A Sceptical Review of Benedict Anderson’s Imagined Communities

This piece was originally published on 03/08/20

I took three big ideas from Imagined Communities. First, Anderson defines a nation as a “sovereign limited imagined political community”. Second, Anderson argues that the origin of nationalism is in Latin American creole resistance to the metropole between 1760 and 1830, not in Europe. The development of European nationalism was characterised by ‘official nationalisms’ — “an anticipatory strategy adopted by dominant groups which are threatened with marginalisation or exclusion from an emerging nationally-imagined community.” (p.101 in the 2016 Verso edition). Third, Anderson is focused above all on language. The entire book is concerned with vernaculars, languages-of-state, print-language and forms of words. Language is the main subject of his analysis, and it is language that binds these imagined communities together — in contrast, for example, with pre-bourgeois aristocracies bound together by kinship. He returns repeatedly to the idea of ‘Peruvianization’, which “shows that from the start the nation was conceived in language, not in blood” (p. 145).

These points are central to the book, and have been immensely influential; yet throughout the book Anderson makes arguments which seem wrong, especially when he’s outside his own area of expertise in southeast Asia. His assessment of medieval history is seemingly based entirely on two books by Annales historians, Marc Bloch’s La société féodale (1939) and Lucien Febvre’s L’apparition du livre (1958). As such, when he makes generalisations, they rest on a limited evidence base. For example, I think his characterisation of Latin speakers as “tiny literate reefs on great illiterate oceans” (p. 15) and his argument that “in a pre-print age, the reality of the imagined religious community depended profoundly on countless, ceaseless, travels.” (p. 54) make a lot of sense; but his assertions that “the universality of Latin in medieval Western Europe never corresponded to a universal political system.” (p. 40), the non-existence of medieval slavery (p. 60), or that “before the age of print, Rome easily won every war against heresy in Western Europe because it always had better internal lines of communication than its challengers.” (p. 39) strike me as misleading.

To take more modern examples, in his chapter on racism Anderson makes a strong case that “the dreams of racism actually have their origins in ideologies of class, rather than in those of nation: above all in claims to divinity among rulers and the ‘blue’ or ‘white’ blood and breeding among aristocracies.” (p. 149); he also talks about ‘solidarity among [aristocratic] whites’ from different metropoles in a colonial context as being evidence for this thesis. This argument comes from a study of Nazi Germany, but ignores the work of scholars on racism in the Atlantic world, where there is a powerful argument that racism originated in a post-hoc justification for slavery and economic domination. Another inconsistency comes in his analysis of colonial schools, like the Ecole Normale William Ponty in Dakar; students at these schools “knew, even if they never got so far — and most did not — that Rome was Batavia [i.e. the centre of Dutch Indonesia], and that all these journeyings derived their ‘sense’ from the capital, in effect explaining why ‘we’ are ‘here’ ‘together’.” (p. 122). But if nationalism in the colonies was created by travel to the centre, then we must question his characterisation of modern nations as “boundary-oriented and horizontal”, in contrast to pre-vernacular societies built around sacred scripts which were “centripetal and hierarchical” (p. 15).

It’s both a strength and a weakness of Imagined Communities that its central three points aren’t actually invalidated by these problems, as Anderson himself recognises: after identifying three main factors behind “the revolutionary vernacularizing thrust of capitalism”, he allows that “it is quite possible to conceive of the emergence of the new imagined communities without any one [of these three main factors], perhaps all, of them being present.” (p.42). The attraction of the book lies in its combination of a simple definition for nationalism with a broad range of examples and digressions; but it seems that the latter isn’t necessary for the former. Even if all his arguments concerning the period before 1500 (the date which he quotes Benjamin in calling the beginning of ‘the age of mechanical reproduction’) are wrong, it’s not clear that his three theses fall.

It feels strange to raise so many issues with such an acclaimed work; all the more so when those disagreements are concerned with minutiae rather than the grand shape of the argument. Nonetheless, when faced with a thesis like Anderson’s, it’s difficult to engage with it at its highest level of abstraction. So we are left with a few questions: What are the historical facts on which the narrative really depends? Does the narrative still have explanatory power when we extend it to new facts? And is there any way to properly engage with historical narratives except by analysing their factual foundations?