Search

What is rigour? What is it meant to achieve? Suppose the defining features of a rigorous process to be not-trivial, auditable, rule-based, precise, integrable, and based on a valid argument. Using randomised controlled trials as an example, we argue that it is not enough to satisfy such requirements. What these can establish is very narrow in scope. They do not of themselves make it likely, let alone guarantee, results that are reliable for most of what we want to do with science, for example, predicting the outcome when you put a policy into action. This is a problem with all such methods, not just RCTs: they can, even if successful, merely certify through rigour the truth of very narrow claims. The reliability that science delivers requires more than such certification.

This chapter provides a final caution: a virtuous tangle doesn’t ensure a scientific product is reliable. Using the example of gravitational wave energy we show how two scientific accounts, aimed at explaining the same phenomenon for the same purpose, can both have overwhelmingly virtuous tangles and yet be incompatible. So one must be giving the wrong answer, despite looking like a good account according to our theory of scientific reliability. The tangle, then, is a guide to reliability, rather than a guarantor.

This chapter puts the concept of the tangle to work, showing how it can be used to solve an ongoing debate in political science. Here we show how the theory of the democratic peace, an agreed-upon principle without a broadly agreed-upon explanation, is constrained by the way that its underpinning products interact under the tangle. These constraints, both divergent and convergent, create a single, overdetermined, version of the democratic peace principle—the democratic peace by anybody’s book. This new version of the democratic peace principle can support predictions about a smaller range of cases than most others. But, we claim, these predictions are likely to be more reliable. That’s because the principle has a far richer tangle backing it up.

In this chapter we explain and defend three objections to scientific method. Objection (1): Any characterisation that is broad enough to cover everything that reasonably gets counted as science, from gravitational wave theory to thermometry to Covid-19 vaccine research to measures of democracy to the theory of gender performativity and so much more, will be so general as to lack content. Objection (2): Using the term ‘scientific method’ is dangerous. Privileging methods narrows the scope of what science is up to and of what is needed to evaluate it. Objection (3): Scientific method is too bound up with the attempt to establish scientific knowledge—true claims and explanations. But that leaves out the bulk of what science produces. For that the focus should be on reliability, and that requires a whole new research orientation in methodology and philosophy of science. This is key to our approach in this book.

The Introduction outlines what The Tangle of Science aims to do and describes how the contents of the book serve those aims. It explains that the central question the book addresses is ‘Why can we rely on the products of science to do the jobs we expect of them?’ It identifies ‘scientific method’, ‘rigour’, and ‘objectivity’ as the usual answers, which in Part I of the book we argue are not up to the job. It then introduces the central claim of the book: the reliability of any one product of science rests on a vast, amorphous network of other heterogeneous scientific products. This is ‘the tangle of science’, which is explained in detail in Part II.

This chapter illustrates the concept of a tangle using two episodes from the history of science. In the first episode, we examine how natural philosophes of the eighteenth century ‘fixed’ the boiling point of water. In the second, we shift to the twentieth century to analyse the development of the antibiotic penicillin. We argue that the virtuous features of the tangle—its richness, entanglement, and long tails—help to explain successes in both episodes.

In this chapter, we offer three different arguments to support our claim that a virtuous tangle is a symptom of scientific reliability and is conductive to it. First, standard theories of confirmation presuppose tangles of the kind we advocate. Second, standard post hoc analyses of the failure of scientific endeavours generally find as part of the reason for failure that something from the tangle that should have been there was missing or faulty. Third, the more numerous and well-defined are the other pieces that a scientific creation must work with to achieve a goal, the fewer options there are about what it must be like. Finally, we provide a fuller account of what we mean by a tangle, what should be in a tangle for it to merit the label ‘virtuous’, and why that could make for reliability.

Science is remarkably reliable. It puts people on the moon, performs laser eye surgery, tells us about ancient civilisations and species, and predicts the future of our climate. What underwrites this reliability? This book argues that the standard answers—the scientific method, rigour, and objectivity—are insufficient for the job. Here we propose a new model of science that places its products front and centre. This is the ‘Tangle of Science’. In this book we show how any reliable piece of science is underpinned by a vast, diverse, and thick network of other scientific products. In doing so we bring back into focus areas of science that have been long neglected, emphasising how every product, from the screws that hold the space shuttle together to ways of measuring the consumer price index to Einstein’s theory of general relativity, work together to support results we can trust.

Most current techniques to deal with invasive species are ineffective or have highly damaging side effects. To this end suppression-drives based on clustered regularly inter-spaced short palindromi...