le Nord-Ouest de l'epoque offre trois types d'experiences communes: l'hivernement du male etranger, son mariage a la facon du pays et son alliance socio-politique avec les freres de son epouse autochtone. L'hivernement est ce qui distingue les hommes du Nord et les mangeurs de lard et ce qui place bien haut les premiers dans la hierarchie des engages de la fourrure . C'est generalement lors de 16  ces hivernements que s'organisent les manages mixtes. Ces manages deviennent particulierement importants a ces engages qui une fois leur contrat termine decident de rester en permanence dans la region et de profiter de leur statut d'hommes libres . Non seulement la :  11  femme joue-t-eTfe un role essentiel dans l'economie et l'exploitation de la traite des fourrures (Brown, 1980, p. 64; Foster, 1985; Van Kirk, 1980, p. 54-61), elle pave egalement la voie a l'integration de l'homme libre dans le reseau de parente autochtone. Or, c'est cette integration qui permet a l'homme libre de s'etablir dans le Nord-Ouest et de subvenir au besoin de sa famille (Van Kirk, 1980, p. 4).  La deuxieme phase de l'ethnogenese metisse identifiee par Foster concerne l'enculturation de la progeniture sang-melee. Cette enculturation depend grandement de la capacite du male etranger a passer d'un univers culturel a l'autre et a se faire influent aupres de 1'Autochtone comme du traiteur. Ceux qui reussissent — tous n'y arrivent pas, ce que demontre Heather Devine dans son etude sur les Desjarlais (2001) — se voient donner l'opportunite de s'etablir avec leur famille a part des bandes indiennes et a l'exterieur du voisinage immediat des postes de traite, soit souvent parmi d'autres families mixtes. Une telle isolation sociale et geographique favorise ainsi l'emergence d'un environnement socio-culturel ou les valeurs autochtones cotoient la culture euro-canadienne, environnement propice au developpement de l'identite d'entre-deux de l'enfant metis (Foster, 1985). D est possible aux specialistes d'observer cette unicite naissante par l'analyse des comportements particuliers de ces enfants dans l'espace.  Le mangeur de lard fait la navette entre Montreal et Grand-Portage a l'extremite occidentale du lac Superieur et revient normalement a temps pour passer l'hiver dans la vallee du Saint-Laurent. Les hommes du Nord s'occupent de la seconde portion du voyage, soit la distance qui separe Grand-Portage des districts eloignes du Nord-Ouest tels que P Athabaska et le bassin du MacKenzie. L'eloignement de ces regions explique l'hivernement de ces hommes dans le Nord-Ouest. Ces hommes sont liberes de 1'emprise du monopole des fourrures et peuvent ainsi, en toute independance, commercer pour leur propre profit. 16  17  40  De par leur nature economique, les archives de la CBH sont particulierement pertinentes. Parce qu'elles enregistrent les operations commerciales et comptables affectant la vie quotidienne des postes de traite, et parce qu'elles conservent aussi le nom des individus impliques dans ces operations, les archives de la compagnie permettent ainsi de penetrer la particularite des deplacements des families mixtes et de saisir la nature unique de leurs interactions avec les societes autochtones et euro-canadienne (Devine, 2001; Ens, 2001; Foster, 2001; Thistle, 1997).  II manque toutefois un element a cet echafaudage theorique et cela concerne 1'apport du pere canadien aux processus d'enculturation. Si la litterature sur le sujet a raison de mettre l'emphase sur les valeurs euro-canadiennes dont il est porteur, elle neglige en retour l'influence de son intermediarite dans l'ethnogenese metisse. Apres tout, l'homme libre ne doit-il pas son influence, son independance et l'isolement socio-spatial de sa famille a sa capacite d'occuper l'espace entre lTndien et le traiteur euro-canadien? II y a tout lieu de penser que 1'intermediarite fut elevee au rang de valeur fondamentale dans ces families mixtes, a tout le moins inconsciemment. Tout compte fait, le Canadien de l'Ouest est aussi un modele metis pour sa progeniture . 18  Cela laisse presumer la confusion identitaire et territoriale du Canadien de l'Ouest et du proto-Metis qui, en raison de l'environnement socio-economique de la traite des fourrures et de leur intermediarite respective, sont appeles a entretenir des comportements sociaux et spatiaux similaires. Bien entendu, le voyageur du Nord est d'abord decrit comme un Canadien, au meme titre que ses compatriotes et parents mangeurs de lard et habitants . D n'est pas originaire du 19  Nord-Ouest comme l'est le proto-Metis et son identite, comme sa territorialite, ont pris racine sur les berges du fleuve Saint-Laurent. Par contre, le voyageur du Nord est un Canadien d'une fibre bien particuliere, la meme dont etaient tisses les coureurs de bois d'autrefois. L'homme libre presente dans le Nord-Ouest le second episode du Canadien metisse, cette figure marginale de la canadianite. Marginale, bien sur, aussi longtemps qu'elle reste tapie entre les escarpements du Boucher canadien et ceux du piedmont appalachien; une fois (de)placee dans un environnement  On peut supposer qu'il en va ainsi de la mere indienne, laquelle est 1'intermediate oblige entre son epoux canadien et sesfreresautochtones. Tous les auteurs ne sont pas aussi categoriques. C'est le cas de John Whittier qui, dans son ouvrage intitule The 18  19  Red River Voyageur, offre une description plutot metissee du Canadien du Nord-Ouest: «Type of a hardy race was he: brawny and muscular, with a tinge of savage origin showing darkly in his sunburnt skin, his high cheek bones and coarse black hair. [...] His language was a French patois, embroidered with Indian and English words and phrases; and his dress was of the same piebald character (1892, p. 1).  41  qui prone 1'intermediarite, comme c'est le cas du Nord-Ouest, elle a tout lieu de s'affirmer comme une figure predominante. Autrement dit, l'identite metisse du Canadien ressemble a un gene recessif qui n'attend que les conditions favorables — un endroit ou les barrieres de l'ethnicite et de l'identite mosaique peuvent etre abaissees — pour entrer en scene.  Le Canadien metisse nait deux fois. U nait une premiere fois sous l'egide de 1'Eglise catholique en Canada. II renait ensuite sous la tutelle de la societe des voyageurs. Cette seconde naissance est d'ailleurs mise en relief par l'historienne Carolyn Podruchny dans une these de doctorat portant sur l'univers social du voyageur. Podruchny a su identifier au moins trois sites ou avaient lieu des rituels de bapteme, dont un premier tout juste a l'Ouest du Grand-Portage sur le lac Superieur, un site ou symboliquement (mais aussi geographiquement!) le mangeur de lard passait dans le domaine du voyageur du Nord (1999, p. 354). Ces baptemes n'etaient ni plus ni moins que des rites de passage, des zone «liminales» ou de transition pour reprendre les mots de Turner (1967, p. 93-111), enfantant de nouveaux hommes.  2.2.3  La cartographie du metissage: entre l'identite mosaique et l'identite metisse.  Malgre le fait que l'entreprise cartographique britannique dans le Nord-Ouest ne repose pas entre les mains d'une meme et seule entite — qui comme en Nouvelle-France centralise les moyens de production et les savoirs geographiques —, elle montre pourtant des caracteristiques assez homogenes. Ce fait s'explique en grande partie par le contexte de production, lequel est marque par la dominance de 1'aspect economique. La cartographie de la region releve done d'une fonction principale qui consiste a recueillir l'information geographique necessaire a 1'elaboration des strategies commerciales dans le Nord-Ouest (Ruggles, 1991) en identifiant les routes de traite, les postes, les populations autochtones etc. Ainsi, bien qu'il existe a l'epoque quelques cartes a grande echelle (1 : 25 000 max.) et a moyenne echelle, la plupart sont des cartes regionales a petite echelles , tout comme les trois cartes qui composent cette analyse: 20  celle de Peter Pond (1785, Figure 2.5), celle de Aaron Arrowsmith (1802, Figure 2.6) et celle de David Thompson (1814, Figure 2.7).  Ce portrait est evident lorsqu'on se penche sur les cartes de la CBH. Le «Cataloque A» de A Country so Interesting de Richard Ruggles est utile (1991, p. 193-236). De ce tableau, on note, entre 1709 et 1813, que pres du trois quart  42  Compte tenu du bas niveau de connaissance geographique, les cartographes britanniques, qu'ils soient a la solde de l'une ou l'autre des compagnies de fourrure ou bien employes par le gouvernement anglais (ce qui est le cas de la firme des Afrowsmiths), dependent passablement de l'apport informationnel autochtone.  Ce qui est nouveau ici, contrairement a la cartographie neo-francaise, c'est l'ajout d'un troisieme ordre culturel represente par le fait francophone. Bien que celui-ci soit sans surprise marginal, sa presence sur un document britannique n'est pas insignifiante. Elle pourrait etre le symbole du pouvoir et du controle qu'exercent les Canadiens dans le cadre de la traite des fourrures (Podruchny, 1999, p. 398) ou le signe de leur apport dans l'edification du savoir cartographique et geographique britannique. Bien entendu, la presence de l'element francophonepourrait aussi etre une indication de l'existence et du developpement des groupes «proto-metis». L'inscription cartographique des preconceptions ethniques: les transparences ideologiques.  De la maniere dont sont apposes sur la carte les elements euro-canadiens et autochtones, il est clair que la carte du Nord-Ouest suit les schemes ideologiques laisses par les cartographes francais et qu'elle fait de la conception mosaique de l'identite l'element central de sa representation. Que Ton s'attarde a lire la carte de Peter Pond, celle de Aaron Arrowsmith ou celle de David Thompson, on notera, en guise d'exemple, que l'espace du poste de traite n'est pas celui de 1'Autochtone. Cela est mis en evidence par les noms affubles a ces postes; une poignee sans plus des forts de traite portent des noms autochtones ou d'origine autochtone («fort qu'Appelle» par exemple). En revanche, ces cartographes ne semblent pas mettre autant d'emphase sur la hierarchisation des elements europeens et amerindiens et sur la suprematie des premiers sur les deuxiemes. La representation des postes de traite ne suggere pas une importance materielle et symbolique qui soit demesuree relativement a la realite. II faut meme un regard aiguise pour discerner la presence de ces postes sur la carte de Thompson.  des 154 cartes repertoriees sont a petite echelle. Ce portrait s'inverse apres les annees 1830 et 1840 alors que les  43 L'inscription de l'hybridite cartographique: le Canadien, le proto-Metis et le portage.  Les cartes du Nord-Ouest presenters elles aussi une image refractaire aux perceptions ethniques et spatiales de l'Euro-Canadien et qu'on a deja decrit comme une image hybride. Dans 1'ensemble, cette image s'exprime de maniere analogue et avec la meme subtilite que celle identified sur les cartes de la Nouvelle-France, soit a travers l'entremelement des univers toponymiques et graphiques. Bien que les noms de lieux euro-canadiens se fassent plus importants aux abords des routes de traite et des postes et que les toponymes d'origine autochtone soient legion a l'interieur des terres, les limites que dessinent dans l'espace ces aires toponymiques demeurent poreuses et diffuses. II est aussi difficile, par endroit, de discerrier la ligne qui separe visuellement le savoir d'origine autochtone et celui de provenance eurocanadienne. II y a des regions sur ces cartes ou le dessin et les traits n'ont ni la precision topographique generalement associee aux cartes europeennes, ni tout a fait la qualite geometrique et schematique des cartes autochtones. Ce fait est particulierement evident sur la carte de Peter Pond (Figure 2.5). Meme certaines regions ayant ete explorees et arpentees en personne par le Nor'Wester (specialement la route du portage La Loche et de 1'Athabasca) suggerent partiellement, par l'approximation relative de leur dessin, 1'assimilation de cartes autochtones . On pourrait en dire autant de la representation des rivieres Churchill (entre le fort 21  Churchill et la confluence de la Beaver River) et Hayes sur la carte de Aaron Arrowsmith (Figure 2.6) . La carte de Thompson est quant a elle particuliere. Dans l'ensemble, la precision du 22  dessin laisse entendre que la carte de l'ancien apprenti de la CBH et alors arpenteur-geographe pour la CNO repose typiquement sur une information geographique decoulant d'observations directes et scientifiques; les routes de traite (les principaux cours d'eau), lesquelles forment l'essentiel de l'image cartographique, sont faites d'un dessin precis et detaille. Un second regard permet toutefois d'identifier bon nombre d'affluents secondaires a peine esquisses — aux formes rectilignes et imprecises — et qui rappellent les conventions cartographiques autochtones.  cartes a grande echelle dominent a environ 60 pour cent (sur un total de 427 cartes). En fait, le travail cartographique de Pond fut fait essentiellement de memoire et repose en bonne partie sur une information mixte, a la fois directe et a la fois de seconde main autochtone (Belyea, 1992a, p. 272). 21  44  L'hybridite cartographique du Nord-Ouest est aussi encherie par l'ajout de l'element francophone. Cet element marginal de la carte, qui s'exprime essentiellement sous forme toponymique, s'insere entre les spheres socio-culturelles britannique et autochtone et suggere l'existence, visuellement parlant, d'un espace d'entre-deux. Effectivement, le fait francophone dessine sur la carte un espace liminaire qui s'affirme comme le passage (ou le lieu commun) oblige entre ces deux spheres dominantes, ce que l'ascension du francais au rang de lingua franca de la traite des fourrures vient confirmer dans les faits. Si la carte d'Arrowsmith contient relativement peu de ce type de noms (il y a bien mention de Bonnet L., de Malign pt. et de Brocket Head sur le lac Winnipeg et de LaleCross [lac L a Crosse] Lake), la carte de Thompson en est truffee sur 1'ensemble de sa surface. On retrouve surtout ces toponymes sur la route entre le lac de Bois et le lac Winnipeg (Rat portage, Terre Jaune, Isle portage, Bonnet lake), sur la route du Portage la Loche (Lac la Ronge, Lake Isle a La Crofse, Loach lake) et sur le long de la riviere de la Paix (Grand Marac [Grand Marais], Fort Liard, Fort du Tremble et Fort Vermilion).  Ces toponymes trahissent ainsi 1'intermediarite du Canadien et du proto-Metis. Ce n'est pas un hasard si les toponymes francais se concentrent dans les regions dominees par la CNO et qu'ils designent en majorite des lacs, des rivieres et des portages. Ces types d'entite geographique forment dans le Nord-Ouest 1'infrastructure de base de la geographie de la traite des fourrures, et consequemment, constituent la scene geographique ou se jouent les interactions interculturelles. II n'est pas surprenant qu'ils soient aussi, comme l'est le Portage La Loche, les lieux de la plupart des pointes aux baptemes identifiees par Podruchny (1999, p. 354). Ces lieux de baptemes president a l'eclosion de la figure metisse du Canadien du Nord; ils sont a la fois des passages physiques et mentaux d'entre-deux mondes, des espaces liminaires a mi-chemin entre deux bassins identitaires. Lieux d'ethnogenese, le portage, le lac et la riviere incarnent aussi, du coup, la territorialite naissante des proto-Metis.  Si on connaissait deja l'importance de l'element francais au fonctionnement de la traite des fourrures (cet element est une main-d'oeuvre et un moyen de communication), on peut maintenant apprecier, grace a la carte, toute la centralite de cet element dans l'univers d'entre-  Arrowsmith fait un usage appreciable de la production cartographique des employes-cartographes de la CBH dont Peter Fidler. Or, plusieurs des cartes de Fidler sont des transcriptions integrates de cartes autochtones, notamment celles d'Ac ko mok ki, d'Ak ko we ak et de Ki oo cus, toutes reproduites en 1802 (Ruggles, 1991, p. 63).  45  deux qui emerge dans le Nord-Ouest de l'epoque. L a carte revele le fait francais comme un ciment unissant les univers autochtone et europeen, comme l'essence meme de la realite metisse dans la region. Ainsi la carte met-elle en evidence un fait que la litterature sur l'ethnogenese metisse a jusqu'ici sous-estime et constitue-t-elle une «preuve» visuelle de la participation primordiale du Canadien metisse dans le developpement de la territorialite a l'origine des groupes proto-metis.  2.3  Les effacements cartographiques dans the Province of Quebec (1764-1857).  Le Quebec d'apres Conqu&e fut marque par deux evenements complernentaires: de un, la diminution constante et presque complete du commerce des fourrures dans les regions peripheriques; et de deux, l'avancee des fronts colonisateurs dans ces memes regions, autrefois «inhabitees» (Delage, 1991, p. 24). Partenaire incontournable dans l'industrie de la traite des fourrures, 1'Autochtone devint rapidement un obstacle au developpement agraire et a l'industrie forestiere, ces grands consommateurs de foret et d'espace (Fortin & Frenette, 1989, p. 31). Le metissage ne pouvait done que battre en retraite devant Faction concerted de la faux et de la hache.  2.3.1  Ambivalence (anti-)metisse.  II semble que le metissage n'a jamais ete aussi central a 1'implantation britannique en Amerique d'avant Conquete qu'il ne le fut pour la France. L'image que la litterature renvoie de l'experience coloniale anglaise en est une de froideur et de distance a l'egard de l'Autochtone. Si la traite des fourrures fut aussi substantielle a l'economie coloniale de 1' Angleterre, elle ne reposait pas sur les memes premisses socio-culturelles et diplomatiques qui, comme en Nouvelle-France, furent a l'origine du middle ground. Cantonnes dans l'espace europeanise de quelques postes de traite etablis le long des rives arctiques de la Baie d'Hudson, les Anglais attendaient que les Autochtone viennent d'eux-memes porter leurs fourrures. Plus au sud, les treize colonies representaient l'antagonisme territorial de la traite des fourrures. II s'annongait deja inevitable que leur expansion demographique et spatiale se ferait aux depens des territoires  46  de traite et autochtones situes a l'Ouest des Appalaches et que 1'interpenetration culturelle se retrouverait compromise (Eccles, 1983, p. 5).  Dans de telles circonstances, on est naturellement porte a penser que le Quebec d'apres la Conquete fut marque par la predominance des discours anti-metis. Le tableau est cependant plus complexe et plus nuance qu'il n'apparait au premier regard. L a negation du metissage durant le regime britannique est loin d'etre automatique et loge souvent a 1'enseigne de l'ambivalence; elle est evolutive, rarement explicite et souvent a peine perceptible . L a carte offre toutefois, comme 23  nous le verrons, une conception beaucoup moins nuancee de 1'anti-metissage. En fait, par sa capacite prospective, la carte du Quebec sous le Regime britannique aura tendance a predire l'effacement de 1'Autochtone et la separation ethnique du territoire, avant meme que ces deux phenomenes ne se concretised reellement dans l'espace.  -'  La nature et l'intensite du discours anti-metis dependent largement des interets en jeu. Au debut du Regime anglais, ces interets etaient divers et parfois meme opposes. D'une part, les autorites coloniales agissent comme si l'assujetissement de l'Autochtone des pays d'en haut etait le corrolaire de la chute de la Nouvelle-France. En demandant le retour des Britanniques faits prisonniers durant la Guerre de Sept-Ans et, surtout, en abolissant la politique des presents, les autorites faisaient ainsi un accroc aux sentiments des populations autochtones et rejetaient du coup la relation d'interdependance au coeur du commerce franco-indien et du metissage (White, 1991, p. 256-261). II n'est pas sur qu'une telle strategie servait les interets des marchands anglais. Reprenant a leurs comptes les activites de traite dans la region des Grands-Lacs, abandonnees avec la Conquete par la plupart des negotiants francais, les marchands anglais reconnurent l'avantage de tirer profit de l'experience des Canadiens et de l'influence considerable de ceux-ci sur les nations autochtones (Brown, 1980, p. 35; Innis, 1999, p. 168171). On comprend que ces marchands n'adhererent pas aussi profondement au discours antimetis colonial. Cela dit, avec l'eclatement, en 1763 et 1764, des revokes autochtones dans les pays d'en haut, les autorites durent reviser leur attitude conquerante, une revision qui fut fortuite a l'emergence d'un nouveau middle ground (White, 1991, p. 315-365).  II y a toutefois une chose qui soit consistante du debut jusqu'a la fin: l'absence d'une categorie legale definissant les droits et les responsabilites des Metis.  47  Le discours anti-metis est aussi evolutif dans le temps. Dans la premiere moitie du xrxe siecle il prit un ton particulierement pejoratif en condamnant la «degenerescence» des Canadiens (Delage, 1991, p. 26; Smith, 1974, p. 74). Les autorites coloniales decrivirent ces derniers comme des Francais ayant troque leurs vertus europeennes en echange de quelques charmes autochtones. Cette vision peu flatteuse du metissage canadien cherchait a legitimer une conviction depuis longtemps inscrite dans l'esprit colonial anglais, a savoir que les Canadiens etaient voues a s'assimiler (Christie, 1848, p. 60 & 73). Une telle conviction est d'ailleurs fondamentale sous la plume de lord Durham, lequel, dans le fameux rapport qui porte son nom, depeint le Canadien comme un Francais inferieur et propose son assimilation complete (Bouchard, 2000, p. 95 ; Brunet, 1958, p. 195). L'Acte d'Union de 1840 fait d'ailleurs suite aux propositions de Durham. En gros, le metissage est une affaire essentiellement canadienne dans la perspective coloniale.  La reponse de la bourgeoisie canadienne a ce discours pejoratif fut elaboree sur un canevas anti-metis d'une tout autre nature. Au lieu de renier le caractere depreciatif donne au metissage franco-indien par Durham et ses acolytes, les elites canadiennes proposerent un discours de la survivance repudiant 1'existence meme de ce metissage. Ce discours chercha a rappeler la memoire et la culture heritees de la France — a entourer d'une aura de purete la race canadienne — vues comme les seules armes veritables contre 1'assimilation (Bouchard, 2000, p. 107-110; Dumont, 1993; Lamonde, 2001, p. 53-62). En d'autres termes, le Canadien se defend bien d'etre degenere, quitte pour cela a reecrire, comme le firent Francois-Xavier Garneau et Lionel Groulx , l'histoire de la nation canadienne en la presentant, genealogie a l'appui, comme 24  la descendante directe (enfin presque) de la noblesse francaise. Dans l'esprit de la bourgeoisie canadienne, s'il y a bien un metissage franco-indien, celui-ci se limite a cette contree qui a vu naitre Louis Riel, le Nord-Ouest.  Quoi qu'il en soit, pendant que les autorites coloniales et la bourgeoisie canadienne s'opposent sur le terrain de l'ideologie et de la rhetorique, dans les marges coloniales quebecoises, la realite metisse continue a unir par le mariage des univers ethniques distincts et a  II ne fait aucun doute que l'opposition du chanoine envers l'heritage metisse du Canadien est avant tout une contestation de la vision negative d'un tel metissage par le Canada anglais. Dans une note de bas de page Groulx affirme: «I1 n'y aurait pas lieu de mettre la moindre ardeur a dissiper cette legende [le metissage] si elle ne servait d'appui a des theories d'ethnologues sur l'inferiorite des races metissees et n'avait permis a quelques historiens d'esquisser des fantaisies assez peu complaisantes sur le caractere du peuple canadien-francais» (1930, p. 25).  48  enfanter une progeniture illegitime. L'«illegitimite» de cette descendance tient a deux facteurs: de un, ces enfants sont souvent nes de manages d'abord celebres a la mode du pays et catholiquement consacres en seconde instance seulement; de deux, ils rappellent cette «degenerescence» dont le Canadien est accuse et doit se defendre. En consequence, on soupconne que plusieurs de ces naissances furent officiellement camouflees (Hebert, 1976, p.xii; Trudel, 1960, p. 262). Grace a une lecture «deconstructive» des registres jesuites de Tadoussac, Leo-Paul Hebert (1984) releve plusieurs de ces alterations officielles et montre qu'une part non 25  negligeable de la population du Saguenay et du Lac-Saint-Jean d'avant colonisation (1842) fut d'origine sang-melee. Une commission speciale sur les affaires autochtones du gouvernement federal note d'ailleurs, en 1856, la presence de quelques families «of half breeds» a la reserve de Mashteuiatsh (alors nommee Pointe-Bleue) au Lac-Saint-Jean; d'apres cette source, les Metis auraient d'ailleurs plus de succes en agriculture que leurs congeneres indiens (Anonymous, 1858). JJ est fort probable que d'autres regions peripheriques quebecoises, telles l'Abitibi, le nord de l'Outaouais, la Mauricie et le Temiscouata, aient ete les theatres de semblables alterations.  II est meme vraisemblable que des populations metisses aient developpe, dans ces regions, une certaine distinction identitaire. Si l'on jette un coup d'oeil attentif au journal de Neil McLaren (Bouchard, 2000), chef du poste de Chicoutimi entre 1800 et 1804, il semble bien que les conditions socio-spatiales necessaires a l'ethnogenese metisse du Nord-Ouest aient ete aussi presentes dans les King's Posts . U y a bien quelques hommes libres influents qui, comme 26  Francois Verreaux, «gardien des Terres-Rompues» (situees a quelques kilometres au nord-ouest de Chicoutimi), se sont isoles avec leur famille metisse de la proximite immediate des bandes autochtones et des postes de traite. Les Verreaux n'ont pas les memes comportements dans l'espace que les «sauvages» dont fait mention McLaren. Leurs contacts avec les traiteurs sont plus reguliers, sans toutefois trahir une quelconque attache envers le poste de traite comme c'est le cas des engages canadiens. On ne peut certes pas comparer I'ampleur du phenomene metis du Saguenay, ou celui d'autres regions du territoire quebecois, avec celui s'etant developpe a la meme epoque dans le Nord-Ouest. Les realites economiques, politiques et demographiques des  Comme l'indique Leo-Paul Hebert dans sa presentation du troisieme registre de Tadoussac: «Parfois la redaction francaise se poursuit en latin par souci de discretion: baptemes d'enfants nes d'une union illicite...»(1976, p. xii). A proprement dit, le journal de Neil McLaren ne dit rien sur les Metis, laissant croire que son auteur ne fait aucune distinction particuliere a leur endroit. On doit 1'identification de l'origine metisse de ces individus au travail de presentateur du journal, l'historien sagueneen Russel Bouchard. Les registres de Tadoussac represented de toute evidence la source premiere de l'historien. 26  49  faits metis du Nord-Ouest et du Quebec n'ont aucune commune mesure. Cela ne veut toutefois pas dire qu'il faille negliger 1'existence possible de l'identite metisse quebecoise.  2.3.2  Les puretes et les clartes: la fonction de la carte et ses consequences.  La politique coloniale des Britanniques a un effet tangible sur la cartographie du territoire quebecois, a commencer par son mandat lui-meme. Plus que jamais, depuis la prise de possession du fleuve Saint-Laurent par la France, la carte a pour fonction premiere l'organisation administrative du territoire (Boudreau, Courville & Seguin, 1997, p. 3-40) . La nomination, en 27  1764, de Samuel Holland au poste d'arpenteur general est une indication des desseins britanniques a l'egard de 1'administration des territoires nouvellement acquis (Boudreau, 1991, p. 39). Le cartographe britannique s'attarde a rationaliser l'espace par l'application de divisions administratives clairement definies. La carte agit ici comme un outil de planification, un cadre graphique sur lequel les strategies colonisatrices seront elaborees. Dans un espace aussi bien delimite et cloisonne, i l reste bien peu de place a l'intrication visuelle et a la representation des espaces d'entre-deux. L'anti-metissage est roi au pays des cartographes et des arpenteurs. Le cloisonnement ethnique et la mosaique coloniale: la cartographie d'un espace anti-metis.  La carte du Regime britannique endosse sans complexe la «theorie de la ligne a tracer», ou si Ton prefere, la conception mosaique de la territorialite. L'image mosaique n'est plus, comme auparavant, 1'expression souvent inconsciente des preconceptions ethniques et spatiales des autorites coloniales. Selon Mark Warhus, specialiste en histoire de la cartographie, cette image serait plutot garante des intentions coloniales, c'est-a-dire l'etablissement de nouvelles logiques socio-culturelles et spatiales et la prise en charge du Canadien conquis et de l'lndien «pacifie» (1997, p. 159).  II y a bien eu, durant le Regime francais, des cartes comme celle de Gedeon de Catalogne qui avaient pour objectif de representer en detail l'occupation seigneuriale de la vallee du Saint-Laurent et celle de ses tributaires les plus importants. Mais sur toute la production cartographique neo-francaise, ces cartes font figure d'enfants pauvres. Comme le mentionnent Courville et Labrecque, «ce n'est done qu'apres la Conquete et plus exactement apres la  50  On observe d'abord, a l'echelle des possessions britanniques, l'importance du lien entre les delimitations cartographiques et les politiques coloniales. La carte de Thomas Kitchin (1764, Figure 2.8) — graveur, editeur, hydrographe du roi et prolifique cartographe —, laquelle se veut une representation du territoire quebecois dans son contexte regional immediat, est particulierement eloquente. Elle s'accorde avec le premier temps de la segregation socio-spatiale du colonisateur britannique. Les limites de la Province of Quebec, elaborees dans le cadre de la Proclamation royale de l'annee precedente, sont pour une des premieres fois mises sous forme cartographique . L a Proclamation redessina la geographie politique americaine et attribua a 28  chacun un espace propre: le Sauvage dans le Pays indien — vaste territoire comprenant la region des Grands-Lacs, des Illinois, de l'Ohio et meme une mince partie du Nord quebecois (excluant les eaux se deversant dans la baie d'Hudson appartenant a la CBH); T Anglais dans les treize colonies atlantiques; et enfin, le Canadien dans the Province of Quebec. Ces limites entrainent l'isolement visuel du Canadien et, conjointement, son expulsion cartographique des pays d'en haut et des regions plus nordiques du territoire quebecois d'aujourd'hui, telles que le Nord-Ouest du Lac Saint-Jean. L'absence cartographique des Grands-Lacs et du Pays indien, sur la carte de Kitchin, renforce d'ailleurs cet effet d'isolement du Canadien . En outre, la nature rectiline et 29  artificielle d'une bonne part des limites de la province ajoute une dimension «planifiee» et prospective a la separation geographique de 1'Indien et du Canadien, laquelle correspond a une distinction sociale dans l'esprit anglais: lTndien est un allie oblige — surtout apres la revoke de Pontiac — alors que le Canadien est un vaincu.  La carte de Samuel Holland  (1829, Figure 2.9) opere a une echelle plus grande que la  carte de Kitchin. Ce faisant, c'est de maniere beaucoup plus precise et appliquee qu'elle presente les limites administratives a la base de 1'organisation coloniale et ethnique du Bas-Canada. U ne s'agit plus de gerer les differentes parties de l'empire britannique en Amerique, mais de planifier l'utilisation meme du sol. A ce changement d'echelle correspond le deuxieme temps de la segregation ethnique et geographique, laquelle commence avec l'arrivee des Loyalistes (1792),  nomination d'un Arpenteur general des Terres en 1764 que la veritable cartographie des seigneuries s'amorcera» (1988, p. 12). Ces limites reviennent souvent dans la cartographie britannique, meme apres qu'elles furent officiellement modifees par l'Acte de Quebec de 1774, comme en fait foi la carte de Jonathan Carver en 1776. II est a noter que l'absence des Grands-Lacs n'est pas une constante dans la cartographie de l'epoque. Kitchin luimeme produisit d'autres cartes montrant cote-a-cote the Province of Quebec et le pays indien, dont celle de 1769. Cette carte fut publiee a titre posthume des 1802 (Courville & Labrecque, 1988, p. 13).  28  29  3 0  51  premiere vague serieuse d'arrivants britanniques: la seigneurie au Canadien, le canton au nouvel arrivant britannique et, un peu plus tard (1850-1853), la reserve a l'lndien (van Schendel, 1994, p. 108). Cette segregation est particulierement evidente sur la carte de Holland. D'une part, la representation du systeme de canton fait ressortir les limites des vieilles seigneuries et cree un effet d'encerclement de l'ancien territoire canadien . Les deux systemes de regime foncier apparaissent distincts, bien qu'ils soient tous deux des abstractions geometriques. D'autre part, la toponymie se charge d'ethniciser l'espace du canton et de la seigneurie. Alors que cette derniere conserve son caractere canadien et ses anciennes denominations francaises, le premier est incontestablement de nature anglaise. Les noms attribues aux cantons proviennent essentiellement des Ties britanniques, et cela meme si la plupart des cantons n'etaient toujours pas habites et mis en valeur par les immigrants britanniques. Le choix de ces noms n'est bien sur pas innocent. Ce faisant, on cherchait possiblement a attirer les colons de langue anglaise en leur presentant une terre d'accueil au paysage culturel familier. Cette carte fut re-editee huit fois entre 1802 et 1843 (Courville & Labrecque, 1988, p. 13), ce qui suggere une certaine popularite et une assez large distribution.  La carte de Joseph Bouchette fils (1857, Figure 2.10), 1'aine de celui qui succeda a Samuel Holland au poste d'arpenteur general en 1803 (Boudreau, 1991, p. 40), perpetue en gros le travail de la carte precedente. II accentue meme l'effet de la division administrative du territoire en epurant sa carte de details geographiques «superflus». On y observe en effet qu'un minimum de caracteristiques physiques et humaines. Le cartographe cree ainsi un contraste visuel faisant ressortir davantage la logique discriminante et classificatoire de la representation cartographique. Claude Boudreau (1993), specialiste quebecois en cartes anciennes, considere une telle epuration informationnelle comme une pratique courante a cette epoque.  La carte de Bouchette met aussi en evidence les plus recents fronts pionniers qui s'ouvrent un peu partout sur la rive nord du fleuve. Bien que le canton regne en maitre dans ces nouvelles zones, la nature canadienne de la colonisation est parfois bien marquee par une toponymie francaise: les cantons Simard, Tremblay, Laterriere, Tache, Caron etc. au Saguenay;  les cantons Joliette, de Salaberry ou Morin en Outaouais et dans les Laurentides. On note aussi la presence de plusieurs cantons «anglais» (Rawdon, Chertsey, Brandon...)  31  qui sont pourtant  Et pour cause, car le systeme de canton avait pour but d'occuper les regions situees aux confins des vieilles  52  occupes majoritairement par des Canadiens. Deux consequences decoulent de cette ouverture «cantonnee» des fronts pionniers francophones sur le vieux Bouclier canadien. D'abord, i l devient clair que le Canadien n'est plus prisonnier de la vallee du Saint-Laurent comme le suggeraient a tort les cartes precedentes. Ensuite, la francisation du canton laisse supposer que les frontieres ethniques ne sont plus aussi rigides qu'au depart, ce qui s'explique en partie par la faiblesse de l'immigration britannique au Bas-Canada.  Toutefois, i l ne faudrait pas penser que Bouchette soit totalement sourd aux ideologies en cours. En fait, une lecture plus poussee de la toponymie permet de constater que le modele mosaique est toujours aussi dominant. Bouchette impose sur sa carte une hierarchie toponymique qui ravive les vieilles distinctions ethniques et spatiales. Ainsi, les seigneuries, canadiennes par leur origine, sont creditees d'une visibilite plus considerable que les cantons britanniques, quand bien meme ceux-ci porteraient-ils des noms canadiens. L a distinction entre les deux types d'organisations spatiales se retrouve meme dans le titre: «...the Townships district from the Seigneuries». L'occupation coloniale des marges: les «reductions» cartographiques de l'Autochtone et de 1'hybridite.  L'occupation effective des marges coloniales par les colons et les compagnies forestieres fut quelque peu anticipee par les cartes sous le regime britannique, ce dont temoigne la disparition visuelle de 1'Indien. D'abord, la visibilite materielle de l'Autochtone est a peu pres nulle sur les trois cartes. II est surprenant de ne voir aucune mention des nations autochtones sur la carte de Thomas Kitchin (Figure 2.8), car i l est encore de coutume de faire ainsi a la fin du x v i i f siecle . Pareillement, on s'attendrait a voir poindre sur la carte de Bouchette (Figure 2.10) 32  les premieres reserves indiennes issues de la loi de 1851-1853, Acte pour mettre a part certaines  etendues de terre pour 1'usage de certaines tribus de Sauvages dans le Bas-Canada (Fortin Frenette, 1989, p. 31). A u bout du compte, seul Samuel Holland semble laisser a l'Autochone une petite place, toute relative par ailleurs. Les quelques villages autochtones qu'il mentionne sur la carte (les reserves contemporaines de Kanesatake, de Caughnawaga et d'Odanak) se perdent  seigneuries francaises, regions toujours inhabitees (dans une perspective coloniale bien sur) a l'epoque.  53  dans la densite graphique du document (Figure 2.9). Bien sur, la presence de portages dans la region du Temiscouata laisse presager l'importance toujours aussi accrue du fait autochtone sur cette carte (Figure 2.11). Le plus important de ces portages, celui-la meme d'ou derive le nom de la region (Timiscouata Portage), est particulierement bien represente, beneficiant d'un trait aussi large que celui accorde aux plus importantes routes coloniales dans la vallee du Saint-Laurent. L'influence autochtone du portage est toutefois temperee par deux realites. D'abord, remis a l'echelle de la carte entiere, ce portage reste plutot marginal. Ensuite, la mention d'un ferry a l'extremite est du portage (au lac Temiscouata) semble rappeler qu'au bout du sombre «tunnel» primitif brille une lumiere civilisatrice.  L'effacement autochtone s'effectue aussi sur le terrain symbolique. L a diminution des toponymes d'origine amerindienne et leur remplacement par des noms de lieux plus fideles a la prise de controle du territoire par les autorites cartographiques est un processus graduel mais constant. Ainsi les toponymes indiens, relativement nombreux sur la carte de Kitchin, disparaissent-ils presque totalement sur les cartes de Holland et de Bouchette, remplaces tantot par la toponymie anglo-saxonne du premier (laquelle se surimpose aussi passablement a la toponymie frangaise) et tantot par les noms de lieux majoritairement francais du dernier.  Comme de raison, avec la chute de la visibilite autochtone, suit l'effacement de toute l'hybridite cartographique observee sur les cartes de la Nouvelle-France. Avec le recul constant des toponymes autochtones, l'image geographique se fait davantage homogene. S'ajoute a cela une diminution evidente et constante des zones suggerant l'apport informationnel autochtone. Si la carte de Kitchin offre encore a voir bon nombre de ces zones, au nord comme au sud du SaintLaurent (y compris a l'interieur de la Province), celles-ci se limitent, sur la carte de Holland, a une mince barre situee entre Quebec et le Saguenay, pour enfin disparaitre completement sur la carte de Bouchette. Ce dernier prefere cacher avec des «blancs» ces zones qui etaient auparavant propices a la geometrie autochtone. L a cartouche de Samuel Holland semble jouer un role semblable, masquant ainsi le Nord de la Mauricie et l'Ouest du Lac-Saint-Jean.  Evidemment, ces zones ou l'hybridite cartographique se concentrait, sont aussi celles ou avait lieu l'essentiel des interactions interculturelles sur le territoire quebecois jusqu'a la moitie  A preuve, Kitchin fait lui-meme mention des «anciens Hurons» sur sa carte de 1769.  54  du XIX siecle, c'est-a-dire ces regions que Ton a auparavant qualifiers de marges coloniales. D s  est done utile de rappeler que la marginalisation de lTndien (sa mise en «reserve») et de la geographie metisse fut une realite cartographique et mentale avant meme de devenir un fait geographique observable. * **  En definitive, la carte du Quebec britannique et pre-confederatif se revele done le discours colonial le plus fidele a la primaute des idees anti-metisses. Alors que les marges coloniales s'articulent toujours autour de 1'interaction culturelle et d'une certaine ouverture a l'Autre et que la place publique engendre la conception mentale d'un Canadien metisse (ou «degenere»), la carte persiste a presenter la «purete» des separations ethniques et geographiques. Contrairement a ces contemporains, 1'arpenteur-cartographe n'a jamais cesse de suivre la logique «de la ligne a tracer».  2.4  Conclusion.  La carte a joue un role non negligeable dans le rattrapage mental et 1'intellection d'une realite incontournable dans le contexte de la traite des fourrures, la geographie metisse. C'est une realite, a notre connaissance, sur laquelle peu de specialistes se sont penches jusqu'ici. Bien que la conception coloniale du metissage reste partielle — la France comme 1'Angleterre hesitent a reconnaitre le «sauvage» en elles — la carte expose une influence autochtone qui s'affiche comme une porte ouverte sur l'univers materiel, politique et symbolique de l'Autre. II y a deux raisons principales a l'apport de la carte dans l'intellection du metissage et de la geographie metisse. Premierement, comme nous l'avons laisse entendre dans le chapitre introductif, la carte repose sur l'image. Elle permet, comme toute image, de communiquer directement et dans toute son integralite rinformation ou le message qu'elle contient (Bertin, 1961). L'image cartographique est une suite de codes (graphiques et textuels) qui permet de voir et de rendre intelligible des interactions territoriales (physiques et/ou humaines) qui ne sont pas visibles autrement (Wood, 1992, p. 5, 7, & 108). Cela represente un net avantage dans 1'apprehension d'une realite nouvelle comme celle de 1'Amerique. Deuxiement, la carte est investie, par rapport  55  aux sources ecrites, d'une credibilite supplementaire qu'elle doit a la «neutralite» de sa 33  representation (Harley, 1989a, p. 7). A titre d'exemple, la description que les jesuites font de l'influence de 1'Indien sur le Francais se fait sous le signe de la reprobation et trahit une prise de position ideologique. Pour sa part, l'image hybride presente cette influence de maniere impartiale, comme un fait pur et simple, comme une realite implaquable (Black, 1997, p. 12). Dans 1'ensemble, la carte aura favorise ce que Denys Delage appelle «la connaissance de l'Autre» et la remise en question des paradigmes europeens (1995, p. 55).  Cela dit, la remise en question de ces paradigmes, ceux a la source de l'ideologie des peuples primitifs par exemple, ne fut pas un processus soutenu et coherent. En consequence, la representation cartographique et coloniale de la geographie metisse fut constitute d'autant d'avancees et de reculs. D'une part, la presence du fait francais sur la carte du Nord-Ouest a amplifie la visibilite cartographique de la geographie metisse. En ajoutant a l'intrication symbolique du territoire et en occupant l'espace laisse vacant entre le monde autochtone et 1'incursion euro-canadienne, le fait francais etablit un lien entre cette geographie d'entre-deux de facture recente et cette geographie metisse plus ancienne qui decoule des alliances francoindiennes au temps de la Nouvelle-France. Ce lien est d'autant plus frappant qu'au moment meme ou cet episode metis originel participe au developpement de l'ethnogenese proto-metisse du Nord-Ouest, i l subit les assauts repetes d'un anti-metissage colonial de plus en plus oppressant a mesure que se concretised le retrait des activites de la fourrure et la progression des fronts pionniers sur le territoire quebecois. U est tout de meme ironique que la source premiere du metissage au Canada soit aussi la plus sujette a l'assechement.  Evidemment, la «neutralite» de la carte n'est qu'apparence. En fait, la carte est un systeme de «connaissances situees» (situated knowledge) et n'est done pas permeable au contexte socio-culturel et ideologique dans lequel prend place sa production (Harley, 1988, p. 278-279). On peut done considerer la carte comme un texte au meme titre que les commentaires jesuites (Harley, 1989a, p. 3; Harley, 1989b, p. 84). 33  i-Ti /  n j / it /1 it a / u,  7 / . <»; -1.19 j  t»o ;  ;~  Q  Figure 2.1  Sr de Champlain, Carte de la Nouvelle-France augmentee depuis la derniere, servant a la navigation, 1632.  Sh^dO J6 3 5 <3«3  lie  57  W vE N L  Figure 2.2  Guillaume DelTsle, Carte du Canada ou de la Nouvelle France et des. Decouvertes qui ont ete fades..., 1703.  A N D  58  Jei dCerreins  On J  ei  »<>if -Ut J#auBu J**,** offuUrs  »  ******  S^'r-Uur. Jute tu Hi'nitres  Us  it  I  ..^vyh-^  s i s *-  ;  I* - C"'  .'fcl  Li ^•3 7.  '-> "NT  / Figure 2.3  Philippe Buache, Carte physique des Terreins les plus eleves de la Partie |  j § }£,  \  Occiden du Canada..., 1742. Ie  GO  o  Figure 2.4  60  Figure 2.5  Peter Pond, Copy o / a map presented to the Congres by Peter Pond, a native of  t  Milford  in the State of Connecticut...,  1785  61  Figure 2.6  Portion de Aaron Arrowsmith Map Exhibiting all the New Discoveries in the Interior Parts of North America..., 1802.  62  Figure 2.7  Portion de David Thompson, Map of the North-West Territory of the Province of Canada...,  1813-1814.  Figure 2.8  Thomas Kitchin, A New map of the Province of Quebec in North 1764.  America...,  Samuel Holland, A new map of the Province of Lower Canada describing all the Seigneuries, Townships...,  1829.  65  Figure 2.10  Portion de Joseph Bouchette, Map of Lower Canada shewing the proposed land agencies and the townships district from the seigniories, 1857.  f  67  Cartography and the Vecu in the West — 1819-1895.  Chapter 3:  1  [...] nous savions qu'un certain nombre de Metis hivernait dans la region des rivieres Tortue et Souris de sorte que nous ne pouvions point esperer une bonne chasse dans ces environs-la. Done, nous avons decide de prendre une direction mediane allant vers le sud-sudest pour changer plus tard vers le sud-sudouest. Cette derniere route nous menerait par le lac des Branches, les Buttes des Trous, le lac du Diable, les Petites Fourches de la Sheyenne, le lac du Bois Blanc et la Maison du Chien. — Pere George Antoine Belcourt  2  Father George Antoine Belcourt describes more than a Metis geography here. His words portray a socio-cultural appropriation of space that placed the Metis at centre stage and was 3  little influenced by such geographical considerations as the British-American boundary. Moreover, in describing the buffalo hunt and listing toponyms the Roman Catholic priest depicted the material and symbolic attributes of Metis territoriality. As components of Metis orality, these toponyms also outline the importance of oral tradition in defining the Metis sense of identity and territoriality.  One of this chapter's objectives is to explore Metis oral richness and to show how it accounts for the unique Metis experience of space. The second part of the chapter is devoted to Metis narratives — stories, songs, and toponyms — and reveals them to be important subjects for understanding the Metis sense of territory and identity. But before treating Metis oral perspectives, I will first consider how Metis territorial makers find their place in the colonial cartography of the Canadian West.  The juxtaposition of those two very different objects of investigation — colonial maps and Metis oral tradition — should yield contrasting and complementary views of Metis  1  A portion of this chapter has been published in RIVARD, Etienne (2002). "Territorialite metisse dans le Nord-Ouest  canadien au xrx siecle: Exploration cartographique et toponymique", Cahiers franco-canadiens de l'Ouest, 14 (1e  2); 7-32. Copy of a letter written in 1845 found, in Dossier "Charrette, Guillaume", Provincial Archives of Manitoba, 2  MG9A6.  By using the term "Metis," instead of "Prairie Metis," I am aware that I may foster the idea that "Metis" exclusively applies to those who live in today's Prairie provinces. This is not my intention. My use of "Metis" in this chapter includes the Metis born in Riviere-Rouge and the Metis groups who emerged outside Riviere-Rouge and were closely tied to the fur trade economy, such as those of Lac La Biche or Lac Sainte-Anne for example. The latter 3  68  territoriality. It should also display two major phases in Metis territoriality: 1) its development, expansion, and dominance from the early 19 century to the Riviere-Rouge uprising in 1870; and th  2), its gradual erosion with the rush of new settlers after that date. Overall, as Antoine Belcourt's comment suggests, these two analyses establish the Metis reality as a dominant feature of the th  geography of the Northwest for most of the 19 century.  3.1  Cartographical inscription of Metis territoriality: from recognition to erasure.  Colonial maps of this period may be categorised by their function. I note four distinct functions, which, of course, are not totally exclusive. I made my map selection in order to represent all of these categories. Thefirst funtion is economic. Not surprisingly, the maps of the HBC are predominant as they principally focus on the fur trade routes and organisation of the different districts within which the company operated (Ruggles, 1991). In that respect, the principal function of HBC maps did not change very much from the early mapping of this part of the world to about the 1840s when cadastral surveyings became more important (idem, p. 96). Peter Fidler's A Map of Red River District (1819) is an example. The second function is scientific. The Canadian and British colonial gouvernments launched a series of scientific expeditions in the mid-19 century to evaluate the possibility of colonising the Northwest. I present one th  example of this type of map, S. J. Dawson's Plan Shewing the Region Explored by S.J Dawson and His Party Between Fort William, Lake Superior and the Great Saskatchewan River (1859). Land  management is another characteristic function of this period. Maps produced for this purpose were intended to serve as tools for supporting the land tenure system and for organising the division and allotment of the land to settlers. These maps represent a huge portion of the cartographic production of the Department of the Interior created in 1871. They were produced in a variety of scales, representing areas as diverse as parishes (Beaudry's Plan d'une partie de la Riviere Rouge comprenant lesparoisses de St.Boniface, St.Vital, St.Norbert et Pointe-Coupe d'apres les  operations de 1871), provinces (Russell's map of Manitoba, 1871), and the whole Northwest. Finally, some maps of this period were promotional in nature; they were used to present a specific perspective of the Northwest to eastern Canadians. These maps generally advantageously portray the political situation and the advance of colonisation by emphasising  Metis are often refered to as "proto-Metis" (see chapter 2). It is understood that Red River Metis had been protoMetis as well.  69  colonial control over the land, railroads, and other lines of communication (colonial roads and telegraph). They were part of a propagandist discourse that aimed to recruit new settlers and to prevent them from settling in nearby American settlements (Courville, 2002, p. 451 & 463-466). These maps were produced or distributed by diverse agencies, public — the Office of the Minister of Agriculture in the case of Holland's Map of the Seat of Mel's Insurrection (1885) — or private like the "Canada Bank Note co.", printer of Map of part of North-West Territory shewing the Locality of the HALF-BREED Rebellion. The Canadian Pacific Railway, the Roads, Trails, and Telegraph Lines (1895).  Whatever their function, maps were essential to colonial authorities in the economic, political, and social management of the Northwest. Moreover, and except for the HBC'S maps — which were generally produced to assist the governors' decision-making — most of these maps were broadly distributed. Their message reached a large audience.  The maps under investigation here reinforce the observation made in the previous chapter that the inscription of Metis territorial markers is an ongoing process. Unlike the maps of the previous chapter, it is easier to distinguish between M&is and Canadien territoriality on the 19  th  century maps; Metis territorial markers on these maps are generally clearer. More interesting is the fact that Metis visibility increases on maps the function of which had nothing to do with the representation of Metis reality. These maps were conceived to serve colonial purposes, such as the appropriation of "empty" and fertile space for colonisation (Friesen, 1987, p. 109; Sprague, 1980b, p. 74).  The Northwest is a vast area and it took many years for colonial authorities to get a fairly precise picture of it. The Metis, along with the Indians, contributed in various ways to the colonial acquisition of knowledge about this region. First, like their voyageurs relatives, the Metis were important components of the commercial and scientific explorations launched in the "unexplored" Northwest over the years. Irene Spry mentions several of these expeditions in which Metis were employed as paddlers, guides, and interpreters: the Rae, Palliser, and Hind expeditions, the Boundary Survey of 1872-1876 (1985, p. 106-107). The geologist Henry Youle Hind made many comments during his expedition of 1858 about the Metis ability to provide geographical information (1860. v . l , p. 181). The scientist added: "I had an opportunity of meeting, at this isolated settlement [Portage la Prairie], with John Spence, a Cree half-breed of great experience in Rupert's Land. He drew a small chart for me, showing the position of what he called "coal" on the Assiniboine" (idem, p. 143). This observation is particularly interesting for it  70  stresses Metis mapping contributions to the colonial representation of the land, a fact that has not attracted the attention of the specialists on Aboriginal maps. In communicating first hand 4  information, translating Indian knowledge, and providing means by which mapmakers could travel across the prairie — the carts and the trails (Hind, vol. 1, 1860, p. 147 et 154) — the Metis sustained and stimulated the colonial mapping of the Nord-Ouest.  Those who wanted to understand this land — voyagers, explorers, fur traders, or mapmakers — inevitably confronted and recorded the Metis presence. From the early 19  th  century to about 1870, the Metis gradually affirmed their prominence in forging the geography of the Nord-Ouest and imposed themselves as an inescapable reality. This is the reason why the inscription of Metis territorial markers figures so prominently on colonial maps.  But the situation changed in the last decades of the 19 century. It is broadly accepted that the Metis way of life was drastically modified following their defeat at Batoche in May, 1885. The post-1885 era was marked by the restructuring of Metis society after the material, political, and symbolic changes brought by the Canadian colonisation of the plains. The depletion of the buffalo and withdrawal of the fur trade affected the Metis traditional economy (Purich, 1988, p. 106). The north-south trade route, furrowed by Metis cart wheels, became obsolete with the implementation of the Red River steam-boats; the arrival of the railroad had a 5  similar effect on Metis east-west trade (Payment, 1983, p. 134).  The political situation of the Metis was not any better. With the rush of new settlers from Eastern Canada and Europe, in a few years the Metis were demographically marginalised within  Over the last two decades, there has been an extensive literature on Aboriginal maps within the history of cartography. Volume two (book three) of the collection of The History of Cartography, initiated in 1987 by J.B. Harley and D. Woodward, is entirely devoted to Aboriginal maps produced around the world. Another important contribution to the field is Malcom Lewis' Cartographic Encounters (1998), an edited book that focuses on Native mapmaking traditions of the Americas. To my knowledge, however, never has any scholar treated in detail the mapmaking traditions of mixed-blood populations. Indeed, Malcolm Lewis (1980, p. 17) mentions the Metis Joseph La France's contribution to the 1744 Arthur Dobbs' map published in An Account of the Countries Adjoining to Hudson's Bay..., but he never tries to demonstrate in what way La France's contribution was distinct from a Native one. There is perhaps no distinction. The work of Glen Fredlund et al. on maps of the Yellowstone and Milk Rivers area produced between 1877 and 1880 by Crazy Mule also comes to mind. Although the authors label the mapmaker as "Cheyenne scout," many portions of their paper suggest the mixed-blood reality of the mapmaker and other scouts. The maps themselves appear like hybrid documents. Crazy Mule combined Native use of name-glyphs and autobiographic techniques with European cartographic "accuracy" (rivers with their meanders) and conventions (northward orientation for example). Adding to this was the reinforcement of existing buffalo robe trade tariffs and customs regulations at the RiviereRouge borders with the US after 1870 (Ens, 1996, p 149).  4  5  71  the whole Northwest. Finally, the construction of the railroad and the influx of immigrants brought with them "train stations" and new settlements bearing names that mirrored the newcomers' realities. Many of the old names that once had covered the prairie disappeared from sight, lying underneath rails and crops.  But the portrait is more complex than this. The difficulties the Metis faced after 1885 were neither absolutely unfortunate, nor altogether new. First, in spite of a loss of influence and a process of marginalisation following the 1885 uprising, the Metis, as Diane Payment argues in her study of Batoche (1983, p. 136), were not left without any future, as earlier authors have suggested (Giraud, 1945, in, p. 1211; Stanley, 1936). Although the Metis were partly dislocated after the battle at Batoche, Payment suggests that they adapted to the new socio-economic reality in the West, and that some of them — Xavier Letendre dit Batoche, Salomon Venne, George Fisher and a few others — took advantage of these changes to prosper (1983, p. 132; 1986, p. 174-175). Second, the post-1885 effects were not unfamiliar to the Metis. At best they were the culmination of an ongoing process that started in the 1870s with the creation of Manitoba. As Nicole St-Onge points out, "By 1880 the farming Metis were no longer an elite in Manitoba. The high rate of immigration of non-Metis to Manitoba caused them to lose most of their political clout and they rapidly became a powerless minority within the province and within the FrenchCatholic group" (St-Onge, 1985b, p. 160).  Many of these Metis would then leave Manitoba to settle beyond the limits of colonisation. Two sections of the Manitoba Act dealt with the question of land: ss. 31 et 32. S. 31 stated that 1.4 million acres of land would be put aside for the children of heads of families to extinguish "Indian Title." To meet its responsibilities to these children, allotments, surveyed in 6  quarter section and mostly situated behind existing parishes (Flanagan, 1991b, p. 3), were to be allocated by lottery. In 1875, the Federal government started issuing scrips to the heads of families, redeemable for 160 acres of Dominion Lands or $160. S. 32 concerned the confirmation of the land tenures allocated before 1870 and applied to both Metis and white settlers. In any case, surveys needed to be conducted and patents issued to secure Metis land  In the first place, Dominion Land authorities interpreted s. 31 in a very "liberal" way, including not only the children, but the adults as well. Since the Dominion census of 1871 revealed that there were about 10,000 Metis in Riviere-Rouge, the 1.4 million acres was to be divided into allotments of 140 acres. In the course of the process, however, Dominion Lands came back to the initial wording of the Manitoba Act and only "the children of the half-  6  72  possession. However, it turned out that only a few Metis had succeeded (or showed interest) in settling their piece of land, for more of them migrated westward.  Just why and when this migration started is a matter of contention among contemporary scholars. The historian Douglas Sprague, who was hired by the Manitoba Metis Federation, argues that the Federal authorities were responsible for the Metis land dispossession (1980a; 1980b; 1988; 1991). Beside the racist pressures the Metis faced after the 1870 uprising, the process put in place to deal with their land allocation was delayed in many ways and put the Metis in an untenable situation. It took more than a decade for most of the patents to be issued (Sprague, 1980b, p. 80). In the meantime, the federal Department of Justice had voted amendments to the initial s. 32 (occupancy before 1870) of the Manitoba Act to increase the requirements for land-in cultivation (idem, p. 77). If such requirements did not affect full-time farmers, they could not be met by most of the Metis, who spent part of the year hunting buffalo, which, year after year, grazed further away from Riviere-Rouge. In such conditions, discouraged and threatened by the loss of their land, many Metis sold their allotment or scrip for cash. A speculative land market emerged in which the Metis were the losing party (Sprague, 1991; Tough, 1996, p. 116), a situation the Federal government was unlikely to ignore in spite of its inaction in preventing it in the first place (Ens, 1983, p. 10; Sprague, 1991, p. 147). Thus, it is 7  clear to Sprague (1988) that the Federal government had made every effort, either actively or passively (by maintaining a state of "lawlessness"), to dispossess the Metis and bring them under its control (Sprague, 1980a, p. 421). Nicole St-Onge supports this argument in her thesis on the Metis of la Pointe-a-Grouette,  in which she concludes that the amendments to s. 32 and  speculation were the main reasons for the Metis dispersal (1985b, p, 163).  For the American political scientist, Thomas Flanagan, "Sprague's thesis shares the problems of all attempts to infer motives from results. It quickly descends into the labyrinth of conspiracy thinking, where things are never what they seem" (1991a, p. 6). Flanagan — who was hired as historical consultant by the Department of Justice to discuss the Manitoba Metis Federation land claim case — affirms that the Metis had never been deprived by Ottawa and  breed heads of families" were eligible for land grants; the allotments were enlarged to 190 acres accordingly (Flanagan, 1991b, p. 3-4). According to Gerhard Ens, the provincial government acknowledged speculation in Metis lands in the early stage of Manitoba and voted the "Half-breed Land Protection Act" in 1873 to prevent it (Ens, 1983). No similar legislative action was put in place by the Dominion.  7  73  speculators, and prevented from getting land (idem). Flanagan supports his argument in two ways. First, he draws on the historian Gerhard Ens to present evidence of the Metis ability to 8  adapt to new economic situations. According to Ens, and to Flanagan, the Metis were anything but the poor victims of the "evils" of western capitalism (Ens, 1996, p. 5). They had integrated perfectly well its logic and functioning to develop, from the 1840s, new markets and profitable activities (Ens, 1988; Ens, 1996; Flanagan, 1991a, p. 17; Flanagan, 1998). In fact, Ens (1988; 1996) suggests that the migration started before the creation of Manitoba and was the result of the development of the buffalo robes trade in which many Metis were involved, and consequently, the necessity, from the 1860s, to winter outside Riviere-Rouge as the buffalo progressivly withdrew westward (Flanagan, 1991b, p. 9). Second, Flanagan tries to quantify the debate, and to show that the Metis were quick to get rid of their land (1991a, p. 65 & 121-127). For him, most of the Metis understood very well the speculative nature of the market, got a fair price for their land (Flanagan, 1991a, p. 122; Flanagan & Ens, 1994), and reinvested it in their "traditional" activities such as freighting and hunting (Flanagan, 1991a, p. 15).  Douglas Sprague (1991) answers these critiques with supporting data. First, he opposes Ens' views with regard to Metis migration from Riviere-Rouge before 1870. While Sprague admits that old parishes became crowded in the 1850s and 1860s, he suggests that many Metis moved near to newer parishes (and not outside the settlement) and criticises the fact that Ens makes generalisations from a study based on the comparison between St.Andrew's and SaintFrangois-de-Xavier,  old Metis parishes (idem, p. 139). Second, Sprague compares the Dominion  census of 1871 and the notes of the commissioners who took Metis affidavits in 1875 to show that most of the Metis had not sold their allotment or scrip in the first years (idem, p. 141), but had rather held it (at least until 1875), "waiting for the terms of the 'Manitoba treaty' to come into effect" (idem, p. 151).  Although it is not within the scope of this thesis to take a position in this debate, my analysis of colonial maps reaches a conclusion closer to Sprague's. Maps of this period tend to show the colonial "intention" to take over the land, especially after the creation of Manitoba. 9  For the first time Euro-Canadians were engaging themselves in colonisation with the intention of occupying the whole Northwest. In comparison, the Selkirk colony was a much more modest  Ens has also regularly appeared as a government expert in Metis issues.  74  enterprise, spatially limited by the fur trade. For colonial authorities, the Northwest was to become a land for agriculture. Using the land otherwise — in a "primitive" way — was to waste its potential (Sprague, 1991, p. 151). In fact, any Metis who fit this Euro-Canadian agricultural scheme had little problem to secure a piece of farm land for himself and his family (idem). One can review the amendments to the Manitoba Act regarding the cultivated land requirements less as a way to disqualify Metis applications than as a means "to separate the wheat from the chaff among the mixed-blood population. As noted above, while these requirements were generally met by full-time farmers, they were an impediment to those Metis who were also involved in the buffalo trade. Accordingly, and as I will show, the colonial maps exhibit the gradual extinction of Metis territorial markers and their replacement by colonial ones. What is more interesting is that not only do the maps depict the growth of colonisation at the expense of a "primitive" way of life, they sometimes predict it. Given their prospective quality, maps may be ahead of the spatial reality they intend to represent and, therefore, they may convey colonial intentions or spatial ideologies (Clayton, 2000; Harley, 1988). Thus, they can also be considered as good indicators of the federal "responsibility" in the territorial dispossession and dispersal of the Metis, another way to say that the maps convey, to a certain extent, colonial anti-metis sage discourses.  3.1.1  Material territorial marks: taking over the land.  The material dimension was the best way for colonial cartographers to mark the difference between the land-wasting primitive way of life and the more prosperous and effective colonial use of land. Colonial maps of the West clearly demonstrate this. On the other hand, these maps also allow much room for the representation of material elements of Metis territoriality, in spite of the fact that they had always been made to emphasise primarily the EuroCanadian perspective of the land. The increasing demographic importance of the Metis, and the specific socio-economic niche they had found for themselves within the Northwest economy, especially before 1870, had much to do with this reality (Ens, 2001, p. 162; Foster, 1985). Two types of Metis material markers prevail on colonial maps: transportation networks and land tenures.  9  "Intention" is understood here less as a planned course of action than as a result of spatial ideologies.  75 Transportation systems: Metis' tentacles in the Nord-Ouest.  In the Northwest, as in other parts of early Canada, rivers and lakes comprised the primary transportation system (Tough, 1996, p. 44-62). Considering the contribution of the Metis in the fur trade geography and in "the development of Northern water transportation," Diane Payment says, the region "could be considered a distinctive Metis sphere of influence (2001, p. 159). A s I showed in chapter 2, portages and rivers are good, although indirect, indicators of the Metis mastery o f the water system. Despite the fact that Metis shared these spaces with their Canadien and Indian relatives, the portage pinpoints map metissage in action, and identify emerging Proto-Metis groups. Again, it is no surprise to note the prominence of French place names in-northern water systems: " L . des Isles," "Portage C r . , " "Point a la Mittasse," "Bonnet Portage," "Rivier L'Oiseau," and " L a Barriere" are relevant examples on the 1858 S.J. Dawson's map (Figure 3.1). The cart trail network, however, was of undoubted Metis origin. Although the network borrowed to a great extent from the pre-existing trajectories of Native people (Letourneau, 1978, p. 15; Ray, 1998; Spry, 1983), it was generally associated with three basic elements of Metis economic life: the buffalo hunt, pemmican production, and trade. Put another way, the cart trails were essential to the Metis material existence. Most of the maps of the Northwest produced in the 19 century represent cart trails. The th  representation is particularly explicit on Dawson's map, which presents the cart trail infrastructure in all its complexity and extension. This infrastructure appears as a dominant 10  material element of the region. The map exhibits a network that spreads its reach along the most important rivers such as the Red (Saint-Paul's Trail), Pembina, and Assiniboine (Carlton's Trail), and also on the plains, as marked and labelled by the term "Hunters Tracks," to the north-west of "Turtle Mountain" or near the Pembina Hills (southern Manitoba). The fact that, beside the rivers, there is no alternative means of transportation in the Northwest accentuates the material and cartographic importance of the cart trails. It is likely that such emphasis of the Metis material importance was not Dawson's intention. The mapmaker and scientist had a responsibility to document the Northwest's suitability for colonisation. The identification of transportation routes  was only one element of his physical description along with the recognition of land improper for cultivation such as the swampy areas near Lake Winnipeg. Whatever Dawson's intentions, the visual effect remains. The cart trails prove the Metis presence. Land tenure: le paysage de la Riviere-Rouge.  The first official system of land tenure was made of long, narrow strips with a water frontage, the river lot. This system was proposed by HBC authorities who asked their servant and mapmaker, Peter Fidler, to survey it in 1813 while he was chief factor at the Brandon post (Warkentin & Ruggles, 1970, p. 183-184). The 1819 Fidler map (Figure 3.2) reveals the first river lots,to be surveyed in Riviere-Rouge, located on the western bank of the Red River (on "Frog Plain") and on the two sides of la riviere la Seine.  In spite of their colonial origin, it is safe to say that the river lots are an important type of Metis territorial mark. The Metis adopted the river lot system and implemented it outside Red River, notably in northern Saskatchewan in the late 1870s-early 1880s. When, following the creation of Manitoba, surveyors again measured and established river lots along the RiviereRouge (idem, p. 233), it was not to satisfy the needs of newcomers, but simply to secure the 11  land occupied by the Metis and white settlers before Manitoba. As the Metis were demographically dominant at the creation of Manitoba, one can argue that s. 32 of the 12  Manitoba Act and the resurveying of river lots first aimed to accommodate the Metis population. Wherever the Metis were found cultivating lands, river lots were generally surveyed (idem, p. 234). The necessity of resurveying these lots also highlights the relatively weak influence of the prior colonial legal and spatial structure on Metis land occupancy. Douglas Sprague notes that before the creation of Manitoba the division of lots was based on both formal and customary boundaries. If the former boundaries were fixed by surveyors — as Peter Fidler and George Taylor had done for the older parishes — the Metis people determined the latter. The Metis 13  This is no surprise, for the trails were essential for expeditions launched into the prairie (Hind, vol. 1, 1860, p. 147 et 154). The HBCfirstresurveyed the lots in the 1830s (Ens, 1996, p. 32). According to the census of 1870, near four-fifths of the Red River population was Metis (Sealey & Lussier, 1975, p. 95). Some archaeologists make a similar point. David Burley et al. show that Metis agricultural settlement sites — that of Petite-Ville, near Batoche for example — featured a lack of boundary definition and symmetry that does not suit 11  12  3  77  used the land they needed — they sometimes needed less than a whole lot — and such use was proof of ownership (1991, p. 143). Staying in Riviere-Rouge in 1857, Hind observed that many Metis occupied lots for which they held no official titles (1860, vol. 1, p. 190-191). From a colonial perspective, these Metis were squatters.  The importance of surveying the river lots in the early 1870s is made obvious by A . L . Russell's map (Figure 3.3) — recognised as the first official map of the province of Manitoba (Warkentin & Ruggles, 1970, p. 248). This is not surprising, given the fact that this map was made in order to prepare the land for the arrival of the new settlers by putting in place the Canadian surveying system. Although Russell did not map the river lots (the scale of his representation was too small), he clearly established the areas where river lots were to be surveyed. These areas are on the banks of the Red and Assiniboine Rivers, on the southern shore* of the Manitoba Lake (site of the Roman Catholic and Metis colony of Saint-Laurent), and along the "White Mud River" and "Rat R.," near Portage la Prairie. Over the decade that followed the creation of Manitoba, the progress in surveying resulted in an increasing number of large-scale maps, which allowed more detailed information about land and its division. In the early 1870s, J.A.U. Baudry [Beaudry], a Canadien surveyor employed under the direction of J.S. Dennis,  14  produced a large-scale map entitled Plan d'une partie de la Riviere Rouge comprenant les paroisses de St.Boniface, St.Vital, St.Norbert et Pointe-Coupe d'apres les operations de 1871  (Figure 3.4). The river lots are represented as they had recently been re-surveyed and their owner's names are indicated. Beyond the few lots owned by the churches or the HBC, most of them were assigned to the Metis, visual confirmation of their demographic and material importance in this part of the settlement (south of Fort Garry). Socio-economic assimilation and the railroad: to "put on track" the Metis material marginality.  As mentioned above, 1870 marked a shift in the mapping of the region. As the colonial involvement in the management of land increased with the advancement of colonisation, maps  the principle of property division central to the Euro-Canadian organisation of space. Accordingly, the author suggest that this rather loose spatial organisation may be linked to Metis social structure (1992, p. 49, 120 & 159).  78  became an efficient if gradual means of erasing the Metis reality. The erasure, however, was the result of a complex, not always conscious, array of cartographic manipulations. It was much more than just ignoring the Metis reality as the Russell map of 1871 shows. The Ontarian surveyor pays more attention to selecting the Metis material elements to be represented then to trying to put them entirely aside as if they had never existed or had disappeared. In fact, while Metis material markers are easily found in Riviere-Rouge, they are almost completely invisible on the prairie. Examples of the representation of the Metis include the river lots, the cart trails, and the mention, in the "statistical" inset situated in the upper-right corner, of the 1.4 million acres of land reserved for "Resident Half breeds." In the meantime, Russell's map does little, if anything, to represent the Metis prairie materiality. Indeed, the prairie is less the mapmaker's focus than is the Province of Manitoba. However, there is nothing in the representation of Manitoba to justify the fact that the province is not quite centred on the map, but is rather displaced towards the left. This has an obvious two-fold effect. First, it reduces space left for the prairie and Metis territorial markers and, second, it allows a better representation of the area dominated by Euro-Canadians. How intentional this was is hard to determine. But the visual effects are obvious. The Metis material reality in the prairie is less represented than its importance would suggest. A n incomplete representation of cart trails, pale and visually insignificant compared to the rest of the map, is all that is left from detailed cartographic representations such as those on Dawson's map.  Basically, Russell represents only one aspect of the Metis, the one that could assist the government to integrate them peacefully and successfully into Euro-Canadian society. He certainly pays little or no attention to the "wandering" life of buffalo hunters, which was 15  generally seen as a limitation to this integration by Euro-Canadians (Hind, 1860, vol. 1, p. 181182). This cartographic representation is at odds with the spatial reality of many Metis. Not only was their life on the prairie not yet extinct, but with the bison moving steadily westward, buffalo hunters and traders were compelled to travel further south and west, to spend even more time in the prairie and, consequently, to devote even less time to farming their lots in Riviere-Rouge (idem, p. 180; Ens, 1996; Grant, 2000, p. 109). Russell's map does not give an accurate picture of the material reality of Metis life. But the accuracy of this map should be measured less by its  Dennis was responsible for the making and implementation of the quarter section survey system in Manitoba. He was appointed Surveyor General in 1871.  79  accordance to a contemporary reality than by its capacity to mirror the land as colonists intended it to become. This map proved to be a fair conception of a future reality and well illustrates the prospective potential of maps. The mention of a "probable railroad route" in the upper-right of the map is another element of this prospective vision.  Beaudry's large-scale map of the old French parish river lots may be analysed in the same way. His focus is on the settled portion of Metis reality. There is no representation of the prairie. Rather, he surrounds Riviere-Rouge by some considerable blanks. Even the cart trails appear to connect only different parts of the settlement; they do not seem to be connecting Riviere-Rouge to the "outside" world. Moreover, his precise inscriptions of Metis names on each single lot give the impression of Metis compliance with a settled life. The impression may be justified for some, but not for the majority of the Metis as their subsequent dispersion attests. More importantly, this inscription of Metis names suggests that government did properly address the Metis land issue. Surveyors, as the map shows, ran the river lot boundaries in accordance with the terms of s. 32 of the Manitoba Act. Actually, they surveyed river lots even where they considered that Metis improvements of the land were insufficient. One can find a considerable number of large-scale maps of river lots in the NAC that represent new French parishes (SainteAnne, Saint-Laurent, Sainte-Agathe, Baie Saint-Paul, and Lorette), precisely where a greater  proportion of Metis were overlooked (Sprague, 1991, p. 141-142). In this light, it could be 16  argued that maps initiated the colonial mental conception that the Metis were offered a fair opportunity to secure a piece of land. Put another way, maps would confirm the colonial belief that Metis failed to secure land in Manitoba because they had no will to cultivate and/or because they left the province, not because they were dispossessed.  Another cartographic manipulation involves information density and visual attractiveness. Even when a considerable number of Metis territorial markers find their way onto a map, they can be hidden by a more striking emphasis on the colonial appropriation of land. This is particularly obvious on the 1895 Map of Part ofNorth-West  Territory Shewing the  Locality of the Half-Breed Rebellion... (Figure 3.5). Although Metis cart trails appear on this map, what first attracts the reader is the CPR, which sits on the lower section of the map and  I use this 19 century pejorative term to emphasise the Eurocentric perpective behind the mapping of Metis reality. For example: McPhillips & Whitcher (1880), and McPhillips & Doupe's (1875). 16  80 crosses it from left to right. The railroad stroke is darker and more pronounced, a reflection of its relative importance. The map insets and the cartouche add to the CPR visibility i n two ways. First, their positioning on the representation creates a visual corridor in the centre of which lies the railroad. Second, one of the insets is a Reduced Map Shewing Route to Winnipeg and NorthWest via Canadian Pacific Railway that connects the region to its "civilised" and more developed eastern province; the other inset is a table of distances that also shows the connection between Eastern Canada and the Northwest territory. A s the Quebec geographer Serge Courville points out, such maps were instrumental in promoting colonisation and were largely addressed to potential settlers (2002, p. 466). Their function was less to "erase" the Metis than to highlight (should I say "overemphasise") the steady and rapid progress of colonisation. The end result is the same. The Metis had become, cartographically speaking, a secondary element of the Northwest, although this was not yet the case.  3.1.2  * " ^ k;  !v  Political territorial marks: delineating the Metis sense of common identity? O f all its dimensions, the political component of Metis identity has caused the most ink to  flow over the years. The narrative thread of Metis history has been built on events that marked the evolution of a Metis sense of common political identity: the battle of la Grenouillere  (Seven  Oaks) in June 1816, Guillaume Sayer's trial in 1849, the creation of Manitoba, and the uprisings of 1870 and 1885 are among the most repeated examples (Friesen, 1987, p. 75-79, 100; Sawchuk, 1978, p. 24). In spite of its importance, however, the Metis political dimension is hardly perceptible on colonial maps. In the first place, mapmakers had little concern for the political dimension of the land as a whole. The centrality of the geography of the fur trade encouraged them to focus on the commercial characteristics of the region (Ruggles, 1991). But by the turn of the 1850s, Ontarians showed more interest in the fertile plains of the Northwest and started thinking about colonising and reproducing their way of life on such promising lands. While the colonial political dimension became more apparent on maps of this time, the Metis sense of identity was not directly inscribed on maps. Nonetheless, some maps reveal indirect evidence of Metis (geo)political characteristics. Implicit recognition of Metis distinctiveness is revealed by subtle elements and by the context within which maps were produced. The representation of the province of Manitoba, which is central to the Ontarian surveyor Russell's map (Figure 3.3) of 1871 is one example. The creation  81  of the province in 1870 marks the apogee of Metis political influence in the West, apparently making it impossible for "outsiders" to rule or take over the land without the Metis' explicit consent. The Manitoba Act is best conceived as a negotiated agreement or as an alliance, and necessarily implies the interdependence and mutual influence of the two parties. Even if the Metis conceded the Dominion's right to send its surveyors to prepare the field for the arrival of the new settlers, their political mobilisation awarded them many advantages, among which was the reservation of land for them. In the Dominion's plan, Manitoba was meant to be, at least in its beginnings, a territory with little autonomy that was to be principally administered from Ottawa. In forcing the creation of a province with its own legislative assembly, the Metis sought to ensure the continuity of their culture.  The 1885 Northwest uprising offers more indirect evidence of the Metispolitical influence. W.H. Holland's map of the Saskatchewan and Assiniboia territories, 1885, is a good example (Figure 3.6). Although the map does not clearly represent battle sites, its title says it all: Map of the Seat ofRiel's Insurrection. Considering that "insurrection" defines an organized opposition to authority, the title of this map indirectly confirms that the Metis were considered as a specific people with a certain military organisation. As the precise locations of the battles are hard to identify, the Metis "military" and geographic importance seems to embrace the whole region. Yet this is contrary to the general emphasis of the map, which is to assert the material mastery of civilisation as represented by the Canadian Pacific Railroad (CPR) and the telegraph line. The Metis are not intended to be dominant. The same comments can be made of the 1895 Map of Part of North-West Territory Shewing the Locality of the Half-Breed Rebellion.. .(Figure  3.5). Again, the importance of Metis political influence in the Northwest is revealed less by the visibility of the Metis on the map than by the title itself. Surveys and battlefields: measures of a political and military seizure.  The Metis may have been politically influential, but this is not what the mapmakers aimed to show. It is unlikely that Russell meant to portrait Manitoba as the ground of an alliance between primitive and civilised peoples. This was against colonial racial conceptions. If Metis existence is perceptible on his map — the river lots — it remains subordinate to the colonial reality. The river lots appear to be remnants of a much larger survey system dominated by the  82  township or quarter section, and the geometric, rational, and organised spirit of modern western society (Spry, 1983, p. 215). The Manitoba that is sketched on this map is not intended to be the site of the Metis Nation, but is rather the colonial sight of an empty land available to be "transformed into townsites and farmsteads—policed, surveyed, fenced, settled, and threaded with railroads and roads" (Spry, 1999, p. LX).  Such an exercise of re-examination can also be applied to the representation of the battlefields of the 1885 uprising (Figure 3.6). It might well be that any mention of the uprising is a demonstration of Metis influence in the Northwest; however, it has to be recognized that battlefields are shown to emphasise the loss of such influence. The use of the word "rebellion" is calculated. From a colonial point of view, the term expresses the "illegal" nature of the Metis uprising, legitimises its suppression, and validates the Euro-Canadian political (military) supremacy over the Northwest.  3.1.3  Symbolic territorial marks: soul and sense of territory.  Symbols are central to a social system as they generate the attributes that define a society's sense of identity, as well as those on which the formation and development of its territory depend. They also generate communication between individuals. The most obvious form symbols can take is language; as a particular way of thinking, language is really a "tresor commun" (Siblot, 1999, p. 15), a symbolic expression of a sense of identity. Naming the land: making a "home."  Toponyms are the predominant cartographic symbol. Not only do names have a meaning on their own, but they are also made of a system of signs, a spoken and/or written language. Naming and language provide a group of people with the means to distinguish itself in space (Akin, 1999, p. 34). Naming and its result, the toponymic nomenclature (the sum of a society's toponyms in a given time), are then an expression of a group's sense of place (Nash, 1999, p.  83  472) and, as Herve Guillorel would say, the process by which space is tranformed into territory (1999, p. 64).  Metis toponymy largely borrows, by way of translation, from original Indian place names. Given, however, the cultural and familial ties that bound Metis and their Native 17  forebears, I consider these translated place names to be authentically Metis.  Moreover, it must  be said that these toponymie transfers indicate yet again the basic Metis socio-cultural duality. Similarly, Metis nomenclature considerably overlaps non-Aboriginal toponymies, notably those of the voyageurs  canadiens.  French toponyms are most likely to originate with the Metis, especially on the prairie where French-speaking Metis were numerous (Dawson, 1859, p. 24; Hind, 1860, vol. 1, p. 179 & 181). One finds some of them on Fidler's map (Figure 3.2). The best known are "R. 18  Qu'appele" and "Portage des Prairie." Fidler also notes "Fort la Reinne" (near Portage-la-Prairie) — a French Regime toponym —, "Mountain L a Bo/se," and "Soore river" (Souris River).  19  There are also some French toponyms on Russell's map (Figure 3.3) — "Seine," "Isles des Bois," "aux Prunes," and "aux Marais" Rivers; "la Petite Montagne" — and Dawson's (Figure 3.1), as revealed by the "Butte au Carcajou" (at the forks of Assiniboine and Qu'appelle Rivers) and numerous tributaries of the Red River: "R. Salle" (Lasalle River); "R. aux Gratias" (Morris River); "R. aux Isles des Bois" (Boyne River); "R. aux Prunes" (Plum River); "R. du Marais"; "Riviere la Seine"; and "Riviere aux Roseaux". Unusual on the latter map is the use of the French generic "Riviere" in substitution for its English equivalent, "River," which otherwise abounds on the map. Dawson had probably transcribed the toponyms, the specific and generic terms altogether, as they were given to him by his Metis informants.  On the one hand, such a translation can be seen as the Metis recognition of their Aboriginal origins. On the other hand, it could also be argued that it was a way for the Metis to distinguish themselves, a way of appropriating First Nations territory. This is not to say that English-speaking Metis had not named the land on which they lived. However, if they did so, their toponyms were more than likely to be integrated into the colonial geography, which was anglophone as well. Therefore, English-language Metis toponymy is not well documented. The use here of English spelling is striking. Mapmakers would generally do so to transcribe Native place names. Actually, "Souris" might not be the French for "Mouse," but rather an unexpected Native name. According to Carol Leonard (2002, ad verbatim), the toponym could be a corruption of "Missouri." It appears that the Missouri and 17  18  19  84  That said, French toponyms are relatively outnumbered on colonial maps. Nothing prevents explorers, traders, outsiders or cartographers from translating into English the names they encounter. However, many of the original French and Metis place names have persisted until today. Dawson's map exhibits the best examples: the "Caling R i v e r " (Qu'appelle) and "Mouse River" (Souris). Both Russell and Fidler mention the "Stinking River," the contemporary name of which is the Lasalle River or was, as the Metis used to call it, la riviere Sale. Fidler also maps the "Reed River," which flows into the R e d River a few miles north of the US border and is known today as Roseau River.  Maps also include many translated toponyms that have not survived in the West, but for which Metis equivalents are known to have existed. These toponyms leave no doubt as to their Metis origin. Dawson mentions a place called "Where the bones l i e " (site of Regina, Saskatchewan), which was named "Tas d'Os" by the Metis (Morice, 1912, vol. i i , p. 328; Rondeau, 1923, p. 113). One can also observe, on Dawson's map, the "Jaw bone R . " This name is a corruption of the Metis term "la Machoire d'Orignal," or Moose Jaw as both a city and a river are named today. The "Scratching River," Metis' riviere aux Gratias (Morris River, Manitoba), or "Moose H i l l " — "la montagne d'Orignal" — are other examples (Charette, 1976, p. 15, 79 & 101). The Queen sits on a pile of bones: a symbolic infiltration. Few in Canada are aware that Louis Riel was hanged in a place the Metis traditionally knew as Tas d'Os (Rondeau, 1923, p. 113) — the "Where the bones l i e " on Dawson's map. The site was one where the Metis gathered to make pemmican, and the name was derived from the accumulation of buffalo bones (Baker, 1934, p. 5-6). In the 1880s, some Metis, especially those from la Montagne de Bois and La Talle de Saule, moved there to gather these bones (and those spread all over the prairie), which were sent to the United States to be powdered and processed into fertiliser (Rondeau, 1923, p. 112). This coincided with the arrival of the CPR. A train station was built on site and named Regina (Latin for "Queen") after Victoria "the Good." Regina, as the symbolic and geographic throne of the British Crown in the Northwest, is the best illustration of  Souris Rivers are very much alike in terms of their turbidity. They are also, geographically, fairly close to each other.  85  how names can be used to superimpose a dominant group's territoriality, and how they can become political issues for they imply control over the land and express the spatiality of power relations (Guillorel, 1999, p. 65; Kahlouche, 1999; Myers, 1996; Nuessel, 1992, p. 3). As the settling of the land had mostly proceeded from the expansion of the railroads and roads (Stanley, 1961, p. 185-186), the renaming of the prairie followed these same axes. The CPR was certainly the most important source of renaming. Along this railroad emerged train stations, most of which were given Euro-Canadian names. Another look of the 1895 Map of Part of North-West Territory Shewing the Locality of the Half-Breed Rebellion...  (Figure 3.5) confirms  this. If old toponyms such as Portage la Prairie, Qu'appelle, Grande-Coulee or Moose Jaw persisted and identified train stations or new settlements, most names were introductions to the Nord-Ouest. Bagot, Austin, Douglas, Sidney, Fleming, Herbert,, and Bowell are just a few examples of the colonial toponymie appropriation of Metis land. As these toponyms are in the more salient portion of the map, their relative importance is often obvious at a glance, leaving other place names in the dark. Given the promotional nature of this map and its role in attempting to attract settlers, one may suspect a certain colonial intention in this European toponymie emphasis that recalls the cartographic tactics used by British mapmakers in early 19 th  century Quebec (see chapter 2). It would be misleading, however, not to remember that the process of renaming had started years before the creation of Manitoba and the expansion of colonisation. As I pointed out above, Metis toponyms were largely translated into the mapmaker's language or replaced on early maps. The huge "Canada" that crosses the old prairie on Dawson's map is also relevant, as the region was not yet "Canadian" when the map was made.  * **  Colonial maps expose Metis geography as a dominant element in the Nord-Ouest, and reveal themselves as rich sources of Metis territorial markers. However, Metis territoriality can be only partially revealed by maps. The Metis territorial markers represented on maps are those that do not confront the Euro-Canadian geographical project, the integration of the Northwest  86  into the Dominion. As both creator and bearer of knowledge, the cartographic discourse produced by colonial authorities is based on views, representations, and languages (cartographic/scientific) largely foreign to Metis reality. Shaped by the discursive and rational logic of the map, Metis reality is first classified and situated, spatially and socially, as any other human (Indian, Euro-Canadian...) or physical (hills, rivers, fertile plains) feature to be mapped. Imprisoned in its specific localisation and category, Metis reality had become another element of colonial management.  3.2  Oral inscription of Metis territoriality: sketch of a primary writing.  Because of the cultural bias of colonial maps, Metis territoriality needs to be seen within another set of representations and perspectives. Metis oral tradition is one. I will now consider what this tradition can teach us about Metis territoriality. In so doing, I will use three major types of oral information: stories, songs, and toponyms. Exploring Metis oral narratives is not new, altough it is still a marginal component of Metis studies (Dorion & Prefontaine, 2001, p. 20). Many authors have offered their interpretations of such material, notably of the songs (Carriere, 199-; Cass-Beggs, 1967; Clemens, 1985; Complin, 1939). From this literature a multitude of insights emerge about what Metis life had been, particularly with regards to its organisation, its materiality (tools, transportation means, food, clothing or housing), and its cultural distinctiveness. However, as I noted in the introduction, territoriality has never been a major concern in Metis studies, and narrative analyses are no exception. Understanding the importance of the experience of space in the forging of identity as revealed by these narratives is a study that remains to be done — and is what I propose to do here. A central question emerges: how do Metis oral expressions mirror and/or shape their vecu or experience of space?  To suppose that oral tradition can provide indicators of a people's experience of space is to admit that narratives are discourses, and following from that, that they take part in the constitution and/or acceptance of a system of knowledge. Narratives present information that was first culturally selected, defined, and hierarchised by the representations (about space, and history), practices, codes and rules originating from and recognised by a people. Selecting and  87  hierarchising "reality" is a way to put some order in the world, to make it meaningful. As the world is constantly changing, so are narratives: "If we think of oral tradition as a social activity rather than as some reified product, we come to view it as part of the equipment for living rather than a set of meanings embedded within texts and waiting to be discovered" (Cruikshank, 1998, p. 41).  As I want to emphasise Metis territoriality, geographical knowledge (be it social and/or physical) is at the centre of my analysis. My understanding of geographical knowledge generally corresponds to what is labelled traditional ecological knowledge (TEK). TEK is defined as a specific set of geographical information based on practices and cognition shared from one generation to another through orality, and upon which is built the relationship between a society and its environment (Berkes, 1993, p. 1-3; Collignon, 1996;'Ruddle, 1993). TEK or geographical knowledge specifically tells us about land occupancy, which "refers to the group's collective sense of its own territory in relation to that of others. In a sense, it is defined by a summation of historic use of the area, and an expression of the breadth of the community traditional ecological knowledge" (Hrenchuk, 1993, p. 75). As this last quote shows, TEK benefits a society by conveying as much information about geography as about identity, collective memory, or historical-genealogical origins.  TEK offers more than spatially represented information; it is a representation of space in its own right. In fact, the definition of TEK is broadly similar to that of the map! It requires only a small step from this observation to conclude that oral narratives can be read as cartographic representations. This is an issue that has been recently debated by scholars working on the history of cartography, and has resulted in the revision and broadening — beyond its European basis — of their definition of what constitutes a map. Accordingly, an increasing number of studies focussing on "alternative" map productions, mainly those of indigenous people, have emerged over the last two decades and have tended to emphasise the distinctiveness of Native cartographic conventions (Lewis, 1998; Rundstrom, 1990). From this literature, one can identify three outcomes that are relevant to our understanding of Metis "cartography." First, Native people drew "conventional" sketch maps that European mapmakers reproduced by integrating  88  them into colonial maps/ Similarly, it is known that the Metis occasionally produced sketch 0  maps for Euro-Canadian mapmakers, as Hind's comments quoted earlier in this chapter indicate (1860. v . l , p. 143). Second, Native maps were mostly "simple" graphic fragments of a primary, rich, exhaustive, and orally communicated body of geographical information (Warhus, 1997). Little of this oral information but a few Native place names was transcribed on colonial maps. Finally, some scholars have suggested that non-graphic representations of space, such as oral narratives, are types of maps (Brody, 1988; Chatwin, 1987; Pearce, 1998, p. 159; Woodward & Lewis, 1998, p. 6). The intention of such a proposal was to avoid the Eurocentric bias of considering "primitive" people, who had not developed writing to which conventional maps belong, to be incapable of mapping. Some scholars oppose this proposition, as does Denis Wood in affirming that "songlines are not maps" (1993). For him, all peoples are not "map-immersed" (1992, p. 34). He considers that maps are artefacts "(so material, not oral) that can communicate over time and be consulted as needed. Wood offers a bounded definition of map, and rather neglects the communicative and temporal nature of oral tradition or TEK. He ignores the fact that memory and orality form, by their interaction, "mental artefacts." On the other hand, there is something suspect and culturally biased in supposing that mapping is a universal cultural activity just because it is part of a European tradition (Woodward & Lewis, 1998). It is as if a people have to be absolutely map-immersed — materially and/or mentally — to be considered equal to Europeans.  For this reason, I am hesitant to label Metis oral narratives as "maps." To avoid an ethnic bias, I think it is worth considering maps for what they are fundamentally:- a specific — if not exclusive — type of representation of space. On these grounds, comparison between maps and oral narratives is possible, even if one defines them separately. Narratives can be conceived as territorial representations that are made of conventions different from those understood by colonial mapmakers (Belyea, 1992), and that mirror the Metis' unique cognitive experience of space, or geographical knowledge.  These Native maps or sketches were generally drawn on ephemeral and fragile materials — birch bark, animal skins, snow, or sand — and were preserved because of their inscription on European cartographic documents.  89  3.2.1  Metis songs and stories: the words that bound.  Neither "stories" nor "songs" are single categories. I will use two types of stories: 1) collective stories, based on Metis collective memory and transmitted through the generations, some of which remain part of the oral tradition of contemporary Prairie Metis; and 2), accounts 21  relating to an individual's experience of space during the 19 century. B y no means are these th  two types of stories closed categories. Individual accounts may make many references to collective stories (Charette, 1976, p. 21). Songs can also be classified in two distinct categories. There are "original" songs that relate specific Metis experiences, and other songs that come from the old French repertoire shared with the voyageurs.  In order to analyse Metis narratives, I propose three types of investigation. First, I will observe how Metis experience of space is expressed through their oral tradition. Second, I will consider how Metis narratives shape the Metis experience of space. If narratives are to be understood as systems of knowledge, one must bear in mind that "just as traditional knowledge and its transmission shape society and culture, culture and society shape knowledge; these are reciprocal phenomena" (Ruddle, 1993, p. 18). And third, I will show how in-betweenness, the central element of Metis reality, forms overlapping spaces that challenge social and geographical "frontiers" delineating the Metis experience of space. This will reveal how ambiguous categories such as "we" and "other" are, and how rich and diverse the Metis vecu is. Before proceeding with the analyses it is important to recall that narratives are contextual productions that tell as much about the present as about the past. My examination relies on a relatively limited number of sources, which focus on individual's accounts, such as these of Peter Erasmus, Louis Goulet, Antoine Vermette, and Norbert Welsh. Although these individuals provide good insights into the collective nature of the Metis sense of identity and territory, one has to avoid generalisations; these Metis speak on their own behalf, not for the Metis as a whole. Moreover, while these accounts reveal a variety of experiences, they do not cover all dimensions of Metis life in the same way. First, the way in which these individuals identify as Metis differs from one to another. While Goulet, Vermette and Welsh clearly self-identify as Metis, Erasmus's sense of identity is only indirectly suggested. Second, their ethno-linguistic origins were  See, for instance, The Virtual Museum of Metis History and Culture held by the Gabriel Dumont Institute of Native Studies and Applied Research at <>.  90  different: Goulet and Vermette were Catholic French; Erasmus, whose Danish father worked for the HBC, was Protestant and had English as his mother tongue; and Welsh, although he adopted the religion of his French Metis mother and in-laws, kept strong links with his English background. Finally, none of these Metis were full-time farmers or fur trade employees; they were mostly involved in the free trade and buffalo hunt. Consequently, Metis individual 22  accounts offer a particular version of the past influenced by life contingencies and the context within which they were performed and "cannot, therefore, be taken word for word as a source of precise, factual information" (Spry, 1999, p. xiv).  But I am not suggesting that these accounts are fictional and exclusively personal. A l l of these Metis elders (Erasmus was 87 years old when he performed his narrative) were considered good storytellers. Hugh Dempsey, in the foreword of the most recent re-edition of Buffalo Days and Nights, says of Peter Erasmus that he  follows a Native practice in telling his story. When an Indian recounts an historical event, whether or not he was part of it, he describes it as though he was there, complete with conversations among the principal participants. [...] The reader should not think that this is fictionalizing the account—not at all. It is following the ways of his mother's people (1999, p. x). In other words, storytellers often describe events of a collective nature as if they had personally experienced them, "individualising" what was common. Their stories perhaps present "corrupted" versions of the past, but they often rely on real events that had collective meanings. Relating Metis experience of space: three areas of investigation.  While the broad components of Metis territoriality were similar everywhere in the Northwest, the Metis sense of identity and territory varied greatly in time and space. The Metis in the 19 century were not only those who hunted the buffalo and lived at Riviere-Rouge as th  As far as I know, there is no account from those Metis working as farmers or as tripmen and York boat operators for the HBC, other than contemporary ones — the Jasper- Yellowhead Museum and Archives, the Lac La Bich Mission Historical Society or Jean Morisset's Ted Trindell: Metis Witness to the North (1986). I do not kno reasons for this. One may suggest, however, that the "free life of the prairie" was a more interesting reality to depict for both the storytellers and the interviewers in the early 20 century when all these accounts were recorded. th  91  early studies had shown (Slobodin, 1966). The Metis were also at home in the parklandZ3  woodland environment, along fur trade routes, and near the posts (Tough, 1996, p. 5). It is in such environments that groups of proto-Metis had emerged and eventually formed most of the Metis population of the Riviere-Rouge (Foster, 1985). The Metis sense of identity and territory was derived, for example, from areas as distinct from each other as Hudson Bay and riviere de la Paix.  The explanation of such diversity has inaugurated an important debate (Dorion & Prefontaine, 2001, p. 29). On the one hand, Frits Pannekoek (1976), a University of Calgary specialist in Metis studies, has suggested that ethno-linguistic and religious differences created an intrinsic cleavage between the Metis and the Half-breeds at Riviere-Rouge.  His argument is  that very few English-speaking Metis supported Louis Riel during the 1870 uprising. On the other hand, Irene Spry thinks that Pannekoek's argument is shaky at best for it does not take into account the numerous family ties existing between French and English mixed-blood populations at the time. She also suggests the inadequacy of building a study exclusively on clerical records, as did Pannekoek, and calls for the use of other sources, notably Metis oral tradition. Spry leads by example in quoting Louis Goulet's account in which she finds evidence of the unity and friendship throughout the colony of Riviere-Rouge (1985, p. 95-97). Moreover, Spry argues that all French Metis did not follow Riel. In fact, a commercial and agricultural elite of French and Catholic Metis had no grievance about the succession of Rupert's Land from the HBC to the Dominion of Canada (Ens, 1996, p. 129-130). Spry considers social class to be a more suitable explanation for Riviere-Rouge diversity in that it created an economically-based spatial segregation. Gerhard Ens (1988; 1996) and Nicole St-Onge (1985b) give further weight to Spry's argument when discussing variations in Metis responses to legal and economic pressures in the Riviere-Rouge area in the 1870s. St-Onge goes further in diminishing the importance of ethnolinguistic and religious cleavages. For her, [English and French-speaking Metis] seemed to have had very flexible alliances when it came to their denominational affiliation. Several Catholic mixed-bloods were burried in St. Peters Anglican Cemetery. Most of the Scottish Metis could name at least one French Metis ancestor (who would presumably have been Catholic) who had joined the Scottish Protestant mixed-bloods community by way of marriage. This would indicate that the two groups [...] were not as distinct as historians have suggested (1985a, p. 3-4).  Quoted by Peterson and Brown (1985, p. 6).  92  The two interpretations may both be reflected in the geography of the fur trade. The HBC and the NWC had developed distinct strategies and gave birth to distinct mixed-blood populations organised along both ethno-linguistic and social class lines. Initially, the companies acquired their employees from two different cultural basins. While the HBC recruited most of its labourers and clerks in Orkney, the Montreal-based company, the NWC, filled its canoes with Canadiens from the St.Lawrence Valley. This is the source of a fundamental ethno-linguistic divide. There are, however, other differences. Each company had distinctive views of its emerging mixedblood population, views that were greatly influenced by the social position of the father within the company. Many of the HBC mixed-bloods, especially men, would be sent to the British Isles, to Montreal and later to the Riviere-Rouge settlement, to receive a "proper education" — to get "civilised" —, and be taught literacy and a sense of agriculture (Brown, 1980; Coutts, 1988, p. 70). Because of their Euro-Canadian education, these mixed-bloods were more suited to occupy higher position within the HBC, such as clerks, than most of their counterparts within the NWC. The latter, like their fathers before them, would be often confined to the lower ranks of the working force, as labourers, tripmen, and canoemen. These trends were confirmed in 1821 when, with the merger of the two companies and the restructuring that followed, the majority of those laid off were of canadien origin (Friesen, 1987, p. 96; Giraud, 1945, p. 1003).  Diversity also found its way beyond the fur trade companies. Les Hommes libres — as engages (Canadiens, Iroquois or Metis) were generally called when they had completed their contract with the fur trade companies and stayed in le Nord-Ouest to trap and trade for their own benefit — were involved in a variety of economic activities ranging from those that were closely related to the fur trade economy to others independent of it. A l l of these activities had their own spatial logic. Some of the Gens libres who had kept a link with the fur trade were reintegrated part time within the HBC labour system as hunters, boat operators, tripmen, etc. (Ens, 1996, p. 29 & 43; Friesen, 1987, p. 92). Another group was composed of independent buffalo hunters who would produce the pemmican, essential food for HBC employees working west of Riviere-Rouge. A last group was involved in freighting and traded goods, pemmican, and other by-products of the buffalo to the HBC or to Indians. Indeed, some Metis engaged in all of these activities.  Freighting also operated outside the fur trade's tentacles. It was at the source of an emerging Metis commercial elite that "reigned" over the prairie to the point of challenging the  93  HBC's trade monopoly in Rupert's land (Ens, 1988 & 1996; St-Onge, 1985b, p. 155). The episode of the Guillaume Sayer's trial, in 1859, is the best-known example of it. Finally, 24  agriculture in Riviere-Rouge was an alternative way for the Metis to sustain themselves. For the most part, agriculture was a subsistence activity on a few acres of land where vegetables, crops and hay — to feed a few cattle, horses and sheep — were grown while Metis were away on the buffalo hunts (Dawson, 1859, p. 24). For the francophone areas, on the east side of the Red River and along the Assiniboine, "the Metis farming element was made of hunters who, between 1850 and 1870, because of thinning bison herds and the emergence of a market for agricultural goods, were slowly turning away from the chase to take up full-time farming" (St-Onge, 1985b, p. 156).  Although it is understood that such a diversity of origin made complexity the likely characteristic of the Metis vecu, I have identified three spatial categories, the limits of which are, of course, fluid and evolving: 1) experience of the settlements; 2) experience of the "prairie"; and 3) experience of the "margins" and of the "outside world." Unlike the fur trade geography, which divided territory by water basins (Friesen, 1987, p. 6), the categories I propose are based less on physical attributes than on the Metis perception of space.  Experience of the settled life: Riviere-Rouge and other Metis settlements.  In the course of the 19 century, the Metis created many settlements across the th  Northwest. They were generally located either in the parkland, near the fur trade posts and the woodland area, or near the numerous wooded hills scattered over the grassland. 25  26  Riviere-  Rouge was the dominant Metis settlement.  References to the material importance of Riviere-Rouge largely refer to the agricultural activity of the settlement. Matheson mentions the harvest in the fall, for example (Morin, February 25, 2003b). Goulet offer other examples: 1) that la riviere aux Marais (west of today's  The Metis Guillaume Sayer was convicted of illegal trade by the council of Assiniboia, the principal authority in Riviere-Rouge. Waiting outside for the decision was a party of armed Metis. Aware of the gathering and fearing the crowd's reaction to the judgment, the Council decided to free Sayer. The Metis interpreted it as if Sayer was not guilty and some voices raised from the crowd and claimed: "Le commerce est libre! Vive la liberte!" (Howard, 1952, p. 59). From that date, the HBC no longer tried to inforce its commercial supremacy over the Metis. Lac Sainte-Anne, Lac La Biche, Ile-a-la-Crosse, Saint-Albert, Petite-Ville, Batoche, etc. Z4  5  2 6  The montagnes de bois, de Cypres, de I'Orignal, etc.  94  Dufrost) was where the Metis used to "faire du foin" (Charette, 1976, p. 24); and 2), how the tourtes (type of pigeon), known as pests among farmers, make the settlement a rather bad land for agriculture {idem, p. 51). In addition to these comments about agriculture, there are other indications of the material reality of Riviere-Rouge; one example is the durability of the Metis log houses — the Charette house built at the entrance of the village of Saint-Norbert in 1800 was still standing when Goulet told his story (idem, p. 59). Similar comments about Metis material life in settlements had been made by Norbert Welsh who, along with Father Hugonard, was the first to farm in Fort Qu'appelle in 1878 (Weekes, 1994, p. 119). In 1884, after a few years trading, he was back to Prairie-Ronde  —  about 60 miles south from Batoche — where he started ranching. Describing his operations, he quickly sketches Metis land-tenure: "It was now the beginning of September [1884]. I had brought my plow and harness, so I broke three acres of land that could be used for potato and vegetable garden in the spring. The plowing marked my claim. Nobody else would take it. That was the law of the country" (idem, p. 144, my emphasis).  Metis community life is also a recurrent element among Metis oral accounts. Goulet gives an interesting portrait of Metis community life at Riviere-Rouge. He describes the moral importance of Father Ritchot within the community (idem, p. 90-93). He also mentions the 27  changes brought by the newcomers from Ontario in the 1850s by recalling his father who deplored the loss of "[l']esprit d'union et de camaraderie qui avait toujours existe chez les gens d'origines raciales et de religions differentes" (idem, p. 77). Similarly, each occasion Peter  Erasmus had to visit his relatives in Riviere-Rouge was an opportunity to meet with his family's neighbour and friend of Scottish descent, Murdoch Spence, and learn the latest news about the community. Spence would tell Erasmus that the quietude of the old days of the community were compromised by the coming of a young French Metis "agitator" named Louis Riel (Erasmus, 1999, p. 140, 191).  In fact, churches greatly regulated Metis life in settlements (Ens, 1996, p. 7); all central activities of community life — marriages, baptisms, funerals, and masses — were officiated by a priest. The churches also created particular patterns in the colony, dividing it into parishes with their specific religious adherence. Given the demographic dominance of French-speaking Metis, the Roman Catholic Church (RCC) was the most important religious institution. It was established in 1818 by Father Jean-Norbert Provencher in response to Lord Selkirk's and the Metis' expressed demands.  95  Settlements were also expressions of Metis symbolism as illustrated by naming. It is striking how many activities or elements of Metis daily life do refer to Riviere-Rouge. Why would the Metis call so basic and crucial a means of transportation "la charrette de la RiviereRouge" (Charette, 1976, p. 99; Vermette, 1910) if the settlement had meant little for them? The 28  same question could apply to "la jigue de la Riviere-Rouge,"  which is mentioned in W.E.  Ingersoll's narrative, The Man Who Danced Himself to Death (Morin, February 25, 2003a).  Naming is about making sense of things. The words used in the process cannot be insignificant.  Experience of the "virgin prairie, " le pays du buffalo.  It would be difficultto discuss Metis territoriality without considering the prairie.  [...] ces annees de mon enfance et de mon adolescence ont ete si belles! je n'hesitepas a dire qu'elles ont ete les plus enivrantes de toute notre histoire a nous Metis, avec Vaccent, Metifs. Nous avions la prairie vierge ou il y avait encore assez de buffalos pour nous suffire, et les Indiens pacifies n 'etaient plus la pour la disputer. Nous avions avec nous tous les anciens qui avaient vecu le temps de la prairie et de ses guerres (Charette, 1976, p. 60).  This quote pictures the prairie as Metis, and relegates both Indians and whites to the background. Indian competition for this land and the fur trade posts established in the parkland and here and there in the grassland have little significance in Goulet's perspective.  Metis experience of the prairie encompassed a vast region comprising both grassland and parkland. If many consider the parkland an ecological transition zone, with a distinct landscape where aspen is dominant and, consequently, not part of the prairie (Friesen, 1987, p. 3), the Metis narratives do not often distinguish the two zones. The prairie is wherever the buffalo are. If the 29  plains bison, "the Prince of the Prairies" to use Horward's term (1952, p. 292), generally ranged within the grassland, the Metis chased them up to Edmonton, in the parkland (Charette, 1976, p.  It is worth noting that Metis narratives often refered to the "charrette" without mentionning its place of origin. In addition to this is the fact that the "grassland zone" was not exclusively made of low-prairie vegetation (grasses, wild flowers, mosses, forbs...), but also contained some wooded areas. The rivers and hills were, as in the parkland, zones of flora transition dominated by poplars, willows or even oaks. These plant species were essential sources offirewood — lack of which would had entailed the use of les bouses de vache ("bison pies") —, and building products for Metis wintering camps and carts (Charette, p. 59, 98-99).  96  78). Moreover, other species of bison, although economically less significant for the Metis, such as the wood bison, occupied both parkland and woodland zones (idem, p. 36).  The prairie is often identified with the buffalo hunt. The buffalo was a structuring element of Metis geography.  It was a major factor in the ongoing acquisition and development  of Metis geographical knowledge — based as it was on a variety of information regarding resources such as fire wood, game, fishing areas, camp sites or flora (idem, p. 23). Louis Goulet's detailed description of the different species of bison the Metis killed for food and trade is a good illustration of such knowledge (idem, p. 36-37).  Although seasonal, the buffalo hunt occupied the Metis for much of the year. There were no fewer than three buffalo seasons: "We hunted about three times a years in the summer, fall and winter, and each man in the party would get from ten to fifteen buffalo" (Vermette, 1910). Giving even more details, Goulet explains that  Nous avions coutume de partir de bon printemps pour la prairie, des que Vherbe etait assez longue pour etre broutee, pour la pincer, comme on disait. Nous revenions vers le mois de juillet. Nous restions a la maison pendant une, deux ou trois semaines pour repartir et ne revenir cette fois que tard a I'automne, quand nous ne passions pas I'hiver en hivernement, sous la tente, dans une loge ou dans une maison d'occasion construite sur la plaine. Nous allions ordinairement a la montagne de Bois; quand le buffalo recula du cote des environs de la montagne Cypres, nous I'y suivimes. Enfin, plus tard, losque le buffalo se refugia dans les terrains difficiles d'acces du Montana, du Wyoming, du Nebraska et de Colorado, cefut le long du Missouri que nous allions a la rencontre des troupeaux qui restaient encore  (Charette, 1976, p. 32). Considering that certain Metis passed little time in Riviere-Rouge it is no surprise that Norbert Welsh delivered his own baby on the plains in the mid-1860s while trading at Prairie-Ronde (Weekes, 1994, p. 35). In fact, many Metis were born, married and died on the prairie, outside official settlements. One can find no fewer than 13 marriages, 47 births and 3 deaths recorded as "on plains" in the 1885 scrip records in the National Archives of Canada?^ Indeed, this number is relatively marginal compared to the mentions of official settlements such as Riviere-Rouge,  The hunt itself had probably reached its peak in the 1840s; never would a Metis brigade be as big in the following years. The rapid depletion of the herds from the late 1860s to the 1880s was likely the major reason for that. However, the importance of the St. Paul buffalo-robe trade, in terms of both the number of carts involved and the value, had grown during that period (Ens, 1996, p. 80). North-West Territories Metis scrip applications (scrip made in 1885 by Metis living in the Northwest Territories previous to July 15, 1870), R190-44-1 (old # RG15-D-II-8-b),findingaid 15-20. 31  97  Saint-Albert (Alberta), Portage-la-Prairie, Fort-a-la-Corne, Prince-Albert, etc? However, it 1  would appear that a considerable number of marriages made in the first place "a lafagon du pays" (Spry, 1985, p. 103) were afterward recorded in settlements where a church and a priest were to be found (Ministry of Education, Government of Saskatchewan). It was the same with baptisms {idem). There were probably many more marriages and births "on plains" than the evidence shows.  Buffalo hunting expeditions also entailed a specific spatial organisation of the camps. Although the Metis would often sleep out with only a blanket (Charette, 1976, p. 177; Vermette, 1910; Weekes, 1994), settling a camp for the night was anything but simple:  Certains chefs de caravane avaient I'habitude de former le rond a chaque arret afin d'habituer les gens a cette manoeuvre et d'en acquerir la rapidite a force d'exercice. On appelait former le rond, placer les charrettes parallelement a cote I'une de Vautre, roue a roue, puis lever les timons en I'air defacon a asseoir la charrette sur safongure d'arriere et presenter ainsi une cloture circulaire. [...] Grace a ce systeme le camp pouvait se mettre immediatement en etat de siege et preparer sa defense (Charette, 1976, p. 40).  This circular distribution of the carts would create a defensive position and help the Metis secure their family and goods from potential Indian attacks, another important element of the human geography within which the Metis lived.  The winter camps were more permanent and were generally situated near a river and 33  source of wood, exposing again the importance of the vecu in the building of Metis geographical knowledge:  La grande majorite des maisons d'hivernement etaient construites de Hard, qui etait Vessence la plus commune desforets du haut Missouri. Cetait de beaucoup I'arbre le plus abondant et le plus facile a travailler, mais une fois qu'il avait ete equarri, puis mis a secher a I 'ombre il valait le chene comme duree. Alors, je ne serais pas etonne qu 'il y eut encore de ces maisons que j'avais vu construire. Surtoutparmi celles qui etaient d'epinette rouge ou de cypres (idem, p. 59).  Among the 1360 affidivits that contains these files there are 130 marriages and 245 births in Riviere-Rouge area. According to Gerhard Ens, these winter settlements would be used only for a few consecutive years (1996, p. 79). Many of these winter camps also became permanent settlements in the 1860s-1870s. (idem, p. 97; Stanley, 1961, p. 179). 33  98  Another central element of Metis material reality was la charrette de la Riviere-Rouge and the trail network that had been developed from it. The cart was indispensable to buffalo expeditions:  34  Les chasseurs emmenaient avec eux leurs femmes et leurs enfants, sur des charrettes recouvertes de baches de peau ou de toile [...]. On appelait ces charrettes bachees des carrachetehounes. Cetait un spectacle grandiose et unique a la fois que de voir defiler des centaines de charrettes trainees par des bceufs, chargees de grappes humaines cheminant sur deux, trois files paralleles et plus, vers les troupeaux de buffalos (Charette, 1976, p. 35-36). The cart was also essential to the trade (Sealey & Lussier, 1975, p. 22). Both Matheson and Goulet picture the importance and extent of the network of Metis cart trails across the whole prairie.  As yet, the railway had not reached the Settlement, so every year in the late Fall after the harvest was in, it was necessary to make the long trip to St.Paul, Minnesota, the nearest point to which freight was delivered, to claim merchandise ordered from England months before" (Morin, February 25, 2003b). Vers le mois de septembre 1880, je suis retourne a la montagne de Bois avec Chrysostome Poitras, de Saint-Vital. Iln'y avaitpas encore le chemin defer, alors nous avons pris la route de Fort Ellice et de Qu'appelle. Je suis reste a la montagne de Bois jusqu'en juillet 1881, alors que j'ai conclu un contrat pour alter chercher dufret a 100 milles, le long de la riviere Missouri, toujours pour les memes traiteurs americains (Charette, 1976, p. 113).  The numerous place names that accompany the last description, and all of Goulet's comments regarding carts and caravans, suggest the complexity and density of the Metis network of cart trails.  Metis knowledge and mastery of the prairie was expressed in other ways; for example, in facing prairie fires. Grappling with a prairie in flames, in an environment of dry and highly inflammable vegetation, is dangerous, and success or failure -— survival or death — largely depend on collective experience and knowledge. Goulet gives us an impression of the resources needed to come to terms with prairie fire. He describes his own experience with a fire that was progressing into a muskeg (near la riviere Castor) surrounding a Metis buffalo expedition on its way to le Bassin de la Judee on the Missouri River:  There were 1,210 carts in the 1840 expedition (Stanley, 1961, p. 13).  99 A I'occasion de cefeu, lafumee durait depuis quatre ou cinq jours. Nous n'avions pas une idee juste de Vendroit ou il etait, vu la difficulte d'aller le localiser a cause de Vatmosphere enfumee. Nous savionsparl'odeur qu'il exhalait qu'il brulait dans la tourbe, et comme ces endroits n'etaient pas nombreux, nous avions une idee approximative de sa distance. [...] Lefeu qui longeait le muskeg constituait un tres grand danger pour la caravane pour plus d'une raison. La traversee du muskeg n'etait pas chose facile. II y avait au moins trois jours de marche en largeur et encore plus en longueur. Le contoumer imposait un detour que nous n 'avions plus le temps de faire de sorte que, faute de choix, nous avons decide de passer a trovers. [...] La troisieme journee nous vimes venir les decouvreurs a bride abattue, sans detour ni circuit. En les voyant venir ainsi nous avons compris que le feu s'etait engage dans le muskeg. Le conseil s' a cela. II avait pris ses precautions. II avait fait un grand cercle en brulant une longue bande circulaire de la plaine, pour proteger les voitures et pour le pacage. Le conseil avait sonde minutieusement le lit d'une chaine d'etangs, quiformait lefond d'un immense repli allant communiquer avec la riviere La Vieille, distante de plusieurs jours de marche a Vest d'oii nous etions. Enftn par des exercices de manoeuvre, il avait assure le bon fonctionnement des voitures toutes attelees vers les etangs qui avaient de Veau, afin de les abriter contre lesflammes (Charette, 1976, p. 79-82).  Goulet's description exhibits at least three aspects of Metis geographical knowledge. First, his description of the group's calm and control reveals the Metis experience with and knowledge about threatening environmental situations. Without seeing the fire, the hunting party was able to assess its extent, direction, and speed. Second, Metis TEK allowed them to react in a way to protect themselves from the flames and certain death. The Metis knew what to do in these circumstances and how to use the environment to maximise their safety. Third, the Metis also had a general knowledge of the region and of its physical particularities, which were of great help in determining their survival strategy. Without this intimate knowledge, Metis chances of handling such critical situations would have been substantially reduced.  The organised Metis response to prairie fires is, in fact, the tip of an iceberg. Below it lies a much broader collective organisation. If the Riviere-Rouge was the core of the institutional structure in the Northwest for both Euro-Canadians and Metis, the prairie was not a paradise for anarchists. It was the site of what was known as le conseil des chasseurs (Riel, 1889, p. 81).  Much discussed in the literature, the organisation of the buffalo hunt had provided the Metis with a sense of common action and the choice of their leaders (Giraud, 1945, p. 805-807; Howard, 1952, p. 302-303; Stanley, 1961, p. 180-181; Tremaudan, 1935, p. 47).  The temporary and itinerant character of hunt organisation distinguishes it from the socio-political structure in the settlements. The institutions in Riviere-Rouge  — the Council of  Assiniboia, the Legislative Assembly of Manitoba, or the diverse religious orders — were permanent structures. Unlike the hunting council of the Metis, they did not operate for a specific  100  season, then to be dissolved and reconstituted as needed. This is not to say that the organisation of the hunt was simply a loose configuration. There was a hint of "permanence." The formation of the hunting expedition and its organisation were generally decided at specific meeting points such as Pembina (Howard, 1952) or near Beaver Lake, Alberta (Erasmus, 1999, p. 200). First, the Metis would choose a leader. Then, they would elect twelve councillors and appoint guides and a "town" crier. In this organisational structure, any hunter was considered a soldat, for the expedition might have to face Indian attacks (Riel, 1889, p. 81). The rules stayed the same year after year. "There was a law that you couldn't shot cows after July 15, and if a man was found guilty of this he was fined by the chief of the party. They would also fine a man if he could not skin all he had killed" (Vermette[2], 1910)]. The communal sharing of the meat and labour was another rule of buffalo hunting expeditions (Erasmus, 1999, p. 183, 229; Weekes, 1994, p. 20). What was "temporary" in this organisation was the exercise of power itself. Although this type of political organisation would become a base for organising the new permanent settlements in the Batoche area in the 1870s-1880s (Payment, 1983, p. 95; Stanley, 1961, p. 180-181) and affect Metis society in Riviere-Rouge (Ens, 1988; Osier, 1961, p. 6), its authoritative and "coercive" function disappeared once the hunt was over. "Les Metis," wrote Riel, "n'avaient presque pas de gouvernemenf  (1889, p. 81).  Metis narratives are also rich sources of memories that pinpoint the symbolic nature of the Metis experience of the prairie. They reveal an experience that is historically rooted in the prairie. Goulet's description of the prairie fauna, while romanticised, is a good illustration:  Ca! le chant d'ensemble de la gent marecageuse n'etait eclipse que par I'incomparable ronde de la danse des poules de prairie. Quel dommage que je ne puisse ecrire une musique sur ce regal sonore que la nature des vastes espaces offrait a I'ou'ie des coureurs des plaines. Pour ma part, je n'ai jamais ecoute ces chants, ces bruits, ces cris sans mefigurer les generationspassees qui s'etaient endormies a leur rythme (idem, p. 47).  The muskeg fire episode is another case that communicates the historical dimension of Metis experience of the prairie: "Ce muskeg etait recouvert d'un epais humus vieux d'au moins un siecle et qui de memoire d'homme n'etait jamais passe par lefeu" (idem, p. 81-82)  101  "Margins" and the "Outside world": le fond et la fin de l'Ouest, and le bout du monde.  M y third and last spatial category covers spaces that, though less significant, composed nevertheless part of the Metis experience of space. I identify three sources in the making of this category. The first is "virtual" spaces, which were not based on a direct Metis experience, but rather were built on an "inherited" vecu. In this matter, a large part of the Metis song repertoire speaks for itself.  Many Metis songs were drawn from the French and canadien repertoire:  Les chansons qu 'on y entendait [during les veillees metisses] etaient a peu pres toutes tirees des vieux repertoires de voyageurs canadiens: "Via I'bon vent, v'la I'joli vent", "/eve to pied jolie bergere ", "les matelots s 'en vont a leur vaisseau ", "brigadier repondit Pindore ", Trois jeunes soldats sur le Henri IV", "Du temps que j'allais voir les fdles", "Souviens-toi belle canadienne, Souviens-toi de ton ami le voyageur", "Berce par la vague plaintive, a Venise par un beau soir", "Ten souviens-tu Caroline", "J'ai de la tristesse, moi, dans ma maison", "La table est agreable", "Dans Paris y avait une brune" (Charette, 1976, p. 65).  As these titles suggest, these songs often refer to historical and geographical realities that have nothing to do with Metis experience. However, these external places and stories form a "metageography" that allows the Metis to connect themselves to the historical geography of their EuroCanadian ancestors, to France and to la vallee du Saint-Laurent. Whereas their maternal Indian 35  ancestors were rooted in the Northwest,  the existence of the voyageurs in this region is barely  older than that of the Metis. The songs remind the Metis of the existence of an old land the origins of which are distinct from their daily experience. Virtual as this geography was, the old French and canadien repertoire of songs mirror the Metis' own experience in the Northwest. They suggest that the Metis share with the voyageurs not only genealogical ties, but also a common cultural background and a way of approaching the space where they live. It is interesting to note that Goulet's mention of these songs is one of the rare occasions he outlines the links between the Metis and Canadiens. Most of these songs were  I am aware that many Native peoples where "newcomers" for they had once migrated to the Northwest to take advantage of the fur trade economy. Surely when groups moved from the woodlands to the grasslands to take up buffalo hunting there were substantial changes in their spatial behaviour and territoriality (Ray, 1998). However, this major shift had likely entailed relatively less modification in their territoriality than it did for people of European descent, for they were to perform mostly the same basic activities that they had before: hunting, fishing, and trapping.  102  communicated to Metis within the fur trade and confirm the parallel between the Canadiens and Metis appropriations of the river network that composed the basic structure of the fur trade geography as described in chapter 2. Travelling in a canoe with Metis from Saint-Albert and Lac Saint-Anne Erasmus relates that "Someone started a French boat song and the others joined in — a rollicking melody that expressed my own feeling of the joy and freedom of the prairies. I was thrilled to be part of this happy good fellowship of the crews" (1999, p. 133).  The second source comprises spaces the Metis experienced directly, but at the margins of both the prairie and the settlements. Hence their marginality within the narratives. The woodland and the space of the fur trade is the first example. These environments generate few comments in the Metis narratives under analysis here. There are, of course, a fair number of references to the C • H B C and its fur trade posts within the narratives. Goulet makes a short mention of his trade with the HBC (Charette, 1976, p. 74). So do Welsh, who traded with Fort Garry Chief Factor, John McTavish, in June, 1865 (Weekes, 1994, p. 36), and Peter Erasmus who worked in the HBC post at Lac La Biche in the 1880s and who frequently mentioned the importance of Fort Edmonton as a yearly gathering place for HBC officials in the Northwest. Erasmus also describes the Metis experience of the rivers as paddlers and York operators: "These men [from St. Albert and Lac Ste.Anne] knew the river, every rapid, bend, and channel. Had they not traversed its bank in the tedious and back-breaking job of towing cumberstone loads upstream?" (1999, p. 133). Apart from these few mentions, the prairie and the settlements remain the central thread of Metis narratives.  St. Paul, Minessota, a very important trading point for the Metis, was another marginal presence in the Metis experience of space. In spite of his numerous years a freighter, Goulet never really mentions this city as he concentrated his freighting on the British side of the border and scouted for the us Army further west along the Missouri. Both Antoine Vermette (1910) and Welsh (Weekes, 1994, p. 28) mention St. Paul. But again, these references do not weigh much in the balance of their account compared to their descriptions of the buffalo hunt and trade. Les montagnes de Roches Brillantes are the last example. Peter Erasmus is the only one to make any mention of them, as he had been ordered to guide his employer, Dr. Hector of the Palliser expedition, through Howse Pass (Alberta). Erasmus opposed Hector's plans for this expedition and refused to accompany the Doctor (1999, p. 109). In doing so, Erasmus showed himself to be uncomfortable in and ignorant of the land to be explored, and uncertain about the time and food  103  needed for the trip. The narrative reveals the Rocky Mountains as the western limit of the Metis j?  experience of space.  36  The last source of information about the Metis experience of external space relates to the gradual imposition of a colonial geography over the old Metis land. In this matter, the English translation of L'espace de Louis Goulet, the title of which is Vanishing Spaces: Memoirs of a  Prairie Metis, best renders the shifting nature of Metis territoriality. So does the title of Norbert Welsh's account, The Last Buffalo Hunter. For Welsh, "the North-West was turning into a white man's country. The buffalo were scarce, about gone you might say. The Government was shutting the Indians up on Reserves, and everything was getting rather tame" (Weekes, 1994, p. 116). The nostalgia and the sense of loss communicated by Welsh taint much of the Metis narratives. Vermette for example: "We always hunted away to the southwest, where the United States are now; but in those days there was no boundary line" (1910). The two previous excerpts about the Metis cart trails and the railroad are other good examples of the vanishing of the old Metis territoriality. They oppose a past reality, when the Metis cart trails were the only means by which the prairie could be traveled, to a current one which displays the new supremacy of the railroad (Charette, 1976, p. 113; Morin, February 25, 2003b). Goulet also understood the symbolic importance of naming in re-appropriating a land: "S'inspirant de la couleur argentee du bois qui la couvrait, les Anglais etaient en train de changer le nom de Coquille Pilee en celui  de Whitewood, nom qu'ilporte  aujourd'hui" (Charette, p.68-67). But it is perhaps Peter Erasmus  who best summarises the changes:  "Peter," the doctor [Hector] observed, "you must prepare yourself and your associates to adjust to a new order in this country. The progress of civilization renders this inevitable." "Yes," added the captain [Palliser], "your work with our expedition is but a phase of things to come. All the great territory now sparsely populated by a few wandering tribes will someday be the home of thousands of prosperous people engaged in agricultural pursuits, stock raising, and other industries that always follow the settlement of vacant lands." (1999, P- 72).  It is a matter of fact that the Metis had crossed the Rockies' "barrier" to settle areas of British Columbia. But this was mainly done by Metis involved within the fur trade economy, which is rarely discussed in Metis narratives.  104  S h a p i n g Metis experience of space.  Stories and songs generate specific spatial bahaviours and can be interpreted as normative guides. They present information selectively and function as discourses about territory.  Memory shapes spatial  behavior.  Memory is a central component of stories and songs. The personal account of a Metis elder like Louis Goulet is primarily based on his memories. Like narratives, memory evolves. Memory is selective; it is made of "essential rememberings" and "forgettings." The Quebec historian, Jocelyn Letourneau, says of the latter that they should not be seen negatively as "refoulements," but rather as "les aboutissements d'un deuil" (2000, p. 37); memory is also, if not primarily, about making sense of the present. In fact, memory thickens the experience of space by giving meaning to places, justifying collective and individual appropriation of space, and enhancing the sense of belonging to these places.  In the story of her uncle John, Matheson gives a good example of the link between memory and individual behaviour: "The first night on the trail, it was already dusk when they made camp not far from the big swamp that was the landmark for the first day's trek" (Morin, February 25, 2003b). Matheson's uncle and his friends did not stop there just because it looked like a suitable place to camp, but also because they knew this was a place where the Metis had stopped for years. This landmark, like any other, is a specific point in space that orients collective memories and marks the site where a specific behaviour is performed. Although life in the prairie has been associated with "freedom," the Metis did not behave randomly on the prairie; they largely followed the prescriptions of elders, bearers of memory and of Metis geographical knowledge. Finding bison, berries or firewood are examples of activities facilitated by Metis landmarks and geographical knowledge.  The collective memory influences more than material existence. It is closely tied to the symbolic dimension of Metis territoriality. In his description of a buffalo expedition, Louis Goulet explains a segment of the trip in these terms:  105 Les vieux tenaient a passer par [la Coquille Pilee], parce qu'il y a soixante ou soixante-quinze ans c'etait un lieupopulaire d'hivernement. Une annee, un groupe de cent a cent cinquante families metisses de la Riviere-Rouge s 'y etaient installees pour hiverner. A cote, se trouvait un gros camp d'Indiens cris qui fut attaque durant I'hiver par une forte epidemie de grosse picote. Des chiens transporterent des germes de la terrible maladie dans le camp d'hivernement metis, ce qui le decima totalement dans I 'espace de quelques jours. Pas un seul Metis ne s 'en rechappa. II n 'y resta meme personne pour donner la sepulture aux victimes, qui devinrent la pature des loups pour le reste de I'hiver, et des corneilles auprintemps (Charette, 1976, p. 78-79).  No material reason led the Metis party to this place — no bison were to be found. The recollection of those who had died was the rationale for such a detour. Decades later, when an old man, Goulet would use both stories — about the epidemic and about the detour to Coquille Pilee — to give his own narrative some historical relevence and legitimacy.  The question of memory brings me back to my previous comment regarding the "silence" about the Metis involvement in the geography of the fur trade. It is as if this geography had been erased from the collective memory or considered irrelevant in defining ways of being Metis. This is at odds with the extensive academic literature about the fur trade economy and the emergence of the proto-Metis groups. In fact, if scholars study processes of ethnogenesis to better understand the Metis sense of identity, it is not clear from the narratives on which my analysis relies that 19 century Metis had similar concerns about their early identity formation. These th  37  sources may explain the relative absence of this specific Metis experience of space. Except for the few years Peter Erasmus spent as a HBC employee, none of the Metis who recollected and shared their memories was directly involved in the fur trade. On the other hand, it is worth mentioning that Peter Erasmus's father had also worked for the HBC. Would not Erasmus's memory have been affected by stories told by his father? What about the stories of the "most hazardous experiences of the north country" (1999, p. 63) that Erasmus was told by a Metis of the Lac Sainte-Anne settlement, father of his fiancee Florence, who worked most of his life as a tripman in the northern districts? Eramus mentioned the fact, but seems to have forgotten the stories. But certainly Erasmus was aware of their existence.  One reason for such forgetting may be that it helped forge a common sense of what it was to be Metis. What best describes Metis ethnogenesis is diversity. Presumably, such diversity  Actually, ethnogenesis studies do not rely on Metis oral tradition, but rather are documented from HBC or colonial archives. A reason for this may be the fact that there are not many accounts of the Northwest from Metis observers (Spry, 1999, p. XIV).  106  made it difficult for the Metis to define their common history, territory and language and to bound the limits of Metis national identity (Kimlicka, 1995). However, one may find better explanation in the context within which these narratives were perfomed. As I said above, these individual narratives were recorded in the early 20 century. In those days when the socio-spatial th  marginalisation of the Metis was the norm, one may consider that Metis elders were more interested in giving a distinct and positive portait of the Metis as "les gens libres de la prairie," than in focusing on their more subordinate role within the fur trade. Erasmus's antipathy to the HBC may be seen as an indicator of this: "I hated the servile attitude of most of their servants, and the almost autocratic power which some of the officials used to assert their authority" (1999, p. 86). Similarly, it could be argued that the old days of the Metis on the plains offered a "romantic" and "exotic" view more attractive to non-Metis readers. In other words, Erasmus might have omitted the Metis fur trade experience just because he was never asked to elaborate on it during the interview.  Land and rhythm: identity poetics.  Contrary to Marius Barbeau's observation about les chansons canadiennes, many Metis songs do not originate in the repertoire of medieval French troubadours,  but in Metis  experience. Pierre Falcon, Cuthbert Grant's brother in law and known as "le barde de la RiviereRouge," composed many spontaneous songs depicting highlights and daily events of Metis life, which the ethnomusicologist Annette Chretien (1996) names "story songs." These songs do not only record historical events, but they also serve to pass on "knowledge of local people and events, Metis history and social life" (idem, p. 166).  In the evening of June 19, 1816, Pierre Falcon wrote La Chanson de la Grenouillere, a song describing an event of the same day that would be known as the "Battle of Seven Oaks" — the other name for the "Frog Plain" in Riviere-Rouge. During this battle, a party of Metis led by Cuthbert Grant of the NWC encountered the HBC's newly appointed Governor Semple and a few of his men. When the two parties were within shooting distance, an accidental shot was fired by  There are well known exceptions to that pointed out by Barbeau himself such as Guerin-Lajoie's Le Canadien errant, or, more significantly for contemporary Canada, Calixa Lavallee's " 6 Canada" sung a marl usque ad mare since the late 19* century.  107 the HBC party; this shot was followed by a barrage of fire that left one Metis and many HBC servants dead, including Semple. This battle had taken place within a broader struggle for the control of the fur territories, which peaked in 1814 when the then HBC Governor, Miles McDonell, issued a proclamation forbidding the export of pemmican from the Red River area. He sought to reinforce the company's official monopoly in Rupert's Land over its competitor, the NWC. As the Montreal-based company depended on pemmican to feed its voyageurs and canoe brigades, it did not accept its rival's pronouncements. The battle of Seven Oaks rapidly became a symbol of the animosity between the Metis and the "intruders" taking over their land.  La Chanson de la Grenouillere is more than an objective description of a fight that took place in a specific site at a given time. It is a call for Metis mobilisation against the "outsider" and a claim for Metis national sovereignty over the land. 39 Chanson de la Grenouillere 1- Voulez-vous ecouter chanter Une chanson de verite? Le dix-neufde juin la bande des Bois-Briiles Sont arrives comme de braves guerriers.  6- Le Gouverneur qui etait enrage II dit a ses soldats: "Tirez!" Le premier coup, c'est VAnglais qu'a tire; L'ambassadeur a manque tuer.  2- En arrivant a la Grenouillere Nous avons pris trois prisonniers: Trois prisonniers des Arkanys Qui sont ici pour piller notre pays.  1- Le Gouverneur qui se croit empereur, II veut agir avec rigueur; Le Gouverneur qui se croit empereur A son malheur, agit trop de rigueur.  3- Etant sur le point de debarquer Deux de nos gens se sont mis a crier: Deux de nos gens se sont mis a crier: Voila VAnglais qui vient nous attaquer!  8- Ayant vu passer tous ces Bois-Brules, II a parti pour les epouvanter, Etant parti pour les epouvanter. II s'est trompe, il s'est fait tuer.  4- Tout aussitot nous avons devire, Nous avons ete les rencontrer: J'avons cerne la bande de grenadiers, [Ils] sont immobiles, ils sont demontes.  9- II s 'est trompe, il s 'est fait tuer Une quantite de ses grenadiers; J'avons tue presque tout son armee, Rien que quatre ou cinq ca I'ontpu se sauver  5- J'avons agi comme des gens d'honneur, J'avons envoye un ambassadeur: "Le Gouverneur, voulez-vous arreter Un petit moment, nous voulons vous parler?"  10- Si vous aviez vu tous ces Anglais Tous ces Bois-Brules apres, De butte en butte, les Anglais culbutaient, Les Bois-Brules lachaient des cris de joie!  39  Quoted from Margaret Complin (1939, p. 49-50).  108  Falcon's song explicitly delineates the social and geographical boundaries of Metis territoriality. First, it draws definite ethnic boundaries between the Metis and the "outsiders" — the Selkirk settlers and HBC officials. Verses no. 1, 2, 3, 5, 8, and 10 are particularly clear: the Bois-Brules are not to be confused with les Anglais or les prisonniers des Arkanys. Second, the song also outlines sharp geographical distinctions and depicts the Riviere-Rouge as the Metis "home" or "mother land." Third, the song claims exclusive Metis occupancy of the land. The fact that they were also employees of the fur trade and that the companies were also important actors in the making of Riviere-Rouge is ignored. Similarly, one learns nothing of the broader context within which this specific event was inscribed. La chanson de la Grenouillere emphasises the Metis question, while from the companies' viewpoint, this question had never been other than a secondary concern. Rather, the song first promotes Riviere-Rouge as the Metis' most important national space, where a sense of commonality and solidarityTind their greatest expression (see especially verses 2 and 3). Falcon was not the only Metis to write songs. Louis Riel also devoted some of his literary work to songs. While waiting for his execution, Riel composed La chanson de Louis Riel. Not 40  as popular as Falcon's song, nor as explicit in term of Metis appropriation of space, Riel's song is nevertheless another example of how narratives can feed common history, national identity and a sense of territory. La chanson de Louis Riel  41  1- C'est au champ de bataille, J'aifait ecrir' douleurs. On couche sur la paille Cafaitfremir les coeur.  4- Quand ell' r'cevra cette lettre Tout c't' ecritur' en sang, Ses yeux baign 'ront de larmes, Son coeur sera mourant.  2- Or, je r'gois t'une lettre De ma chere maman. J'avals ni plum' ni encre Pour pouvoir leur z 'ecrire  5- S'y jett' a g 'nouxpar terre Appelant ses enfants: Priez pour votrefrere Qu 'il est au regiment.  3- Or, je pris mon canif, Je le trempis dans mon sang, Pour ecrir' t'un' vieu' lettre A ma chere maman.  6- Mourir, s 'ilfaut mourir, Chacun meurt a son tour; J'aim' mieux mourir en brave Faut tou(s) mourir un jour  Barbara Cass-Beggs is not categorical in attributing the authorship of this song to Louis Riel. She argues, however, that the song is likely to come from Riel for he wrote many verses and poems, especially in prison (1967 p. 27). The song lyrics are quoted from Barbara Cass-Begg (1967, p. 11). 1  109 Reference to the land and to its protection from outsiders' encroachments is metaphorically expressed in this song through the blood used by Riel to write to his mother. Riel is writing, the metaphor reads, with the same Metis blood that had been spilt to defend the Nord-Ouest from intruders, the same vital fluid that roots him and the Metis to their "Mother-land." The loss of 42  blood and the tears that choke his mother's heart also represent the end of the Nord-Ouest as the Metis had known it. But at the same time, this blood makes history — as it flows both on paper and on dirt — and provides the foundations for a persistent Metis nationalism.  Metis national pride is also revealed by their collective battle for the land and for control over the buffalo, also a necessary resource for many Indian tribes. Hence the need for Metis to protect themselves from potential Indian attacks, particularly from the Sioux. This is depicted in the Metis story of the Sixty Seven Bois-Brules:  43  1- Of a most incredible tale, let me say That it remains a legend to this very day; Of a buffalo hunt on the open plain A small Metis band who won great fame.  7- The southerly band was very small indeed Three score and seven of the bravest Metis. As they reached grassier plains their hearts did smile And again hope grew with each passing mile.  2- It happened so very very long ago That only the prairie wind is left to know, Of the time sixty seven Bois-Brules Were challenged by two thousand Sioux one day.  8- I believe no more than three days passed When all too suddenly and quite fast, The M6tis scouts while ahead to scan Spotted 2000 Sioux charging overland.  3- One hot summer in eighteen forty nine Nine hundred carts left Pembina on time, They made their way towards the west For the mighty buffalo as their quest. 4- Their home they left with happy hearts The women and children seated on carts The men recalled with anxious talk The number of buffalo they each had stalked 5- Across the dusty sun-baked land The ox-cart trail was like desert sand. Two long weeks had slowly passed And still no buffalo on the prairie vast.  9- Excitementfirstin their hearts took hold Then a terrible fear made their hearts turn cold, Quickly the men shouted out their demands And rifles made ready tofireat command. 10- With careful aim the Metis shot To protect themselves and the lot, The battle lasted all through the day And by night they had turned the Sioux away. 11- Just sixty seven alone that day Became legendary Bois-Brul6s.  6- So one night with unhappy hearts A decision was made that they must part. Some would go north, while others west And southerly bound went the rest.  4 2  "Le Nord-Ouest est ma mere,  Riel would repeat during his trial, le Nord-Ouest  est ma mere patrie, je suis sur que  ma mere ne me tuera pas" (quoted from Morisset, 1997, p. 94).  This story is quoted belowfromBruce Sealey's Stories of the Metis (1975b, p. 90-93). The author associates this story to "the era of Louis Riel." I am not aware of any French version. Nor am I sure that this story has not been written recently; the words might be not authentic. However, the spirit they convey seems undoubtedly genuine. Rememberance of this event had been important for the Metis. Norbert Welsh refers to those years when the Sioux used to attack the Metis (Weekes, 1994, p. 161).  43  no  According to this story, the Bois-Brules are these courageous ancestors who once faced 2,000 Sioux warriors and defended themselves and their right to use the land. The story also emphasises the age of this event, to the point "that only the prairie wind is left to know"; a longtime ago then, when the Bois-Briiles  where already a distinct people, they left "home" with "nine  hundred carts" travelling "the ox-cart trail." The spirit of this story is that the Metis should be proud of their glorious past, of what they are, and of the land they are from. There could hardly be a better way to feed national feelings. From the Sioux perpective, of course, the Metis were appropriators — colonisers — of their land.  Place  hierarchy.  So far, I have presented the Metis experience of space with little regard for the relative importance of each spatial category. There was no doubt, in colonial minds, that settlements were the core of Euro-Canadian colonisation in the Northwest. This was particularly true in the case of the most important settlement, the Riviere-Rouge. Colonial maps perfectly picture a Western "core/periphery" cleavage (Rokkan & Urwin, 1983). Beginning with Riviere-Rouge, settlements are the centres of attention on colonial maps. Their relative importance is suggested by the higher density of information related to them. In consequence, the prairie appears subordinated to settlements and is depicted as their potential hinterland, a fertile land to be exploited; the prairie is, like any peripheral region, "dependant upon one or more centres in at least one of the three domains of behaviour [...]: in political decision-making, in cultural standardization, and in economic life" (Rokkan & Urwin, 1983, p. 3-4). In some ways, this is true for the Metis as well. As agriculture in settlements was susceptible to crop failures and diseases, Metis supplemented their economy with prairie provisions (Ray, 1984, p. 269-270).  However, a Metis hierarchy of places does not emerge clearly from their oral tradition. Although some parts of these narratives suggest the centrality of settlements, other parts picture the lure of the "margins."  One may find many excerpts that set settlements at centre stage. The Riviere-Rouge is described as "la maison" and "le pays natal" by Louis Goulet, when he refers to Saint-Norbert (Charette, 1976, p. 32, 45, 95 & 113) or when he discusses how, in 1870, Louis Riel and the  Ill  Metis prevented the Dominion from taking it over (idem, p. 89). Peter Erasmus, also a native of Riviere-Rouge, talks of the settlement as "home" (1999, p. 14).  The centrality of Riviere-Rouge is also comparative. The settlement is often constructed in parallel with the prairie, and in many cases, the latter appears subordinate to Riviere-Rouge — the vast hinterland that furnishes provision: "[...] nous allongions la distance qui nous separait du pays natal de la Riviere-Rouge,  en nous rapprochant sensiblement de la region de chasse aux  buffalos: la montagne de Bois ou la montagne de Cypres" (Charette, 1976, p. 44-45). Norah  Matheson depicts the Riviere-Rouge/prairie  binary in the same way:  In the year 1858, my uncle John was sixteen years old, and ready for adventure. Life in the Red River Settlement [...] was pleasant, with plenty of work, plenty of friends, and plenty of "home-made" amusement according to the seasons, but John was eager to see the world beyond the confines of the settlement (Morin, February 25, 2003b). In Matheson's mind, the prairie presents a life of adventure in contrast to the comfort of daily life at Riviere-Rouge;  the prairie is a terra incognita to be explored.  But, paradoxically, many other parts of Metis narratives present the prairie as the focal point of Metis reality. Louis Goulet had spent little time in his native Riviere-Rouge and most of his narrative is about his experience of the "bare" prairie: in his childhood during the buffalo hunts; in his adult life as a freighter or as a scout for the US Army. Moreover, his descriptions 44  are fed by a deep and positive appreciation of life on the prairie, and reflect a specific attachment to that space. Peter Erasmus's attraction for the prairie life compromised his chance to marry Florence, a Metis woman from Lac Sainte-Anne: "Suddenly I realized that the real reason for doubt or hesitation in declaring my intentions was not actually the difference in religious adherence but in my own dislike for a settled existence and my love for travel" (1999, p. 61). Another native of Riviere-Rouge, Norbert Welsh, who traded all over the plains for most of his life, mentions that "in all I must have had about twenty wintering houses on the Saskatchewan plains" (Weekes, 1994, p. 96). Metis settlements in the Northwest, no longer appear quite so central to Metis life. In these narratives, the prairie is depicted as a way of life that has a great deal to do with Metis identity and territorial behaviour.  Goulet's family had once spent over two years away from the Riviere-Rouge (Charette, 1976, p. 72).  112  Metis narratives do not reveal a fixed Metis hierarchy of places. The Metis spatial priority consistently oscillated between settlements and the prairie. For the Metis, the "core" and the "periphery" are interchangeable spatial categories. In Metis' minds, they can switch quite rapidly:  C'est au cours des ces voyages memorables [towards wintering places] que s'allumaient les amours qui aboutissaient a d'heureux manages. C'etait la belle occasion pour lesfdles et les veuves de se trouver un mari choisi sur la creme de toute la race! [...] La vie pastorale en commun au cours de la traversee des plaines depuis la Riviere-Rouge aux montagnes de Bois, de Cypres, ou aux bords escarpes du Missouri continuait meme apres que la caravane eut atteint sa destination [Riviere-Rouge], Un repos de quelques jours succedait a notre arrivee et nous en profitions pour retrouver des connaissances ou de la parente. C'etait le temps oil les liens de parente chez les Metis s'etiraient pour ainsi dire a I'infmi. II suffisait a des grands-peres d'avoir une fois echange des chiens pour que leur petits-enfants se considerent comme des parents. Ceux issus des cousins de deuxieme et troisieme degre redevenaientdes oncles et des tantes... (Charette, 1976, p. 61).  Where, according to Goulet, is the centre of Metis socialisation, the focal point of their communal life? Is it in the prairie where months of pastoral life were shared, and where, he said, many of the marriages occurred? Is it in Riviere-Rouge where the extended family assumed all its significance? Actually, it is not in one or the other. Metis life in Riviere-Rouge is as much influenced by the prairie, as Metis life in the prairie is influenced by Riviere-Rouge. This shows that the Metis, like the First Nations, were multi-place oriented. Each place took all its meaning within a network of places. Overlapping spaces of in-betweenness.  The ambivalence that characterises the Metis territorial hierarchy can also be observed in the way Metis define their otherness. Although one knows who is included within the Metis' others — both Indians and Euro-Canadians — such a category remains flexible in time and space (Ens, 2001; Ray, 1995), and suggests that the Metis are neither white, nor Indian, but partly both. Geographic location fully participates in determining the nature of Metis otherness and in explaining its fluctuating condition. Does it mean that the Metis are more likely to be "EuroCanadian" in settlements while more "Indian" in the prairie? It may well be, but this picture seems too simple, too incomplete. Ultimately, it is my argument that the Metis cultivate their Indian origins in settlements as much as they spread their Euro-Canadian nature over the prairie.  113  Their territoriality is made of overlapping spaces that give the impression that they are at "home" nowhere and everywhere at the same time.  Spatial mobility: reality of "wandering"  Spatial mobility  45  people.  is what best describes Metis identity and territorial reality in the 19  th  century Canadian West. For most Metis, life in the settlements was not continuous. Many would leave their settlement for months, if not years, as Louis Goulet and Peter Erasmus did. Settlements were both points of departure and of arrival. They fell within a broader network of places (wintering, hunting, etc.) dispersed over the prairie and connected to each other by Metis mobility. In turn, some of the latter places would gradually become permanent settlements where hunting activities were complemented by some basic agriculture (Payment, 1983, p. 95).  Spatial mobility echoes, to a great extent, the fluctuating nature of the Metis hierarchy of places. It gives birth to a persistent and paradoxical feeling. Reading L'espace de Louis Goulet and Buffalo Days and Nights, one observes both the excitement of departing for the "free life" of the plains and the heartbreak of leaving Riviere-Rouge (Charette, 1976, p. 77; Erasmus, 1999, p. 141). The two narratives also express the pleasure of coming back to Riviere-Rouge. Such a mix of feelings makes it difficult to clearly identify which place is more meaningful for the Metis, the settlement or the prairie. Metis spatial mobility seems to considerably rely on a mental or symbolic disposition for "ubiquity," as if the wish to occupy consistently two spaces at the same time was central to Metis reality. This is addressed by the old French songs. Clemens refers to this type of musical material as "songs of separation," oral expressions of the Metis "itinerant life" (1985, p. 3). The "songs of separation" make some sense of the "wandering" nature of Metis life by expressing fragility or the absence of well defined geographical anchors, and the  1 use this expression for it helps avoid the eurocentric tone that expressions such as "nomad" or "semi-nomad" bear. As the anthropologist Adam Kuper says of primitive society ideology, one can see "nomadic people" as the "distorting mirror" of Western society, as the unconscious strategy of constructing otherness (1988, p. 5). Even labels such as "semi-nomadic" and "semi-sedentary" — often tagged on the Metis —, although attempts to establish connections between these two "irreconcilable" conditions of life, are incapable of overcoming the antagonistic philosophy that ties these conditions together in a Western anthropological tradition. What anthropologists and archaeologists propose in exchange is to broaden the definition of sedentariness to include all settlement systems "in which at least part of the population remains at the same location throughout the entire year" (Rice, 1975). Mobility becomes the variable (it can be nil) and sedentariness the constant; increasing of mobility is no longer thought as a 45  114  nostalgic memory of the land and the people left behind. In 1860, the German ethnographer Johann George Kohl described in similar ways the musical performance of the Metis voyageurs of the pays d'en-haut who "regarded themselves as exiles" (quoted in Ray, 1996, p. 50). On the  other hand, these opposite states of mind (exitement-heartbreak) may simply emphasise the Metis ability to feel at home anywhere in their extended territory. Thus, Metis "ubiquity" would only be an illusion perceived by outsiders, including academics. In any case, both interpretations confirm the changeability of Metis spatial hierarchy and their multi-place tendency.  Identity mobility: to be your own other.  Mon pere qui partageait notre deception eut I'ingenieuse pensee de nous apprendre que si nous ne repartions pas, nous serions obliges d'aller a I'ecole. La question fut reglee. Nous aimions mieux repartir tout de suite que d'etre obliges d'adopter la vie des petits blancs et d'aller a I'ecole. Nous pensions a tout cela quand un bon matin defin d'ete, on alia se joindre a une caravane qui s 'ebranlait deux jours apres, pour les hivernements dans lefond de I 'Ouest (Charette, 1976, p.  77-78). Goulet exposes here a clear-cut distinction between Metis and "white." According to Armando Jannetta, a literary specialist, L'espace de Louis Goulet often constructs Metis identity as binary  and rhetorically emphasises Metis difference and otherness with regards to both Indians and les Anglais (1994, p. 62).  In Goulet's memoirs, Indians are frequently depicted as des sauvages or des mauvais  chiens (Charette, 1976, p. 146). Goulet's comment suggests that the Metis considered themselves more civilised than their Aboriginal forebears. The "primitive" Indian can never been trusted, as Goulet suggests by his comparison of Metis and Indian horses:  De notre temps, Von appelait les chevaux des Metis cayousses et ceux des Indiens, broncos. [...]La proportion des bonnes betes etait plus forte chez les cayousses que chez les broncos. La grande difference entre les deux, en dehors de leur physique etait que le cayousse ressemblait au chien:  social regression or as a lower level of cultural advancement (Crepeau, 1993; Eder, 1984 & 1986; Rafferty, 1985). With such a definition in mind, the Metis can be considered settled with a relatively high level of mobility.  115  vous saviez a peu pres toujours ce que vous aviez dans vos mains, tandis que le bronco, c'etait comme un sauvage, vous ne pouviez pas vous yfier avant de Vavoir connu (idem, p. 97; also quoted  in Jannetta, 1994, p. 63). If Peter Erasmus shows more respect for Indians and usually depicts them positively, he also labels them sometimes as "uncivilized savages" (1999, p. 127). These labels may be partly explained by the increasing conflict between the Metis and First Nations as they both competed for the same depleting resources, especially the buffalo herds (Ray, 1984, p. 270; Ray, 1998, 206, 227 & 231).  Les Anglais and les Americains compose a great part of otherness in Metis testimonies. In a land where the "newcomers" are becoming more and more influential and are progressively modifying the geography of the. whole region, they are often conceived of as "le diable [...] dans la cabane" (Charette, 1976, p. 77). As Jannetta points out, Goulet opposes the Metis and the English other in terms of the antecedence of the former (1994, p. 61-62). A l l the comments made by Goulet, Vermette and others, about the US border and the railroad and that above I have labelled "nostalgic" can, then, be seen as the symbolic marks distinguishing in space and in time the Metis from newcomers. Moreover, the Euro-Canadian is generally depicted as one who can hardly survive on the prairie, ill-adapted as he is to operating in this vast wilderness. Hiding from the sight of some Canadian troops during the 1885 uprising, Goulet says that "Si ce n'eutpas ete des blancs, ga y etait!" (Charette, 1976, p. 174), meaning that if these men had been Indians or Metis, Goulet and his friends would have been detected.  However, a second glance at these narratives indicates some permeability in the Metis "othering" process. Metis narratives are not exclusively constructed on difference, but also on mediation and dialogue. In other words, the Metis also use the others' discourses and categorisations to define themselves (Jannetta, 1994, p. 60). Metis narratives amalgamate diverse voices (idem, p. 65) and move between two cultures, in so doing making Metis identity exceedingly mobile. When the narrator distances himself from the Indian, by emphasising the latter's "savageness," he makes use of a colonial discourse, and approaches the Euro-Canadian way of thinking. When he shows sympathy for Indians or is critical of Euro-Canadian behaviours or ideologies, he moves in the opposite direction. When Goulet describes William Gladu as a  116  "metis sauvage,"™ he admits in veiled terms Metis potential for occasional "savageness" and somewhat contradicts his negative assessment of the savage state. Goulet is trapped by his own in-between condition, and his narrative offers him no escape. Rather, this narrative exposes how its author builds his own sense of identity through a looping movement alternating between divergence and convergence, Indians and Euro-Canadians.  Peter Erasmus is caught in a similarly ambivalent logic. Like Goulet, his categorisation of Indians as savages is contradicted by his critique of the "so-called civilized living" brought to the Northwest by Euro-Canadians (1999, p. 201). His narrative opposes his enjoyable life with Woodland and Prairie Indians and the settled life brought by the inevitable advance of colonisation and by the great socio-economic changes in the region. He also has to make sense of the discordance between Indian and Euro-Canadian beliefs. Irene Spry outlines the "conflict between Indian mysticism and Protestant rationalism" experienced by Erasmus (1999, p. xxviii). The antagonism between scientific and Indian medicines presents another dilemma. If, at first, Erasmus expresses his doubts with regard to the latter (1999, p. 154), later he is "convinced that [his old friend and Methodist minister, Rev. Woolsey] could have been saved his present misery if he had not been prejudiced against Indian medecines" (idem, p. 171).  But ambivalence and in-betweenness have their advantages as well. They make the Metis successful intermediaries between Indians and Euro-Canadians, a role they are often pleased to play. Peter Erasmus, Louis Goulet, and Norbert Welsh all served, at least once in their life, as interpreters (Erasmus, 1999, p. 239-264; Charette, 1976, p. 146, 150; Weekes, 1994, p. 84, 107). They relate proudly to this experience and consider themselves, and the Metis as a whole, to be indispensable on the plains: "Many of the early famous travellers would have been hopelessly lost, starved, or frozen to death without the guidance and advice of the Indians and half-breeds" (Erasmus, 1999, p. 75). In other words, the Metis did not passively experience ambivalence and in-betweenness, they were aware that their intermediacy made them a distinct people.  It is no surprise that this intermediacy became, over the years, a way for the Metis to mark politically their distinctiveness in the Northwest. This Metis political awareness is best  The term chosen for publication was "metis indien." The "metis sauvage" expression was found in the transcripts of Charette's original manuscript notes in the Provincial Archive of Manitoba (MG9A6, fo 112). If the former  117  rendered through the words of Charles Nolin. Attending the meeting that decided to send a party to Montana to retrieve the "exile" Riel, Nolin said that: "La question metisse [...] est comme une charrette. Pour la faire marcher ilfaut deux roues et, dans le moment, il nous en manque une. Si nous la voulons, il nousfaut aller la chercher dans le Montana, le long du Missouri" (Charette,  p. 137). Nolin's two wheels are made of different woods. On the one side of the cart is Gabriel Dumont, the "child of the plains," an image of the Indian way of life; on the other side is the "educated" and literate Louis Riel, symbol of the Euro-Canadian element of Metis reality. For Nolin there exists only one way for the Metis to resist the Dominion, it is to mobilise a cart that contains the two essential components of their identity.  Moving from place to place: a journey in-between identities.  Considering the interactive nature of territory and identity, I would argue that the Metis spatial mobility and identity mobility are two sides of the same coin. This is what Jannetta also suggests when he refers to "Goulet's nomadic ability to move between two cultures and to have sympathies for both..." (1994, p. 65), and argues, at the same time, that spatial mobility — which he considers a "(semi-) nomadic life-style" — is the basis of Metis identity formation and definition of the other (idem, p. 60).  But as far as the Metis are concerned, the situation of the other is fairly fluid, and this modifies the way Metis position themselves both ethnically and geographically. As I have suggested above, it is too simple to consider that the Metis get closer to the Indians when on the prairie and that they automatically adopt "civilised" manners, such as farming, when in RiviereRouge. Metis narratives are not drawn on these fractional lines. Goulet sometimes assimilates the Metis to their Indian conditions in the settlements as he does when reporting Ritchot's words: "Si on sefaufilait pour ne rentrer qu'auxpetites  hemes du matin, [le Pere Ritchot] ne disait rien,  mais a peine etions-nous endormis qu'il nous criait: 'HO! HO! HO! la, les sauvages, c'est le  temps d'aller au fain" (Charette, 1976, p. 92, my emphasis). Similarly, although Goulet admits the influence of Indian reality for the Metis way of life in the prairie — when he associates, for  efficiently shakes the distinction between Metis and Indians, the latter remains more significant for it bears the  118  instance, Louis Riel's superstitions to the sixth sense the old Metis developed among Indians (idem, p. 185) —, his narrative should be mostly interpreted as an account of the Metis bringing "civilisation" among Indian tribes. In other words, the Metis had the ability to move between "civilised" and "primitive" conditions, in both the prairie and the settlements, and in so doing, to adapt their sense of identity and territory to specific social and geographical contexts.  This last comment reminds us that the Metis are neither "whites" nor Indians, but they are rather simulteneously both. I am aware that academic discourse — including mine — does not always make this point obvious. In fact, such analysis tends to emphasise the contradictions and ambivalance that articulate Metis life and provides a bi-polar image of Metis sense of identity and territory as either Euro-Canadian or Indian. These contradictions may well be academic constructions based ona Euro-Canadian experience of identity. In spite of the complex picture of Metis identity Louis Goulet provides, he never refers to this identity in terms of ambivalance and mobility. Goulet is not ambiguous when he says he is Metis.  3.2.2  Metis toponymic nomenclature: The land that knows a name.  47  At this point, it is necessary to come back to toponymy. So far, I have considered place names as symbolic components on colonial maps and in Metis narratives, and as expressions of power in space. M y treatment may give the impression that place names are only elements of broader territorial discourses, cartographic or oral. What I am suggesting in this last section is that Metis toponymy is a territorial discourse in its own right (Nash, 1999, p. 473) — made as it is of an array of ideologies (spatial, historical), practices (naming, mobility, orality), conventions, codes and rules (linguistic ones for instance) that are generated and recognised by Metis society — that draws and conveys a specific mental and cultural representation of space (Guillorel, 1999, p. 64). The descriptive and pictorial nature of Metis toponymy, in which place names are the mental presentations  of the physical (la riviere aux Gratias) and human (lafourche  des  Gros-Ventres, the South Branch of the Saskatchewan River) geography they name, means that it provides a wealth of geographical information. Taken as a whole, place names are indispensable  ideology of primitive people, which in this case, seems stranger to Goulet.  119  to the Metis and can be considered, as Susan Fair argues for Alaska Inuit place names (1997), as a cornerstone of their TEK. Toponymy is an indicator of the Metis way of perceiving, comprehending and forging their territoriality. This is its power.  The toponymie contribution to Metis geographical knowledge is two-fold: historical and spatial. In-betweenness is, again, central to the Metis naming processes. An "oral history" of the Nord-Ouest: collective memory of place.  Place names are closely tied to Metis memory and orality. As the elderly Goulet says: "Savez-vous? Chaque fois que j'y pense et j'y pense souvent, ga me fait ennuyer. Ces noms me rappellent tant de choses, tant de souvenirs! Devant mes yeux eteints, il me semble les revoir passer comme dans les defiles de mon enfance et de ma jeunesse" (Charette, 1976, p. 60).  Toponymy takes an important place in Goulet's narrative. The richness of the Red-River-born toponymie knowledge is outlined by a map that the publisher has produced and printed on the back cover of the book in both the French and English versions. I will come back to this map in chapter 5. Here I note the considerable number of names it contains, most of which have since disappeared or been translated or corrupted. Toponymy is also central to Peter Erasmus and Norbert Welsh's accounts. There is hardly a recollection or a description that is not linked to a specific place name in all of the examples of Metis oral tradition that I have analysed in this chapter.  Metis emphasis on place names is not unique. It is a behaviour that belongs to most oral cultures. As Julie Cruikshank notes of Athapaskan toponyms, they "are more than names; they are metaphors bringing together varieties of information in one small word" (1990, p. 63). In other words, place names are mnemonic devices that follow and lead the narrator through a story (Fair, 1997, p. 473). The mnemonic role of Indigenous place names is reflected in the terse grammatical style in which these toponyms are composed. Toponyms are generally not precise geographical descriptions, but rather pictorial and summarised versions, that act as guides for  Otherwise specified, the principal sources for this analysis of Metis place names are Erasmus (1999), Charette (1976), Coues (1897), Hind (1860), Letourneau (1980), Morice (1912), Rondeau (1923), Tache (1869), and Weekes (1994).  120  remembering and communicating (Pearse, 1999, p. 159-160). Globally, place names convey individual and collective memories, stories, and the mythic baggage of a people. To communicate these names from one generation to another is to preserve much of a people's oral history.  Some names are particularly eloquent in this regard for they are the result and mirror of specific events, whether real or mythical. These names, which one could label anecdotal, are numerous in Metis toponymic nomenclature. Les Mauvais Bois (Butler, 1873. p. 23) on the Assiniboine (surrounding les Grands Rapides de I'Assiniboine,  today Brandon, Manitoba), is an  illustration of how experience of space and specific events can be spatially incribed. Henry Youle Hind refers to this place name in his narrative of the Red River expedition:  Leaving Prairie Portage on the morning of the 19 [June 1858], we took the trail leading to the Bad Woods, a name given to a wooded district about thirty miles long, by the buffalo hunters in 1852, who, in consequence of the floods of that year, could not pass to their crossing place at the Grand Rapids of the Assiniboine by the Plain or Prairie Road. There were found hundred carts in the band, and the hunters were compelled to cut a road through the forest of small aspens which forms the Bad Woods, to enable them to reach the high prairies. This labour occupied them several days, and will be long remembered in the settlements in consequence of the misery entailed by the delay on the children and women (1860, vol. 1, p. 283-284). th  There are many other examples, some of which I have mentioned before: la Machoire AO  d'Orignal (Moose Jaw, Saskatchewan),  la Butte du Cheval Caille (Pinto Butte, Saskatchewan),  la riviere du Calumet (Pipestone), lac La Vieille (Old Wives), le Coteau des Festins (North of  Saint-Francois-Xavier), or la Pointe Coupee near Saint-Eustache, Manitoba (Charette, 1976;  Coues, 1897; Letourneau, 1978; Rondeau, 1923).  The anecdotal nature of this toponym is a matter of contention. If it was once proposed that the name was derived from the story of a white man who had repaired the wheel of a charrette de la Riviere-Rouge with the jaw bone of a moose, such interpretation is now considered eurocentric. The most broadly accepted interpretation is that the name is Native and refers to the shape of the Moose Jaw River (Hamilton, 1978, p. 302). This interpretation is suspect, however, for it supposes the use of maps; it is unlikely that the river looked like a moose jaw from the ground. Although it is now known that Indians produced maps before Columbus, as noted in chapter 2, they were based on quite different cartographic conventionsfromthe European ones. Indian maps' accuracy was measured less in terms of their topographic accordance with the represented area, than in terms of their respect of the topology and hierarchy of water systems; in a colonial perpective, Native maps were at best schematic representations. From such cartographic documents, one would not likely be able to decipher a moose jaw-like design. Paradoxically, this explanation is thus eurocentric as well. To come back to the initial interpretation, I would note that it is possible that the repair of the cart wheel was made by those Metis accustomed to building and repairing their carts — which contained no metal parts — entirely of wood. 48  121  When it is noted how close place names are to geography and culture and how collective memory is inscribed in them, then it becomes clear that Metis toponymy reveals the historical land use of the Metis (Cruikshank, 1990, p. 63) and the fact that Metis history is inscribed (or "written") in the landscape (Linklater, 1994). This is much more than the inscription of a name on a map, which can be done without occupying the land; Euro-Canadians had named the land well before they occupied it. Rather, Metis naming is the result of a close time-related experience of the land. The recitation of names is then a means by which Metis can communicate such a vecu. To list toponyms and to give detailed geographical information are parts of the same process in Metis culture. The name "la riviere aux Iles-de-Bois" (Boyne River, Manitoba) not only reveal what can be found at this river, but also bears on what this river meant for Metis material life. Along this river, only a few miles west of Riviere-Rouge, was a forest of oak where the Metis used to gather most of the wood they needed to built or repair their carts. As the archaeologist Linklater (idem, p. 62) says of Nelson House Cree (northern Manitoba), the Metisnamed landscape is the "text" in which is recorded their historical experience of land. The fact that, as I have noted, numerous Metis place names originate from previous Native geographies is an indirect way of suggesting the Aboriginal and immemorial nature of Metis land occupancy and of reinforcing the durability of the Metis in space (Akin, 1999, p. 10; Guillorel, 1999, p. 65) by tracing their history back to a time when they were not yet a distinct people. An "oral geography" of the Nord-Ouest: topography of mobility.  To say that Metis place names tell about historical land use and geographical knowledge is to admit that they map an oral geography (Pearse, 1999, p. 159), they comment a territorial reality (Collignon, 1996, p. 116). The descriptive nature of these place names has a great deal to do with that for they provide characteristics of the place they name. It is a fact that dedicative toponyms (which comprise most Euro-Canadian place names) are incapable of doing as much. In involving an imported metaphysical, religious, and historical repertory, these toponyms are not site-specific and remain foreign to the places they name. Although Metis names can be mapped, they do not need such a document to express their geographical relevance. In fact, "detaching names from the context in which they are presented as though they can be objectively isolated and filed on a map gives too little sense of how they are actually used" (Cruikshank, 1990, p. 56).  122  Indeed, some Metis place names are dedicative, La traverse a Gabriel for instance, a place named after the Metis leader Gabriel Dumont. But the toponym is also descriptive in that it is a terse expression that literaly means "the place where Gabriel Dumont operates his ferry service," and so reflects both the reality to which it refers and the Metis experience of space.  Toponyms are often close to the geographical reality of the prairie. They can refer to topographical features: 1) the prairie itself, with names such as la Prairie-Ronde  (Saskatchewan),  and la Prairie de la Tete de Bceuf (Calf Mountain area, Manitoba); 2) the numerous hills scattered over the prairie and where the Metis had a number of camp sites — la Montagne de I'Orignal (Moose Mountain), la Montagne de Cypres, la Montagne de Bois, la Montagne Sale,  and la Butte du Foin de Senteur; and 3), the coulees and rivers that incise the prairie such as la Grande Coulee de la Grosse Butte which derived "its name from a large conical hill about two hundred feet high" (Hind, 1859, p 18). Some other place names describe available resources. La Talle de Hart-Rouge (Willow Bunch), is one example; the Metis used to fill their pipes with the bark of the red willow that grows in this region (Rondeau, 1923, p. 104). Finally, Metis place names also refer to the human geography, with toponyms such as Batoche, which refers to the place where Xavier Letendre dit Batoche had his ferry and store.  Place names also picture Metis experience of the northern margins (the boreal forest and the parkland), those lands often shared with the Canadiens, at the very heart of the fur trade geography: 1) in Manitoba, there were notably les eaux-qui-remuent (series of portages near the mouth of Winnipeg River), la riviere de la Tete Ouverte (Brokenhead River), la riviere Bouleau, la riviere Blanche, and the upper section of la riviere au Rat (Geographic Board of Canada, 1933, p. 30; Charette, 1976, p. 37); and 2), in Northern Saskatchewan and Alberta, with places such as Lac La Biche, lie a la Crosse, or Portage La Loche.  Central to the persistence of the geographical knowledge of oral cultures such as that of the Metis is the "practice of place." As Beatrice Collignon (1996) argues in her Les Inuit: ce qu'ils savent du territoire, traditional knowledge can only persist by the action and relation of an ethnic group to its territory. Sedentarisation of the Inuit populations, as the French geographer argues, appeared to be fatal to their mobile experience of territory and traditional knowledge.  123  In the same vein, it can be said that Metis spatial mobility was both a practice of place, and a means by which geographical knowledge can be preserved and communicated to the next generations. As the Metis toponymy was knowledge, it is obvious that it played a role — along with the reasons for travelling (Collignon, 1996) — in structuring Metis mobility. Although described as "wandering" people by many observers in the 19 century, Metis mobility was not a th  loose spatial practice but one that was socially structured (Thistle, 1997, p. 200), and based on a precise and well established network of places recognized and named by them. In return, spatial mobility was sometimes affected by the encounter of exceptional situations, such a prairie fire, or by a collective decision — the elders' will to reach la Coquille Pilee in order to recall those who once perished there. Deflected from its original course, a party of Metis could be exposed to unknown or long unvisited lands. Such encounters could generate new territorial experiences, • and sometimes feed Metis toponymie baggage with more names. The previous quoted example of les Mauvais Bois is relevant; forced by a flood to cut their way through wood, the Metis immortilised the event and the place by a proper name. Overall, a specific place acquires its full meaning as part of a network, confirming what was said above about the Metis multi-place orientation and lack of a clear spatial hierarchy. Mobility was essential to the maintenance of such a network of place and to the persistence and development of Metis geographical knowledge. An in-between geography of the Nord-Ouest: naming distinctiveness.  Metis place names are symbolic expressions of in-betweenness and identity mobility. If many place names derived from direct Metis experience, many originated from both Metis cultural backgrounds, Indian and Euro-Canadian. This variability in naming exposes the duality of Metis collective memories, stories, and mythic baggages. This explains why toponyms of Native origin, such as the riviere Queue d'Oiseau, riviere Calumet or lac la Vieille, lie alongside place names of Euro-Canadian extraction: 1) toponyms like Portage-la-Prairie, derived from the Laverendrye's exploration expeditions of the 1730s-1740s; 2) toponyms shared with the voyageurs, such as lafourche des Gros-Ventres (Tache, 1869, p. 34); and 3), hagionyms (names with religious meaning) brought in the Northwest by missionaries, such as Saint-Boniface, SaintNorbert, Saint-Anne, Saint-Vital, and Saint-Albert. As I have said above, Metis did more than translate into French (or Metchif) Native place names; they also integrated the original stories  124  (real or mythic) associated with these names. La Prairie du Cheval Blanc (today Saint-FrancoisXavier on the Assiniboine River, Manitoba) is one example. The general belief is that the name comes from the Indians and relates the story of a mysterious and unapproachable horse (Rudnye'kyj, 1970, p. 179). Although agreeing with this, Elliot Coues also specifies, in a footnote of his edition of the narrative of Alexander Henry the Younger, that the story from which the name is derived has been often associated with Metis oral tradition (1897, vol. 1, p. 288). If la Prairie du Cheval Blanc is a name that borrows its meaning from the Indians, it draws in return its form from Euro-Canadian language, which adds another dimension to the inbetweenness of Metis naming.  In addition to this general picture, there are more specific indications of the in-between reality of Metis naming. One of them is the co-existence of more than one name for the same' place. Again, la Prairie du Cheval Blanc is a good exemple, for this "Native" place name was used for a while in parallel with the hagionym Saint-Frangois-Xavier.  In this case, the former  had eventually been substituted by the latter. But in some other cases, amalgamation is the norm. This is how la Coulee des Loups was "christianised" to become "Sainte-Anne-du-Loup"  (Morice,  1912), a name that assimilates the "savage" and the patroness of the voyageur.  3.3  Concluding remarks.  The two investigations lead us to a similar conclusion. If the Metis territoriality had become a central element of the Nord-Ouest through most of the 19 century, the advance of th  colonisation, which significantly increased in the 1870s, brought structural changes that deeply affected it. Forced to abandon certain traditional activities and to develop others judged to be more civilised, the Metis image of themselves undoubtedly changed. But the most critical changes occurred in space, the appropriation of which was a colonial affair. The Metis were gradually dispersed and dispossessed of a clear land base. They saw their material superiority on the land melt like snow. What had once been a land of the cart, became one of farms and railroads. Maps leave little doubt about this after 1870. In fact, as I have shown, they often predicted such a reality. Metis narratives also address these changes in clear tones of nostalgia and loss.  125  However, some crucial differences have emerged through these two types of analysis. Metis oral tradition is revealed to be a much more fertile description than maps of the extent and deepness of Metis experience, perception, and representation of space. While maps do display Metis territorial markers, their portrait is partial, in all senses. On the 19 century maps the th  hybridity and ambivalence surrounding Metis territorial markers disappear. The Metis presence on maps becomes clearer, more precise, and homogenous. This creates the impression that the Metis reality can be fixed in space, regardless of its spatial and identity mobility. This had great advantages for colonial purposes. A "fixed Metis" is one awaiting to be taken in charge by colonial means. Such a Metis is better suited for civilisation. On the contrary, analyses of Metis oral tradition reveal identity and spatial mobility to be the main figure of Metis territoriality. Metis narratives convey a greater diversity of identities (Jannetta, 1994, p. 94) and of territorial discourses.  .  On the other hand, maps are the best indicators of the influence of Metis territoriality on colonial representations of space. If there is nothing surprising in the fact that Metis territoriality is central to the narratives I have analysed, one would not automatically suspect the same on maps, the purpose of which was not the representation of Metis reality. As incomplete as the cartographic representation of Metis territorial markers is, it nonetheless proves how inescapable Metis reality was in the geography of the region and how sophisticated the means of antimetissage were to be.  Ultimately, these two investigations provide complementary views of the same phenomenon.  Figure 3.1  Detail from S.J. Dawson's Plan Shewing the Region Explored by S.J Dawson and His Party...,  1859.  127  Figure 3.2  Peter Fidler's Map of Red River District, 1819.  128  OF •JnVMAMf*  '  /Tlttr  r-ft  44—.6- - SJ—*t—  • ftM*****/ !  —:  .. ;  o  THE.  w  1  • .-  STJIT/ST/CAL. • .  • •  J\  "I  50 So' ;  50-3£l..  50  EM  83'  =-  si  I  ' si  s5  H35  49  ^  Figure 3.3  •A-  .  A . L . Russell's Map of the Province of Manitoba,  1871  130  Locality of the HALF-BREED  Rebellion...,  1895.  131  3j  81  oI  o1  ij si 1  m i l  Figure 3.6  W . H . Holland's Map of the Seat ofRiel's  Insurrection...,  1885.X>r^  I  132  Chapter 4:  Metissage and Exclusion in the early 2 0 Century: The Metis, an "Official" th  Nameless Reality.  De-territorialiser une ethnie est la meilleure facon de la voir disparaitre pour sefondre dans un magma sociologique. — Joel Bonnemaison, "Voyage autour du territoire."  Ethnicity, de-territorialisation and assimilation are components of the equation of acculturation proposed by the French cultural geographer Joel Bonnemaison (1981, p. 256). If one accepts that "to see disappearing" ("voir disparaitre") implies "to pass out of sight" as much as "to cease to exist," I would argue that this equation applies to the Metis reality in the first half of the 20 century. During this period Metis territoriality was irremediably changed by the th  reinforcement of economic and political exclusions. The Metis endured material, political and symbolic losses that deeply impacted their experience of space and their collectivity, particularly on the prairie where a national sense of community had arisen in the 1800s (Stanley, 1978b, p. 76; St-Onge, 1984b, p. 1).  This chapter is not a detailed look at early 20 century Metis territoriality. As my aim is th  to explore the historical significance of Metis geographies in Canadian socio-spatial transformations, it is beyond the scope of this thesis to treat Metis territorialities during this period in great detail. If over the years successive Canadian political regimes had to cope with metissage — generally by opposing — this was hardly true in the early 1900s. Then, I will argue, the Metis were "erased" from Canada's cultural, historical, political and territorial discourses, and were considered a marginal, if not "invisible," component of the country's socio-cultural geography. However, Metis territoriality did change during this period in reaction to exclusion and to Canada's Aboriginal policy. As these changes need to be addressed to better understand the contemporary Metis, this chapter is a sketch that summarises the impact of such exclusion and of Canadian policy on the Metis sense of identity and territory. Essentially, the chapter is a link between the historical content of the two previous chapters and the contemporary material of the next two.  133  4.1  Cartographical deletion: the Metis, a "past" people.  As Sealey and Lussier (1975) put it, the Metis were a forgotten Aboriginal people until their re-emergence in the late 1960s and early 1970s. Information relating to the Metis in this 1  period is sketchy and the existence of Metis territoriality is barely "perceptible," a conclusion demonstrated by two sources of information relevant to my study.  The first is cartographic. Apart from a few French place names that persisted on the prairie, the official topographic cartography in Canada makes almost no mention of Metis territorial markers. In fact, this cartography is the logical consequence of the trends I observed for the Quebec Metis in the late 18 and early 19 century, and for the Prairie Metis in the last th  th  decades of the 19 century — the cartographic fading of Metis geographies. In Quebec, as I th  showed in chapter 2, the obliteration of the hybrid maps of the French regime is accomplished. On the prairies, the Metis are no longer substantial informants for mapmakers, and Metis territorial markers are hard to identify. Metis geographies are confined to a marginal, virtually non-existent, status.  Official mapping was produced, in the early 20 century, by three governemental th  agencies: the Geological Survey of Canada (1842-1947), the Department of Militia and Defence (1903-1931), and the Department of the Interior (1873-1936) (Dubreuil, 1989, p. 1). Created in 1871, the Department of the Interior was the most active agency in mapping western Canada, and is particularly known for its "Sectional maps of Western Canada" series. If some of these sectional maps were made at the scale of six-miles to one inch, most were at the larger scale of three-miles to one inch. These maps were never meant to be published for the general public, but were for government use (idem). Like maps of the late 19 century, the Metis presence is not th  totally erased on the Department of the Interior's topographic representations of the West. One notes river lots on sectional maps of La Biche (1918), St. Ann (1933), and Carlton (1935, Figure 4.1), three areas known to have Metis settlements. Sectional maps also sometimes show light dashed lines that the legend identifies as trails; they are likely old Metis cart trails. However,  This view is supported by the Canadian census, which stopped counting the "Half-breeds" in 1941. While a "Metis ancestry" section was added to the 1981 census (Dickason, 1996, p. 365), it is only since 1996 that Statistic Canada started also amassing data about Metis self-identification. Moose Mountain (1915, 1940) and Wood Mountain (1916, 1930) sheets show a rich variety of these trails. 1  2  134  Metis reality is no more than marginal on these maps. Even the Indians are more visible, as the reserves are clearly identified by the legend and by the striking color of their boundaries.  There were also a considerable number of atlases of Canada during that period. These atlases were generally the work of private publishers, although some of them were produced by government agencies (Department of Immigration and Colonization, 1922; Department of the Interior, 1900). They were either adressed to a general public (C.S. Hammond & Co., 1946; Morrison, 1958; Rand, McNally, and Co., 1905; Taylor, 1948; Thomas Allen, 1920) or specifically to school children (Dent, 1949; Lewis & Campbell, 1951; Watson, 1958). In either case (see Figure 4.2 & 4.3), the Metis presence is reduced to few place names, which are surrounded by marks of colonisation such as settlements, roads, railroads and national parks (C.S. Hammond & Co., 1946). The Indians' presence is similarly effaced, except for a few atlases, which meagerly represent reserves (Department of the Interior, 1900; Rand, McNally, and Co., 1905). The textual portions of atlases confirm these cartographic trends as they focus on topics such as fishing, mining, manufacturing, exports-imports, transportation, and cities/towns. The Metis and Indians are not mentioned in the "historical" section of these texts. Although most of the general and school atlases were largely promotional — they emphasised the socioeconomic development of the country, the progress of colonisation and, to a certain extent, the growth of the Canadian Nation — promotion was the explicit function of some of them. This is particularly obvious in the Department of Immigration and Colonization's Canada  Descriptive  Atlas (1922). As the name of the Department indicates, this atlas was intended to promote abroad — in the British Isles in this case — the colonisation of Canada.  Other types of mapping sometimes recognise edges of Metis reality. Yet, such recognition generally remained partial, marginal, and quite inadequate to allow the recuperation of Metis geographies in the first half of the 20 century. The map produced by the Saskatchewan th  Golden Jubilee Committee in 1955 is a case in point (Figure 4.4). This commemorative map is 3  essentially made of historical and pictorial features and provides a scenic representation of Saskatchewan's territorial history. It shows Metis territorial markers, with icons representing Metis settlement (St.Laurent de Grandin, near Batoche on the South branch of the Saskatchewan River), freighting and buffalo hunting activities, and it mentions cart trails. However, such a  3  The HBCA hold a similar commemorative map published in 1955 by Stanley Turner [G4/182].  135  portrayal is, at best, an impoverished representation of Metis geography. The recognition of Indian and Metis geographies is limited, and is strongly downplayed by more dominant information emphasising the Euro-Canadian contribution to Saskatchewan, with scenes and icons referring predominantly to newcomers (fur traders, explorers, missions, and new immigrants) and modern ("civilised") developments (agriculture, mines, oil wells, bush planes, a university, a legislative building, etc.). Part of the information related to Indians and Metis focuses on land dispossession or shows how the Metis "rebels" were brought under Canadian control. Two scenes mention Indian treaties (#4 at Fort Qu'appelle and #8 at Fort Pitt). The Metis "problem" is mostly emphasised by icons depicting two rifles in a cross-like pattern and marking each battle site during the 1885 uprising. More important is the fact that among scenes 4  and icons marking Metis territoriality (positive and negative), none refers to Metis life in the early 20 century. On a map in which nearly 50 per cent of the contents concern the 20 century, th  th  it is clear that the Metis were considered solely as inhabitants of a receding past.  Was this really so? Could one argue that the Metis reality was minimised to hide a disturbing influence on mainstream society? Could maps be tools used to express a subtle antimetissagel What I argued in chapter 2 with regard to the erasure of Native influence on the cartography and geography of la Nouvelle-France  could perhaps be argued here. On the other  hand, if such a "deconstruction" of these maps might reveal a certain instrumental erasure of Metis reality, I doubt its influence was more than marginal. Much more likely, the lack of information about the Metis on the map was a result of the mapmakers' focus on colonisation and development. Representing the Metis reality was a much lower priority in the mapmakers' mind. Metis invisibility on maps is a reflection of the processes of marginalisation and discrimination socially and spatially at work, rather than the opposite. As I will argue below, the dominant society held that the Metis, like other Aboriginal peoples, would be integrated eventually into the "civilised" population of Canada. Maps are socio-cultural representations that mirror and convey "truths" as societies construct them.  The second source of information lies with the early 20 century scholars who th  worked on the Metis. These scholars, for the most part historians and ethnographers, focused principally on 19 century Metis history at Red River and on Louis Riel, at the th  4  There is also a scene representing the Canadian army at Loon Lake that is accompanied by a relevant mention:  136  expense of the contemporary Metis. In English Canada, Stanley's Birth of Western Canada (1936) is a good example. Stanley explored at length the importance of the two Metis "rebellions" and the central role of Louis Riel, but never mentioned a persistent Metis reality. In fact, Stanley did not intend to write a native history of the Metis. As his book title indicates, he rather emphasised the importance of Louis Riel's "rebellions" in the making of the Canadian nation (Ray, 1982, p. 93). From such a perpective, one understands that the Metis of the first half of the 20 century were not of interest to the th  Canadian historian. The same can be said of Joseph Kinsey Howard's Strange Empire (1952). Although Howard was not a professional historian and wrote for a wide public, his book is recognised as one of the best ever written about the Metis nation and Louis Riel (Flanagan, 1974, p. 738). His last chapter is devoted to the hanging of Louis Riel, as if the Metis' fate were tied to that same rope. In this respect, French historiography is much the same. Auguste de Tremaudan's Histoire de la Nation Metisse dans l'Ouest canadien (1936), A . G. Morice's Histoire de L'Eglise catholique dans l'Ouest canadien (1912), and  Clovis Rondeau's Montagne de bois (1923), all offer exhaustive depictions of Metis society and culture, but exclusively study the 19 century. An exception is the work of the th  French ethnographer Marcel Giraud, who made one of the most complete studies of the Western Metis. In Le Metis canadien (1945), Giraud devotes a chapter to the contemporary Metis. His description, however, is mostly negative, portraying ill-adapted "nomad" Metis condemned to disappear by their own physical and moral "agonie" (p. 1251). Such a portrait does little to enhance the image of the Metis in mainstream society. Giraud's portrait is only a reflection of western thought with regard to the ideology of primitive people, the principles of which he adopts (Tough, 1989). In Quebec, as I noted in chapter 2, a similar neglect of the Metis is apparent in the academic opposition — chiefly promoted by le chanoine Lionel Groulx in the 1930s (Mouhot, 1999, p. 56-58; Smith, 1974) — to the "metissage thesis," which challenged the intellectual and nationalistic construction of a "pure" (French) canadienne identity.  Things have changed during the last three decades, and the general interest in understanding and documenting the early 20 century Metis has notably increased. Metis and th  Non-Status Indian organisations have had a great deal to do with this in promoting research on  "last hostilities on Canadian soil: Skirmish at Loon Lake, June 3, 1885".  137  Metis issues (Ray, 1982, p 93). Along with studies focusing on the political leadership of the Alberta Metis (Bell, 1994; Dobbins, 1981; Driben, 1985; Sawchuk, 1998), there has been a burgeoning of primary sources such as Metis (auto)biographical accounts and oral history 5  projects. Although these sources do not primarily aim to describe Metis territoriality of the 20  th  century, they offer a picture of Metis society that includes elements bearing on identity and territory.  But analysis of these sources, especially the oral histories, has its limitations. First, they represent over 200 hours of audio-taped interviews, not always transcribed, nor always devoted exclusively to Metis people, spread all over the prairie. Second, as I noted in chapter 3, oral 6  histories are performed within a specific context — the 1970s, 1980s, and 1990s in these cases — and tell us as much about informants' "present" as about the past to which they refer (Cruikshank, 1994, p. 414; Cruikshank, 1998, p. 41). If this does not disqualify the historical use 7  of this source, it does suggest that the oral records would benefit from comparison with written or cartographic documents, which is hardly possible here. Third, the increase of this type of 8  information reveals more about the contemporary nature of scholars' concerns — they are "performing" them too — and about the influence the Metis have on Canadian society today, than about their influence during the period under investigation.  The best known are Howard Adams' The Prison of Grass (1975) and Maria Campbell's Haljbreed (1973). More recently, the audiotaped Au temps de la Prairie has been published in French by Marcien Ferland (2000), relating the life of Auguste Vermette of Manitoba. In Alberta: "Jasper-Yellowhead Oral History Project", Jasper-Yellowhead Museum and Archives, Jasper; "Lac La Biche Mission", Lac La Biche Mission Historical Society, Lac La Biche. In Saskatchewan: "Metis History", Gabriel Dumont Institute, Regina; Saskatchewan Archives Board (Regina), "Biographies of Two Metis Society Founders, Norris and Brady", "Chipewyan and Metis People of La Loche", and "Ethnocultural Groups of Saskatchewan: The Metis". In Provincial Archives of Manitoba: "MeTis Oral History Project", two phases conducted in 1984 and 1985 by Nicole St-Onge; "Metis Women of Manitoba Oral History Project", conducted in 1993 by Lorraine Freeman and Doreen Breland-Fines (Fortier, 1993). As an example, one could compare St-Onge's (1984a; 1985) oral history projects with Freeman and Breland-Fines' (1993). A quick glance at both sets of interviews reveals that respondants in the latter project refer more to the 19* century Red River tradition, emphasising the role of Louis Riel and Gabriel Dumont. One could think that the nationalist movement that drained the prairie Metis affiliated with the MNC of Yvon Dumont in the early 1990s has something to do with the emphasis on the Metis leaders of the past. According to Morisson (1995), this period of "glory days" peeked with the succesfull negotiation of prairie Metis self-government within the nationally denied 5  6  7  Charlottetown Accord in 1992.  Although the previous chapter was partially based on a non-cartographic analysis of Metis territoriality, such information was compared with an exploration of Metis territorial makers as incribed in colonial maps. It was also supported by substantial secondary historical sources. As I have noted, there is hardly an equivalent for 20 century Metis history. 8  lh  138 Moreover, the study o f 2 0 century Metis history still represents an infinitesimal part of th  the work on the Metis. If the rise o f social history has allowed researchers to decenter their focus from the Red River settlements and to express broader visions of Metis reality (Miller, 1988), it has done little to stimulate their attention i n the 2 0 century. This period of Metis history th  remains in the specialist domain and is hardly passed on to a larger audience. The marginality of such a topic in general histories supports this point. Gerald Friesen's Canadian Prairies (1987) is an appropriate illustration. Friesen exhaustively documents the Metis contribution to Prairie history in the 1 9 century, but says nothing about the Metis during the 2 0 century. The same th  th  can be said o f edited books such as John Foster's Developing West (1983) and Howard Palmer's Settlement of the West (1977), both of which include only "conventional" studies about the Metis. Missed by such accounts are, for example, the foundation o f the Metis settlements in Alberta in the 1930s and their consequences for the Metis arid for provincial authorities, which have received expert attention: e.g., Catherine B e l l (1994), Paul Driben (1985), and Joseph Sawchuk(1998).  The slight importance academics have attached to early 2 0 century Metis history is th  echoed by the Metis themselves. Without disregarding this part of their history altogether, the Metis accord it minor historical significance. Rather, they emphasise their roots among the Red River Metis o f the 1 9 century where a positive identity is readily generated. This emphasis is th  indicated by the MNC definition quoted in the introduction and by the symbolic promotion o f a material culture and a way o f life that refers to the Metis' "Golden A g e " (Red River cart, sash, bead artwork, and buffalo hunt). A similar conception of Metis history can be observed in Quebec, as I w i l l demonstrate i n chapter 6.  4.2  Canada's Aboriginal policy: ambivalence, metissage, and spaceless Metis.  During these years the administration of aboriginality in Canada was marked by ambivalence. A t the same time that the Metis were officially expelled from the "Aboriginal family" and previous Metis geographies were absorbed by the colonial perception and cartography o f space, Canada was generating, through its legal handling of Aboriginal people, a new Metis geography — or "metisation" to use Jacqueline Peterson and Jennifer Brown's  139  expression (1985, p. 5). The 1876 Indian Act may be considered to have introduced a major 9  shift in the processes of metissage in Canada.  The idea behind the Indian Act was, ostensibly, the protection, civilisation and assimilation of indigenous people into mainstream Canadian society (Cairns, 2000; Dickason, 1992; Flanagan, 2000, p. 45; Tobias, 1983, p. 39). There was nothing new about such objectives, for they were the logical conclusion of the Euro-Canadian perception of Aboriginal people as primitive (Laroque, 1983, p. 86). Since Montaigne's myth of "du Bon Sauvage" in the 16  th  century (Delage, 1992, p. 169; Dickason, 1992), the belief prevailed that primitive people were ill-adapted to the modern way of life introduced to them by western civilisation, and did not have the means to protect themselves from eventual Euro-Canadian encroachments to their land and way of life (Tobias, 1983, p. 40). Europeans had developed an evolutionary perspective of history and the human condition. Not only were primitive and civilised people different in this perspective, but they were also differently located in time, civilisation being the advanced stage of human achievement. Primitive people, these "wandering" souls, were thought to make a  v  superficial use of the land. Farming was seen as a more advanced and efficient way of using land and cultivated land became a visual evidence of the superiority of western civilisation (Harris, 2002, p. 51), and some scholars (e.g. Adas, 1989; Headrick, 1981) have argued that in the 19  th  century the increasing rates of industrialisation and technological change reinforced the ideology of progress, and the sense of superiority over non-western people that Europe had developed since the 16 century. Hence, the colonial vocation for "civilising" and assimilating les Sauvages th  du Canada. Under these same postulates, Canada justified its opposition to the Northwest Metis grievances in the 1880s, thereby forcing these Metis to "rebel" (Sprague, 1988, p. 1).  The context of the Indian Act set it apart from previous attempts to assimilate Indians. The principal limit of earlier efforts was the need to preserve the "primitive" way of life on which the fur trade economy depended (Innis, 1999, p. 392). "Civilised" agrarian Indians would make ineffective hunters. Consequently, policies of assimilation had always suffered a lack of cohesion and had failed to assimilate. With the decline of the fur trade — which can be traced as far back as the first decades of the 19 century in Quebec — this limitation to the "civilising" of th  This is a unique legal text that had brought together disparate policies which were spread from 1851 to 1876. Although it has been subjected to modifications since 1876 (notably in 1952 and 1985), the "spirit" and rationale of the Act has mostly remained unchanged.  140  Indians was greatly diminished and the path to assimilation cleared. Although an amalgamation of diverse policies and legal actions that were implemented over the years, the Indian Act was drawn in more coherent lines and was a more consistent effort to coordinate the means by which Indian integration would be realised.  The Indian Act was based less on the belief that Indians were of an inferior race than on the premise that they had the potential to become civilised and full citizens of Canada. If the "primitives" were given the opportunity, they would eagerly abandon their Aboriginal identity and take part in the country's development. They could not be forced, but could be persuaded to assimilate. Persuasion was conceived as a two-fold strategy (Simard, 1992, p. 719). First, the Indian was to be legally defined and categorised by the state, with the creation of a specific status. Thought of as a temporary condition (Green, 1985), this status would allow Canadian authorities to direct measures of assimilation or "enfranchisement" (Tobias, 1983). Second, the Indians would be granted a specific piece of land, the reserve, which was to be allocated and situated in such a way that it would act as a "training" device (Tobias, 1983, p. 39). Overall, 10  this double strategy was based on the creation of a social-legal boundary (Indian status) and a geographical boundary, the reserve, both conceived as ways of demonstrating the advantages of assimilation and of convincing the Indians to cross the thresholds between their "state of nature" and Canadian culture.  Metissage was central to this strategy, although Canadian authorities never admitted that this was the case. No one believed that the Indian would assimilate into an Euro-Canadian way of life overnight as a result of legal and geographical pressures. There was implicit belief in a transitional phase that would allow the gradual integration of the indigenous component into the dominant society. The historian William B. Cameron concluded in 1900 that:  although the future of the redman looks gloomy it is impropable that his blood is altogether to disappear, for there is one agency at work which promises to preserve to posterity some of  This, of course, proved to be true mostly in the east, especially in Lower Canada where colonisation was well advanced when Indian policies were implemented. Aboriginal land was already encroached by the plows and axes of new settlers. The situation was different in treaty areas, notably on the prairie. As the treaties generally preceded the massive arrival of settlers from Canada and Europe, the reserves were not only pieces of land set aside to protect Indians from Euro-Canadian encroachments, but were also the residual land the Indians managed to keep. Put another way, the reserves were the only sections of land that colonial authorities would not take over and open up for colonisation. lu  141  the best characteristics of his race — and only one. It is the sole method of ci