Skip to content

Response to Review of Science in the Twentieth Century and Beyond

I’d like to thank Graeme Gooday for his review of my book. In general, it is generous, detailed and sound. (It’s not so kind – and unfairly so – about another author’s work, John Pickstone’s Ways of Knowing, but I’ll come back to that at the end.)

Certainly being compared to Kuhn is humbling indeed. Gooday is right to remind us that Thomas Kuhn’s Structure of Scientific Revolutions, half a century old, is still our most influential picture of the historical process of science, and one which contrasts sharply – at least on the surface – with my account. While Kuhn portrayed science as a social process, most successfully in the depiction of ‘normal science’, in which scientists work within a paradigm, these paradigms were also insulated from a wider world. What I argue is that while science has sought institutional forms which seek to preserve autonomy, in fact science is nevertheless animated by its engagement with the problems that the wider world generates.

This engagement is captured in the main conceptual innovation of Science in the Twentieth Century and Beyond: the idea of ‘working worlds’, what I define as ‘arenas of human projects that generate problems’. Crucially these problems cannot be solved directly, and here is where science comes in (and is why science is of central importance in the modern world). Once problems have been articulated – not a trivial achievement, and one with a social history of its own – then science builds simplified, abstracted ‘representative’ models, of microworlds, if you like), which can be subjected to all the well-developed, sophisticated techniques so well traced by historians of science: manipulation and measurement, comparison and reasoning, theory-building and experiment. Some of this work goes on in institutions and disciplines that are so tightly self-referential that the working world prompts may be forgotten by those who do it. Nevertheless, from such manipulation come candidate solutions to working world problems. Again this process of ‘solutionisation’ (my apologies for the ugly word) is an achievement with its own history.

So, for example, take the working world of civil administration. One problem administrators have encountered is simply knowing the population to be administered – its strata, movement and make-up. This knowledge problem cannot be solved directly. So a simplified representative is built: the census. The census is complex, but the set of data is much more tractable, and can be subjected to sophisticated techniques of inquiry (such as statistical reasoning) in ways that the raw population cannot.

My big point is that there are several dominant working worlds within which much of 20th-century science makes sense. They are: the construction and maintenance of technological systems (especially, for the century in question, military ones), the maintenance of life (especially the human body) and, as in the case of the census, civil administration.

Gooday praises the scope, substance and open-endedness of this working worlds interpretation of science in the 20th century. If, as the reviewer suggests, ‘the provocations and gaps in this volume will inspire research and teaching in the subject’, then I will be more than happy.

However, I think Gooday is too kind where perhaps a more stringent critical approach would be deserved. What would happen, I wonder, if we took Karl Popper as our reference rather than Kuhn? We ought to subject our most cherished theories to the most sustained and honest scrutiny.

What would a savage but fair review of Science in the Twentieth Century and Beyond look like?

Let me start such a review here. Let’s take Gooday’s nuanced criticisms, add some more, and see what remains. I do this to the working worlds thesis not because it is a bad idea or generates few insights but precisely because it is promising. If the working worlds framework ends up being rejected and replaced by something better then we will have progressed.

First, just how many working worlds are there? Science in the Twentieth Century and Beyond mentions both the working world of the military but also the working world of the Cold War. Should the latter be considered a subset of the former? And does it matter that working worlds overlap? For example, military medical science is a response to both the working world of the military and of the maintenance and repair of the human body. The short answers, I think, are that working worlds can have smaller working worlds within them, and that overlap is unproblematic, since two working worlds can prompt the articulation of different problems to which the sciences respond.

Second, critics can offer examples of science that seems to be done in an insulated fashion away from working world problems. Gooday offers the example of Robert Oppenheimer and colleagues’ theory of black holes. How can a black hole in anyway be part of an answer to a working world problem? I have several answers to this point. The first is a reminder that, as I argue in the book, some of the best scholarship in the history of modern theoretical physics demonstrates, when read sympathetically but without distortion, that achievements as towering as Planck’s quantum physics and Einstein’s special theory of relativity are indeed responses to working world issues. The internal, disciplined, carefully cultivated (but only apparent) autonomy of the sciences might obscure these links, but they are there. Perhaps the same can be shown for Oppenheimer’s black holes. Furthermore, I am entirely comfortable with the picture of science that retains much of its detailed apparently internal development. The starting point for new physics remains, mostly, older physics.

I would add that this specific point of criticism only seems to have force if we take science to be mostly the great intellectual achievements – the bright shining lights – rather than the more commonplace, more ‘typical’ science of the kind usually called ‘applied’. Take a scientist at random from the 20th century and they are more likely to be responding to the problems of working worlds in a direct and uncontroversial fashion. Working worlds analysis can stretch to help us understand Einstein, but it certainly describes Einstein’s dull cousin.

And let us not underestimate the power of the value system within science’s institutions that seek to preserve and defend its image of autonomy and its distance from the problems of the world. I am reminded of one, local UCL response to the book. ‘Rubbish!’, said the critic, ‘none of my science has ever been like that’. He is a chemist. Need I say more? It would be an achievement to write a history of chemistry that treated its subject only as an insular commentary on and response to previous, academic chemical ideas. Yet it is undeniably the case that some chemists perceive their science in such terms. What needs to be explained, by the historian, is how and why such perceptions can be sustained.

Third, what are the policy implications of the working worlds approach? I think there are several. Gooday says, approvingly, that I stick ‘my futurological neck out’ when I speculate about what working worlds might be relevant to 21st-century science. My sense is that working worlds might well make a useful tool for horizon scanning, since new working worlds might be identified, or changed working worlds might be projected, and their problems anticipated, all with a view to suggesting how future science might respond. The framework is also relevant in other ways to thinking through how contemporary science policies can be improved. Attention might be focussed on each of the steps in the process: on identifying the circumstances in which problems can be articulated earlier, more clearly and by the right range of stakeholders. What about the cases where problems are articulated but no solutions have worked? For example, why is it the case – and this might be called ’the Steven Packer objection‘, in honour of my father-in-law who raised it with me – that the problems of third world poverty have been visible for decades yet remain stubbornly unresolved? Does the failure lie in the process of problem articulation, or in model building, or in understanding what the conclusions of modelling imply? Is there perhaps guidance that can be offered about how to design representatives of working worlds better? In which case we must also ask: better for whom? Moving back again, who can or should interpret the conclusions of manipulating representatives so that solutions travel more effectively? There are new tools for talking about policy here – let’s sharpen and use them.

But a note of caution should be sounded. The working worlds framework ties science to the influence it can have on wider problems. I think the historical evidence supports the thesis that this model helps us understand the roles and actions of scientists in the 20th century. But I also worry that I am reading contemporary values into history, specifically the predominant anxiety that animates contemporary science policy: that science should be ‘translated’ – moved from laboratory to wider world – so that it contributes effectively to desired outcomes, usually improved economic performance and health. This suggests the fourth line of criticism: am I merely taking recent values and writing them in to the past? Is the working worlds model a Whiggish rewriting of history so that the forces shaping older science are aligned to those shaping the present? Perhaps. But, to play devil’s advocate for a moment, Whiggism has its advantages as well as its well-known dangers: it gives one something to look for in the past. The past is otherwise inchoate. So long as we are explicit and critical about what our presentist values might be they can be used with caution. Alternatively, I just might be right.

A fifth, connected point concerns time. Historians of science should ask what are the reasonable limits of working worlds as a framework for analysing past, present and future science. How far back might it apply? Does it make any sense, for example, to talk of a working world for alchemy? If so, what is it? Is the working worlds framework an analytical approach which has purchase for the modern sciences only? Could a stronger claim even be defended – that working worlds as a frame for the sciences demarcates something distinctive about modern science?

There might be a problem with very recent science too. Gooday suggests that the working worlds model begins to falter when used to analyse changes in the sciences of the 1990s and after. My ’commitment to the working worlds thesis’, he writes, ’is at the expense of understanding how the new sciences … have become more socially accountable in the post-Cold War era‘. What he is referring to is the movement towards public engagement that has sought to build a dialogue between science’s decision-makers and its stakeholders and wider publics. My view is that this engagement is simply a specific and recently expanded technique of articulation of working world problems. It is not the only method of articulation by any means, but it is one that has received a degree of political support, at least in the United Kingdom, and perhaps only for the duration of New Labour, and even then perhaps not entirely wholeheartedly. Nevertheless, I agree that the ways that publics for science can be theorised within the working worlds model is, as Gooday says, ‘a fertile theme for critical engagement in future scholarship’.

Sixth, there’s the problem of over-extension. Let’s call this the ’Paskins objection‘, after my PhD student Mat Paskins who made this criticism. Are there practices which do all the things that sciences do in the working worlds model – notice articulated problems, build representatives, manipulate those representatives, and offer interpretations that can be translated as solutions to problems – which we would, nevertheless, feel uneasy about describing as ’science‘? There are two ways of thinking this through. The first applies only if I want to defend working worlds as somehow definitional of the character of (perhaps just modern) science. In this scenario only sciences relate to working worlds in the way outlined, and, vice versa, the only things which relate in such a way are sciences. It is an ’if and only if’ relationship. Second, only the first of these relations holds. There may then be other entities – not sciences – which also respond to problems, build representatives, manipulate representatives, and offer solutions. Either is possible. Neither might be true.

Finally, ‘working worlds’ has been used before as a term of analysis. Gooday notes that it can be found in sociology and healthcare. This is correct. A quick and dirty google ngram survey also shows that the term had peaks of appearances in books in the 1930s, 1950s and 1980s. However I don’t know what these usages were, and any influence must have been unconscious, if present at all.

Nevertheless, it is a reminder that there is a history of neighbouring ideas to be explored. In this context, I would certainly reject Gooday’s passing comment about the usefulness of Pickstone’s ‘ways of knowing‘ historiography. Pickstone has offered an invaluable set of terms that describe both ways of knowing (natural history, analysis, experiment) – and, crucially, ‘ways of working’ (craft, rationalised production, and systems of innovation) that map on to them. I don’t think these are ‘sterile and static taxonomies‘, and, on measured reflection, I would hope that Gooday would agree.