A brief reply to Bhambra and Holmwood

Abstract

This a reply to Gurminder K. Bhambra and John Holmwood’s response to my article discussing their book Colonialism and Modern Social Theory.

I thank Gurminder and John for their response to my article. I am also grateful to them because their book gave focus to a recent preoccupation of mine: what exactly I got myself into when I decided to study sociology at university 60 years ago. I have come to the conclusion that this ‘discipline’, to which I am still committed, has no core and no clear boundaries, despite efforts to present it as if it has. This relates directly to a point I made in my article: that Bhambra and Holmwood are highly selective in characterizing the discipline they are wishing to decolonize, that they ignore the huge range of types of work coming under its heading, largely reducing this to social theory, whose relationship to sociology is a matter of dispute (on both sides). They insist in their response that ‘the association of sociology with the modern, and the modern as deriving from social changes taking place from the late 18th century onwards, is quite standard within social theory’, supporting this with an appeal to the work of Habermas (Bhambra and Holmwood 2026: 155). But this simply reinforces my point: the relationship between this metanarrative and what many sociologists do today is remote, at best.

We also seem to disagree quite sharply about one or two other key issues. However, understanding the exact nature of any disagreement is often difficult; and, while dialogue is the only means available, it is never straightforward. Unfortunately, this seems to be illustrated by our exchange here. While Bhambra and Holmwood declare that ‘intellectual argument is advanced through criticism and disagreement’ (Bhambra and Holmwood 2026: 160), I don’t see much sign of advance in our discussion. ‘Puzzlement’ (see Bhambra and Holmwood 2026: 157, 158 and 160), and complaints about misinterpretation, seem to be the main products.

In fact, there even appears to be a misunderstanding over the nature of dialogue. According to Bhambra and Holmwood, I suggest ‘that dialogue and criticism must take place within an agreed framework and, therefore, that some categories are the condition of sociological reason and not open to transformation’ (Bhambra and Holmwood 2026: 156). I presume this refers to my statement that ‘for dialogue to be productive, it must operate within constraints, not least a shared devotion to discovering the truth’ (Hammersley, 2026: 138). A shared devotion is not a category but does clearly involve assumptions, in this case: that discovering the true answer to a factual question is both possible and desirable (even if we can never know for sure that we have been successful). What is ruled out here is complete epistemological scepticism (see Burnyeat and Frede, 1998) or the pursuit of other purposes than inquiry, whether this is ‘winning the argument’ or ‘serving a cause’. Similarly, if the aim of the dialogue is not knowledge but a sound value judgement, then the essential constraint is a belief that coming to reasonable judgements about value issues is possible and desirable; what is ruled out here is the assumption that value judgements are entirely emotive or are simply attempts to exercise power over others. Without these constraints, attempting dialogue would be futile.

Bhambra and Holmwood complain that I have put words into their mouth, they say that the terms ‘racist’, ‘racism’ or ‘colonialist mentality’ ‘are nowhere used by us’ (Bhambra and Holmwood 2026: 154). And they insist that they did not argue that European colonialism is distinctive in being ‘the domination of black peoples by white, and in being founded on white racist ideology’ (Bhambra and Holmwood 2026: 154) or that some of the writers they discuss had a ‘colonialist mentality’. Yet, in representing others’ views we are not restricted to using the same words they do; Bhambra and Holmwood did not adhere to this when discussing the work of early sociologists, nor in their response to my article. I leave readers of their book to decide whether I have misrepresented them on these topics. Personally, I wouldn’t deny the validity of either of the two statements to which they object.

Bhambra and Holmwood’s main concern about my article seems to be that I present them as demanding a radical transformation of the discipline, and that this involves rejecting or cancelling the work of at least some of the sociologists they discuss. They say that what they proposed was a more moderate ‘resituating and renewing’ (Bhambra and Holmwood 2026: 158) of early sociological work. And they claim that ‘some commentators, including Hammersley, have regarded our engagement with the sociological tradition as hostile’ (Bhambra and Holmwood 2026: 160), whereas they insist that their approach was ‘sympathetic and generous’ (Bhambra and Holmwood 2026: 158). However, nowhere do I say that they wish to reject or cancel any of the authors they discuss, nor do I describe their approach as ‘hostile’ or use any equivalent label. They were highly critical, but that is not the same. ¹

Nevertheless, for reasons explained in my article, I do believe that Bhambra and Holmwood misrepresent the views of the early sociologists they discuss in some important respects (and early modern social thought too – see Hammersley, 2025). But I will not revisit that argument here. Instead, I want to focus on three difficult issues about which there seems to be genuine disagreement. These concern:

1. The effects of relying on ideal types and value relevance, in the manner of Weber;

2. The problems involved in interpreting the work of authors in the historical past; and

3. Whether sociology should restrict itself to factual inquiry or must be evaluative.

Unfortunately, in responding to what I said in my article about these issues, Bhambra and Holmwood largely repeat the statements in their book, so I am uncertain about the precise nature of our disagreement. For example, in the article I sought to explain why Weber’s proposal that factual claims operate within a value relevance frame does not mean that a postcolonial or decolonial analysis of modernity must be viewed only as one perspective amongst others, and as no better than a colonialist one (Bhambra and Holmwood, 2021: 138–139; see Hammersley, 2026: 10). However, in their response Bhambra and Holmwood complain that, on the basis of value relevance, ‘there could be no argument that colonialism was central to the development of modern capitalism because any issue of its role depends upon a value-relevant framing which could only be one among others’ (Bhambra and Holmwood 2026: 156–157). But value-relevance determines what questions are addressed, not what answers are produced. Bhambra and Holmwood add that ‘we are asked to accept that there are no requirements of mutual coherence among the constructs developed for different purposes’ (Bhambra and Holmwood 2026: 157), though I didn’t make any such request nor do I see any need for one.

What seems to be at issue here is whether sociological knowledge captures the essential nature of social processes or, instead, only comprises answers to various questions about them. Perhaps Bhambra and Holmwood are adopting the first of these positions, whereas I take the second? So, for me, we can try to answer particular questions about reality, but we cannot represent it ‘in its own terms’, as if it were a person who speaks. This may well be disappointing for an Aristotelian or a Hegelian, indeed it may be claimed that we can have no genuine knowledge about the world if this is true. But this is cutting off one’s nose in spite. In any case, it seems to me that Dewey’s pragmatism (which Bhambra and Holmwood sometimes appeal to) adopts an analogous position to Weber’s neo-Kantianism on this topic: for him, knowledge consists of solutions to particular problems rather than a complete representation of reality in its own terms. While he may have believed that all the solutions fit together to provide a comprehensive picture, I’m not sure on what grounds he could have assumed this. However, we are in deep philosophical water here, and I may not have represented Bhambra and Holmwood’s position accurately.

As regards the second issue – historical versus rational reconstruction (see Hammersley, 2026: 131–133) – in their response Bhambra and Holmwood again largely repeat what they said in their book: that they combined the two, and that each can serve the other. But in neither place do they explain how this is possible, as against the insistence of an author whom they cite that it is not (Jones, 1997: 150, 1999: 2–3). In any case, in my article I showed that this contrast does not capture all the ways in which we may approach the work of past writers, I presented a more adequate typology, and I located Bhambra and Holmwood’s approach within this. I argued that the type of critique they employed was legitimate, though difficult to apply successfully. However, instead of engaging with this in their response, they effectively ignore the problems with which the dispute over historical versus rational reconstruction was concerned.

The third issue concerns Bhambra and Holmwood’s apparent rejection of the ‘fact-value dualism’ (Bhambra and Holmwood 2026: 156). Presumably, what they are referring to is not the distinction between factual and value arguments, since they rely on this themselves in their response, but rather the contrast between non-normative and normative approaches within sociology. Their argument would perhaps be that all sociological work, even that which claims to be ‘value-neutral’, is inevitably normative. However, they provide no justification for this view, and it is one that I have challenged elsewhere (Hammersley, 1995, 2017, 2024).

There are also a few minor points in Bhambra and Holmwood’s response that I will comment on. They claim that, in their book, they demonstrated that classical sociologists did ‘address colonialism within their writings’ but that the secondary literature elides this (Bhambra and Holmwood 2026: 155–156). This is a complex claim since in their book they write that: ‘In none of the writers who make up the usual canon of modern social theory is there a discussion of race as central to the social structures of modernity. We trace this absence to a failure to account for the centrality of colonialism and empire within the modern world’ (Bhambra and Holmwood, 2021: viii). So the point, presumably, is that these early writers discuss colonialism, to varying degrees, but do not treat it as central to the development of modernity. I suggest that, given this, it is not surprising that, in summarizing their views, the secondary literature does not treat what they say about colonialism as centrally relevant.

Contrary to what Bhambra and Holmwood (Bhambra and Holmwood 2026: 160) claim, I did not argue that it was ‘unnecessary’ to add Du Bois to the canon: in the part of my article to which they are refer, I simply noted that, contrary to what they claimed, his work had not been neglected ‘until very recently’ (Bhambra and Holmwood, 2021: 177), that it had been given somewhat more attention than they implied, for example that a volume in the Heritage of Sociology series had been devoted to it (Green and Driver, 1978). Furthermore, I said that ‘I agree with Bhambra and Holmwood, and others, that Du Bois’s work ought to be included in the sociology curriculum’ (Hammersley, 2026: 145), and I meant it.

Finally, I am grateful to Bhambra and Holmwood for pointing out my citation error in relation to Sangiovanni, though I am not sure why they find it ‘puzzling’ (Bhambra and Holmwood 2026: 158). It is a minor error that is very irritating. Fortunately, I don’t believe it will have misled any readers since I provided the page numbers. It is on a par with confusing Louis-Philippe (not Phillipe) with Louis Napoleon (Bhambra and Holmwood, 2021: 86).

I wish I could be as confident as Bhambra and Holmwood about the capacity of dialogue to advance knowledge. I have tried my best here to clarify the differences between us. But I don’t believe we have made much advance in determining what is to be done about the current state of sociology. Nevertheless, I thank them for engaging in dialogue – that is our only hope.