Until the late 1920s, anarchism was still a significant presence in Chinese radical thinking and activity, and till the middle of the decade, gave serious competition to the Communists. The essay discusses the nature of the anarchist movement in China, anarchist criticism of Bolshevik Marxism, and anarchist revolutionary strategy and activity during 1921–1927. It argues that while anarchists were quite innovative with regard to revolutionary strategy, their repudiation of organized power deprived them of the ability to coordinate revolutionary activity on a national scale, and what success they achieved remained local and short-lived. Indeed, the Communists were able to make better use of anarchist tactics than were the anarchists themselves. Anarchist critique of power rested on a denial of a center to society (and history). While this undercut the anarchists' ability to organize the revolutionary movement, it is also revealing of a basic problem of socialist revolution: the problem of democracy. In ignoring the anarchist critique of power, the successful revolutionaries deprived themselves of a critical perspective on the problem of socialist revolution, and were left at the mercy of the new structures of power that they brought into existence. Hence the importance of recalling anarchism.

Historical truth, like other kinds of intellectual understanding, proceeds through argument and discussion. Thus, we should be grateful to Jonathan Zeitlin for his rigorous responses to a body of scholarship which he categorizes as “rank-and-filist”. Although much of Zeitlin's argument is unexceptional, its proposal to consign as erroneous and irrelevant the scholarship that fits his “rank-and-filist” paradigm requires some examination. Most troubling are the procedures he uses to make his case; the prescriptive advice he offers though not without merit is also highly problematic. I shall discuss these objections in a moment. But first we must define the problem as Zeitlin sees it.

Labor history continues to progress. It continues, too, to transform itself from a field staffed by enthusiasts and partisans into a discipline peopled with scholars sympathetic to their subject but also highly conscious of the methods, theories and interpretive frameworks within which they work. The last few years in particular have witnessed a considerable intensification of scholarly controversy within labor history and a growing sophistication of debate. To some extent the discussion parallels the ongoing conversation among social historians about the boundaries of social history, the power of primarily social explanation and the role of language, culture and politics in social history. Unfortunately, however, the stepping up of debate in labor history coincides with a dip – temporary, one hopes, but perhaps more long-term – in the fortunes of labor movements themselves. This conjuncture seems to have produced a sense of pessimism and malaise as the sentimental accompaniment to many of the debates within labor history.

Few theoretical paradigms in labour history are so deeply entrenched as “rank and filism”. It was only to be expected, therefore, that a frontal assault on its assumptions would provoke a vigorous reaction. The responses to my article by Richard Price and James Cronin thus offer a welcome opportunity to clarify the theoretical claims of “rank and filism” and reassess its empirical plausibility as an interpretation of British labour history. But as in any clash between rival theoretical perspectives, the points at issue in this debate extend beyond factual disagreements to the meaning of basic concepts and the standards of proof involved in their assessment, and neither party fully recognizes itself in the account of their ideas presented by the other. No accumulation of discordant facts can conclusively disprove a theory, as students of scientific revolutions have demonstrated, but readers will have to judge for themselves whether the counter-evidence I have presented amounts to minor discrepancies which can be satisfactorily accommodated within the assumptions of “rank and filism” or fundamental anomalies which necessitate the abandonment of the paradigm itself.