Fictional Settlements: Footnotes, Metalepsis, the Colonial Effect

I am going to argue that the nineteenth-century novel is anomalous using as an example an anomalous nineteenth-century novel. The anomalous novel, Catharine Parr Traill’s Canadian Crusoes (1852), is not well-known now, although it was well reviewed and popular in its time, and for about fifty years thereafter. A genre fiction in at least two ways—as a young adult novel and as an adventure fiction—it is also an emigration novel, which may or may not be a genre. It was written in Canada by a pioneer who is often described as “British-Canadian” and who began writing children’s books at the age of sixteen to support herself and her family after her father died. The field, in Pierre Bourdieu’s sense, might best be described as that of “colonial letters.”¹ I mean “letters” in both the sense of belles lettres and in the sense of epistles written home.² Anglophone Canadian fiction and travel writing of the nineteenth century was not usually read by Canadians, but rather by Britons in Britain, who might or might not be prospective Canadians. The writers in the field of colonial letters who imagined and constructed fictional settlements such as the ones proposed in Canadian Crusoes “participate in domination, but as dominated agents; they are neither dominant, plain and simple, nor are they dominated.” Parr Traill, as the wife of a British army officer, is mildly privileged in the colonial social hierarchy, but just by virtue of having to participate in emigration, she is among the dominated citizens of nineteenth-century Britain. Her participation in the representation of empire is accordingly complex: her writing encourages emigration to Canada’s forested “north” and also depicts the intense hardship and tragedy that so often attends it.

Bourdieu has argued that “literary fiction is . . . a way of making known that which one does not wish to know.” We can bear novelistic revelations because they remain “veiled.”³ It is this figure that I wish to amplify and revise in what follows. I want to suggest a specifically “colonial effect.” The idea of the “colonial” in this effect must be understood both literally and figuratively. It refers both to the way in which the novel helps us to imagine and colonize actual space, in part through the navigation of represented space, and it also refers to the idea of the colony as a place over which a fantasied domination can always preside. Dorothea Brooke longs for a “colony”; Gwendolen Harleth for an “empire”; Robinson Crusoe is of course not the first or last fiction in which such dreams come true.

As Edward W. Said spent much of his career arguing, most explicitly in Orientalism (1978) and Culture and Imperialism (1993), empire depends on substantial epistemological and narrative support. Novels are perhaps always in some sense colonies for their authors and readers: small worlds under the control of the author who makes them, and then the reader who can turn the pages or not, who can imagine the world represented or not, live in it for a time or not, can believe in this or that aspect of it or not. The colonial effect I want to describe suggests that part of the work that an imperial power requires from the realistic fiction that novels tend to proffer is precisely an important flexibility between fantasy and reality. Realism’s weird—although thoroughly naturalized—combination of fictionality and factuality, in its awkward form in the anomalous Canadian Crusoes and its elegant form in more canonical nineteenth-century novels, makes known that which we do not want to know about our world but which we must know at some level, or at some moments. In some sense, realism makes social reality known literally: actual places and historical events mingle with fictional places and people. Realism insists on some degree of reference. The delicate but persistent connection between fiction and reference makes the form of the nineteenth-century novel anomalous (and this form persists beyond the nineteenth century in any novel that continues to be realistic and thus referential). It is most...