1
Backwards Up Niagara Falls:
Space/Time, Image/Text and the Biases of Information
1
Introduction
James Carey once made a wry observation concerning the relationship between two of
the most important, and probably most idiosyncratic, North American communication
theorists. Appropriating Oscar Wilde’s witticism about Niagara Falls, he argued that the ‘arc’
which ran from Harold Innis to Marshall McLuhan ‘would be more impressive if it ran the
other way’ (Carey, 1992: 142). The same could be said about the enthusiastic resurrection of
McLuhan’s writings in what is inescapably called ‘the information age’: it would be more
impressive if it were to lead to a general reanimation of Innis’ work. Unfortunately, the 1990s
and early 2000s have been almost a blow-by-blow replay of the 1960s and early 1970s. Quite
a number of recently published books and papers have - with varying degrees of
sophistication - proclaimed McLuhan’s relevance to the era of computer-mediated
communication, with some practically canonizing him as the tragic prophet of virtuality.
However insightful the more serious of these reappraisals may be as individual analyses, en
masse they have been accompanied by a sonorous silence regarding Innis. Almost no
published work has engaged with his thought in a way that implies that it might be remotely
useful for understanding the advent and implications of new media.
2
This paper is predicated on the idea that this is a great shame. For Innis’ thought,
notwithstanding the notorious difficulty of his writing style, actually intersects powerfully
with a number of central approaches within media and cultural studies, among them the
political economy of communications and the sociology of culture. Additionally, Innis’
central conceptual tool – the spatial or temporal ‘bias’ of communications media – is
Page 1 of 31 
Convention