The demarcation problem is a long-standing philosophical issue of how to distinguish (or demarcate) science from non-science. Demarcation dates back to the early Greek philosophers, and has been a central and problematic issue in philosophy of science for the last fifty years or more.
Why is demarcation such a difficult problem? One reason is that science is a heterogeneous, moving target, developing and changing significantly over time. And yet paradoxically, despite diverse views on the nature of science and demarcation criteria, there is broad agreement on most concrete demarcation cases, suggesting that demarcation is achievable.
Often, we also want to know if a field is a pseudoscience – non-science masquerading as science. Science has a privileged position in modern society, and those practising pseudoscience desire these privileges. Of course, they may also genuinely believe they are practising science.
Creation science is one field widely regarded by philosophers of science and scientists as a pseudoscience. Its proponents claim to provide scientific evidence for an Biblical creation account. Intelligent design (ID), the subject of this series of posts, is a recent derivative of creation science that attempts to place itself more firmly within the realm of science.
The importance of demarcation
Obviously, it is important for philosophers of science that they are able to characterise their own subject matter.
However demarcation is more than a philosophical problem, as modern society not only privileges but depends on scientific knowledge in various ways. For example, we need to be able to distinguish effective medical treatments from quack remedies; we have limited public funding for scientific research and would like to allocate funds appropriately; we need to make political decisions on scientific issues such as climate change. Science is also a central subject in our education systems, and decisions must be made as to what topics should be included in science curricula.
Consequently, society needs to make informed judgements about what is and is not science – and in all of the above dependencies, the critical demarcation issue is between science and pseudoscience – the primary focus of this series, applied to ID.
A brief history of demarcation
Aristotle was the first philosopher to attempt to comprehensively describe scientific knowledge, but it was not until the rise of logical positivism in the twentieth century that demarcation became a central concern of philosophy of science. The logical positivists decided a claim was scientific only if it was empirically meaningful, and that required empirical verifiability.
Karl Popper rejected verificationism, instead proposing falsifiability as the sole demarcation criterion. Thomas Kuhn thought that science in its ‘normal’ phase resisted falsification, and was characterised by routine puzzle-solving, while Imre Lakatos and Paul Thagard proposed progressiveness as a demarcation criterion.
In the 1970s and 1980s, a variety of multi-criteria demarcations were proposed. These included being guided by natural law, testability, tentativeness and falsifiability. In 1983, Larry Laudan famously announced the demise of the demarcation problem, labelling it a ‘pseudoproblem’ that is ‘uninteresting and […] intractable’ (Laudan, 1983). Laudan complained that there are no necessary and sufficient conditions adequate for demarcation, as they invariably misclassify some fields. Work on demarcation subsequently stalled until a recent resurgence of interest, particularly with regard to distinguishing science from pseudoscience.
Hansson noted that a pseudoscience will ‘deliberately attempt to create the impression that it is scientific’ (Hansson, 1996). Pigliucci criticised the desire for a set of necessary and sufficient conditions (Pigliucci, 2013), noting that the demarcation problem is not sharply delineated. Both Pigliucci and Dupré (Dupré, 1993, 242) proposed Wittgenstein’s family resemblances as a basis for demarcation. Rather than a common set of compulsory criteria, family resemblance concepts are connected by overlapping similarities that may not apply to every instance. Pigliucci suggested evaluating disciplines based on their varying degrees of theoretical soundness and empirical support, with no sharp demarcation line.
Similarly, Mahner argued for a variable cluster of indicators that characterise science, requiring a certain number to apply for a field to be regarded as scientific (Mahner, 2013). He suggested there may be up to fifty relevant indicators.