The economic history of Canada has been dominated by the discrepancy between the centre and the margin of western civilization. (Innis, The Fur Trade in Canada, 1930)

As described by McLuhan in ‘Media and Cultural Change’, his 1964 introduction to Innis’ The Bias of Communication, the structural pair of centre and margin is subject to a range of expression:

Visual technology creates a centre-margin pattern of organization whether by literacy or by industry and a price system. But electric technology is instant and omnipresent and creates multiple centres-without-margins.  

What Innis calls “the discrepancy” between centre and margin names not only their relationship in a particular case (here, the fur trade in Canada) but also, as McLuhan explicitly points out (and as is clearly implied by Innis), it may be taken to name a range of relationships which characterizes ‘events’ as different as “literacy”, “a price system” and “electric technology”. A “discrepancy” can hardly be unique.

It is not the case that these different happenings come first and then the relationships they exhibit between centre and margin are abstracted from them.  Instead, from the vantage of media analysis, the range of relationships is prior and the events with their different modalities of centre and margin are subsequent.  

Innis’ “discrepancy” names the relationship or medium between the two poles of centre and margin which governs or ‘dominates’ (as Innis says) their expression and which thereby determines the meanings of those poles. This structural determination may be seen in McLuhan’s manipulation of the pair. In the “visual (…) centre-margin pattern of organization” McLuhan puts forward an (a≠b) relation where the ‘centre’ pole is preferred to the ‘margin’ one (so either a≠b or a≠b). But when McLuhan comes to write of “multiple centres-without-margins” he is not indicating some centre-only state which might be mapped as a (alone) or b (alone). Instead, as the reference to plural “multiple centres” makes plain, the form of “electric technology” is still complex and certainly does not abrogate the elemental a/b structure. So McLuhan does not use the phrase “multiple centres” to indicate something about the single ‘centre’ pole of a centre/margin structure, but, instead, he uses it to describe an inclusive or “multiple” preference-state of that structure: (a=≠b). This is an a/b structure “without-margins” only in the sense that there is no pole of it which is without “preference” and which would therefore be ‘marginal’. In fundamental contrast to a case with one-sided “preference” and with corresponding marginality of the other pole, here both poles are, in this new sense, “centres”. Absent the literate habit of privileging ‘centre’ over ‘margin’ (aka ‘imperialism’, ‘orientalizing’, etc) this could also be expressed as centre =≠ margin.

In fact, margins are still necessarily present in “multiple centres-without-margins” since (eg) there could be no plurality of “multiple centres” if there were no margin or border or gap separating or differentiating individual “centres” from each other. In this latter sense, ‘margin’ takes on the meaning of R in the aRb relation and is therefore not only not absent but, as the medium, is just the elementary message: “the gap is where the action is”, “the medium is the message”, “the medium is the massage” . . .