A Critique of Crude Positivism: Why the Epistemology of Dawkins and Hawking Fails

Introduction

In this essay I wish to address a particular set of opinions that seem to be quite popular among many contemporary atheists, rationalists, and freethinkers. It is not a single specific position, but rather a patchwork of overlapping ideas and perspectives sharing a more-or-less constant core. Being somewhat amorphous, the position of which I am speaking does not really a distinct name. For the purposes of this essay, however, I shall refer to this constellation of views as ‘crude positivism’. ‘Positivism’ is a complex and controversial philosophical perspective, which broadly speaking is characterised by a strong respect for science and empirical enquiry, and an opposition to truth claims based on metaphysical speculation, faith, or authority. My purpose here is not to attack positivism itself, but rather the relatively crude form of it that is popularised, to varying degrees, by figures such as Richard Dawkins, Sam Harris, Peter Boghossian, Neil deGrasse Tyson, Lawrence Krauss, and Stephen Hawking. While one again emphasising that I am describing a family of related and overlapping viewpoints rather than a single well-defined doctrine, three of the key most commonly-encountered components of this ‘crude positivism’ are the following:

  1. Strict evidentialism: the ultimate arbiter of knowledge is evidence, which should determine our beliefs in a fundamental and straightforward way, namely that we believe things if and only if there is sufficient evidence for them.
  2. Narrow scientism: the highest, or perhaps only, legitimate form of objective knowledge is that produced by the natural sciences. The social sciences, along with non-scientific pursuits, either do not produce real knowledge, or only knowledge of a distinctly inferior sort.
  3. Pragmatism: science owes is special status to its unique ability to deliver concrete, practical results – it ‘works’. Philosophy, religion, and other such fields to enquiry do not produce ‘results’ in this same way, and thus have no special status.

My goal in this piece will be to challenge these three claims. In particular, I will argue that the ‘crude positivism’ typified by these three views presents an overly narrow conception of knowledge, and represents an ultimately fragile basis upon which to ground challenges to superstitution, pseudoscience, and other forms of irrationality. My key contention is that we need to move beyond such crude positivism in order to have a stronger intellectual underpinning for the atheistic/rationalist/freethought movements. A final note on style: when I use the phrase ‘crude positivists’ I don’t mean to imply a well-defined group of people. I just use it as shorthand to refer to those who, to varying degrees, hold to one or more of the three positions outlined above.

Strict Evidentialism

Crude positivists insist that all beliefs, or at least all beliefs concerning anything of importance, ought to be based upon appropriate evidence. While I agree with this as an abstract principle, I have concerns about the manner in which crude positivists typically interpret and apply this maxim in practise. The trouble is that, when challenged, nearly everyone will be able to provide some sort of justification for their beliefs, something that they regard to be ‘evidence’. To consider a specific example, the evangelical Christian may claim to know that God works in the lives of believers because they have seen it happen with their own eyes, and experienced it personally in their own lives. Needless to say, this is not the sort of ‘evidence’ that adherents of crude positivism are likely to accept as legitimate. The question, however, is why not? After all, the justification in question is empirically based, in that it is derived from making observations about the world. Generally positivists respond that such experiences are uncontrolled and anecdotal, and thus cannot be trusted to provide reliable evidence. To this, however, the Christian may simply agree, arguing that while such experiences are anecdotal and thus do not qualify as scientific evidence, nevertheless they do constitute evidence of the relevant sort for the domain in question, namely the domain relating to knowledge and experience of God. According to this perspective, only certain particular phenomena or aspects of reality are susceptible to the investigative methods of the empirical sciences, and the nature of God and mankind’s relationship to him would not be one of these areas that science can study. These phenomena can be empirically studied, but this is done by applying different standards than those used for scientific inquiry, using methods that are much more personal and experiential. Scientific methods are applicable in the scientific domain, while other methods and other forms of empirical evidence are applicable in other domains. I am not attempting to defend this ‘separate domains’ position. Instead, I am arguing that it is not sufficient to respond to a position like this by simply asserting that beliefs should be based on evidence, since that is not the point under dispute. That is, the question is not whether some form of ‘evidence’ is important, but the type of evidence is deemed acceptable, and how that evidence justified claim being made.

A related problem concerns the issue of how evidence should be interpreted. Crude positivists often speak as if evidence is self-interpreting, such that a given piece of evidence simply and unambiguously picks out one singular state of affairs over all other possibilities. In practise, however, this is almost never the case, as evidence nearly always requires an elaborate network of background knowledge and pre-existing theory in order to interpret. For example, in order to understand a historical text, one requires not only knowledge of the language in which it is written, but also a broad understanding of the relevant social and political context in which the text was written. Likewise the raw output of most scientific observation or experiments are unintelligible without use of detailed background theories and methodological assumptions.

Given the important role that background assumptions and perspectives shape our interpretations of a given piece of evidence, it is very common for different people coming from different perspectives to conclude that the same evidence supports wildly different conclusions. For instance, many young earth creationists interpret the fossil and other evidence in light of their pre-existing belief that the bible is the literal and infallible word of God, and as a result they conclude that the extant evidence points to a divine creation event in the recent past, devising various ingenious methods of reconciling their beliefs with the apparent evidence to the contrary. My intent is not to defend creationists, but to illustrate that it is not enough to simply say that creationists ignore the evidence. These creationists are responding to the evidence (indeed they argue that it supports their position), but are interpreting it differently on the basis of different suppositions and approaches. We cannot simply dismiss them as being blinded by their presuppositions, since (as I have just argued) evidence can never be interpreted in a vacuum, free of assumptions or preconceptions, but can only ever be interpreted in the context of an existing methodological framework and various background assumptions. To say this isn’t to endorse some form of epistemic relativism, but simply to point out that if we want to explain why creationists and others like them are mistaken, we have to move beyond the crude positivistic cry of ‘seek the evidence’, and articulate a more detailed set of criteria and epistemological principles upon which certain initial assumptions and modes of interpretation are to be preferred over others. We need to do a better job of explaining what types of evidence are most reliable, how to interpret evidence, and why these approaches are more conducive to the formation of true beliefs than other, competing approaches.

Narrow Scientism

The second aspect of ‘crude positivism’ that I want to discuss is the view I have termed ‘narrow scientism’, which refers to the tendency to dismiss, or significantly downplay, the importance and status of all disciplines outside the natural sciences. Physics, chemistry, biology, and geology produce reliable knowledge, while psychology is a bit of a question mark, and economics and political science are clearly ‘not sciences’, but belong with disciples like philosophy and much of the humanities, the domain of fuzzy opinion and not verifiable fact. This, at least, is the typical perception among my advocates of crude positivism. In my view, however, this disciplinary classification is arbitrary, and fails to demarcate any epistemologically relevant distinction. In particular, what is the justification for the view that the only ‘real sciences’ are only the natural sciences? It cannot be the result of having adopted a superior set of methodologies, since in many cases there is more methodological continuity across different disciplines than within single ones. For example, analytical chemistry and cognitive psychology are both largely focused on laboratory experiments, while in astrophysics and macroeconomics experiments are mostly impossible, and so these disciplines instead rely predominantly upon observation and development of mathematical theories. Likewise, piecing together the evolutionary relationships of different species has more in common with the linguistic analysis of different languages than it does with other subfields of biology. Nor can it be the subject matter of the disciplines which sets them apart, since there is a continuum between the study of primate behaviour in biology and the study of human behaviour in the social sciences, and also between the study of natural history in geology and biology, and the study of human history in the social sciences and humanities. Furthermore, many mathematical models originally developed in the context of physics and chemistry have also been profitably applied to many other fields, especially economics and sociology (e.g. equilibrium theory, network analysis, complex systems theory). My contention here is not that there is literally no difference between the natural sciences and social science or non-scientific disciplines. I do, however, think that there is a great deal of continuity and intermingling between them, both in terms of methodologies and subject matter, a fact which belies the sharp science/non-science dichotomy advocated by crude positivists.

This is not, however, merely a question of whether disciplinary boundaries are sharp or fuzzy. The real point I am trying to make is that crude positivists simply have no justification for elevating the natural sciences (whether their boundaries are fuzzy or not) on a pedestal above all other disciples. That is, I do not think the natural sciences are epistemically privileged in the way that crude positivists claim that they are. After all, what is so special about the natural sciences relative to, say, economics, history, or even blatant pseudosciences like astrology? The most straightforward answer, and I think the one crude positivists have mostly in mind, is that the natural sciences apply a rigorous scientific method not found in any of these other disciplines, and this method is more conducive to finding truth than other competing methods. My response to this is threefold. Firstly, I note that this is not a claim that finds a home in any of the natural sciences (i.e. it is not a scientific claim), but seems to appeal to philosophical criteria that lie outside of science. I do not think there is anything wrong with that, except for the fact that it seems to sit at odds with the crude positivistic view that only science is to be trusted. Secondly, as I have argued above, it is simply not true that the natural sciences systematically apply different methodologies to those used in other disciples. Within any disciple the quality of work varies dramatically, some being much more careful and rigorous than others, and this applies just as much to the natural sciences as to other disciplines. Thirdly, and most importantly, if the superior status of the natural sciences is based on their superior adherence to a particular set of epistemological principles, then it is those principles themselves that are the true bearers of the superior status, not the physical sciences themselves. Applying these same principles to any disciple should yield knowledge justified to similarly rigorous standards. If this is correct, and what is at the bottom of the success of the physical sciences is adherence to a particular methodology or methods of inference, then it is those methods that we should focus on championing, whatever discipline they may be applied in.

It has been argued that the subject matter of the social sciences and other such disciplines is inherently ‘messier’ and more complex than the comparatively simpler physical systems studied by the natural sciences. However even if this is true, application of appropriate methodologies should still result in reliable knowledge – the only difference will be that the knowledge will be less precise and known with less confidence, since our understanding of the system in question is less complete and less detailed. This will not, however, result in a qualitatively distinct and far inferior form of knowledge, contrary to the claims of the crude positivists. Some argue that the subject matter of history and social science is such that it is not suited to study by the rigorous methods of natural science. If this were true, it would seem to leave us with two options: either no reliable knowledge about such things is possible in principle (i.e. we can say little or nothing about human history, how societies and economies work, etc), or the reliable methods of attaining knowledge in such disciples are distinctly different and at odds with those used in the natural sciences.

The former possibility strikes me as deeply implausible – why should we not at least be able to know a great deal about such topics through careful investigation, and furthermore how could we possibly know if this were the case given that we could not study these topics? The latter option seems equally unpalatable, for it is essentially identical to the argument by which the evangelical Christian claims that their supernatural claims are outside the bounds of scientific investigation. Indeed, if it is the case that the appropriate methods for studying any subject outside of the natural sciences are fundamentally different to and at odds with scientific methods, then any ground for objecting to irrational or unscientific claims is lost. Religious claims (“the divine cannot be studied scientifically”), alternative medicine (“human health is too holistic to be subjected to scientific methods”), or the paranormal (“the spirits don’t respond under controlled conditions”), it can always be argued that the subject matter lies outside of the natural sciences, and hence different, non-scientific investigative methods are applicable. In my view, this absurd outcome shows that, if we grant superior respect and status to the claims of the natural sciences, it must be because (when conducted properly) the natural sciences utilise justified and reliable general epistemological processes, processes which should similarly be conducive to knowledge acquisition when applied to other subjects. Crude positivists who instead reject any application of scientific methods outside of the natural sciences cannot then simultaneously berate those making religious, paranormal, and supernatural claims for failing to use scientific standards and methods, since by their own admission such methods are only applicable to certain subjects. Narrow scientism, then, is at odds with the core principle of basing all important beliefs upon reliable evidence.

Pragmatism

The third and final aspect of ‘crude positivism’ that I wanted to discuss in this piece is pragmatism, the appeal to the past successes of science as the primary and overriding justification for its epistemically superior status. Science, so the argument goes, simply ‘works’: it puts men on the moon, builds aircraft that fly, and makes transgenic fish that glow in the dark. Ways of knowing that rely on appeals to authority, esoteric knowledge, or personal experience, are inferior precisely because they do not ‘work’ in this way. While I do think this sort of argument has some validity, I think the crude positivist goes too far in advocating practical utility as the defining feature of knowledge. One simple problem with this approach is that many people think that prayer, mystical experiences, etc, ‘work’ in a very real way – they pray to Jesus, and they feel God’s love pouring out over them. The crude positivist, of course, is unlikely to admit that as being a valid example of ‘working’, however all this shows is that science comes out best when judged by its own criteria of what it counts as legitimate ‘success’, while the types of ‘success’ (e.g. drawing closer to god, becoming one with nature, etc) defined by other ways of knowing are simply disregarded.

Beyond this issue of defining criteria for success, there is a deeper philosophical issue concerning the relationship between the ‘success’ of a theory, and the ‘truth’ of that theory. Most of the examples of science ‘delivering results’ are, properly understood, really applications of engineering, not science itself. Of course, engineers utilise scientific findings and theories, but there is nevertheless an important distinction between the development of theory and its practical application. This is important because some schools of thought in philosophy, especially the sort of instrumentalist, pragmatic viewpoints that crude positivists are most closely aligned with, argue that the ability of a theory to deliver successful applications is insufficient to validate the accuracy of that theory in describing the way the world truly is. One example is that of Ptolemaic astronomy: it was capable of generating accurate predictions of the positions of the planets despite the fact that its underlying model for reality (an Earth-centred cosmos with the planets orbiting about crystalline spheres) is completely wrong. To take a more recent example, scientists and engineers still routinely use chemical and physical models which treat atoms as solid spheres interacting in accordance with the laws of classical mechanics. As a description of reality, this is entirely incorrect – atoms are mostly empty space, and what is not empty space consists of protons, neutrons, and electrons, which according to our best theories behave (very loosely) like smeared-out probability wavepackets, evolving in accordance with the laws of quantum (not classical) mechanics. Notwithstanding this completely inaccurate description of the underlying reality, however, the ‘billiard balls’ approach is still very useful and ‘delivers results’ in a wide range of applications. Such examples are one of the major arguments used by those philosophers who adhere to a position known as scientific anti-realism, which is the view that while science produces very useful predictive models, it does not necessarily describe the way things ‘truly are’. Thus, according to this view, science is not in the business of finding ‘truth’ per se, but merely of producing theories that are ‘empirically adequate’ and useful for prediction and practical application.

My point here is not to argue that anti-realism is correct, or that science doesn’t describe reality. Rather my argument is that either way, these considerations pose a problem for the simple pragmatism of crude positivists. If, on the one hand, scientific anti-realism is false, and scientific theories do truly describe the way the world is, then the extreme focus on scientific theories being special because they ‘work’ becomes difficult to justify, since under this view science is special not predominantly because it ‘works’, but because it yields true descriptions of reality. The simplistic pragmatism defence thus simply cannot work, and the fact that other disciplines (e.g. philosophy or theology) may not ‘deliver results’ does not mean that they cannot accurately describe reality. On the other hand, if scientific anti-realism is true, and scientific theories don’t necessary say much about the way reality truly is, then the crude positivist has no basis for critiquing non-scientific ways of knowing for not making predictions or ‘delivering results’. This is because these other ways of knowing (e.g. faith based) don’t necessarily claim to be able to provide predictive models, but claim to describe parts of reality as they truly are. If science and faith/intuition/etc are not even trying to do the same thing, the one attempting to generate useful models, the other not caring about predictive accuracy but about providing true descriptions of reality, then it is unclear how the crude positivist can even compare the two in the way they seem to want to. This approach also seems hard to reconcile with the fact that many adherents of crude positivism do very clearly make truth claims about subjects like religion and the paranormal. If this form of pragmatism is correct, then science and non-science aren’t incompatible, but rather are incomparable, for they are not even trying to do the same thing.

Conclusion

Some people will doubtless read this piece as an attack upon the value of science, or a defence of pseudoscientific, faith-based or emotion-based methods of reasoning. As I have said throughout this piece, however, this is not my intention at all. My goal is in fact to equip skeptics and rationalists to deliver a robust, cogent defence of the value of science and critical thinking in learning about the world, and the superiority of such methods over various rivals. What concerns me is that the constellation of views that I here describe under the label ‘crude positivism’ is quite popular among many rationalists and skeptics. As I have argued, however, I think these views are philosophically naive and very hard to rigorously defend. Worse, some of the more intelligent defenders of non-scientific practices, including religious apologists, practitioners of alternative medicine, and defenders of various pseudosciences, are aware of the problems with such views, and will vigorously critique rationalists who espouse them. I think we can answer their objections, but to do so requires a greater familiarity with philosophy and relevant methodlogical issues than many rationalists and skeptics have, especially when they so often dismiss these fields as irrelevant. In order to advance the cause of science and rationality, therefore, we need to abandon ‘crude positivism’, and replace it with a more sophisticated, thoughtful, and philosophically rigorous account of science and rationality.

Advertisements