Philosophy of science

Australian pictograms

APPLICATION OF PHILOSOPHY OF SCIENCE IN ROCK ART RESEARCH

ROBERT G. BEDNARIK
Anthropos metron hapanton – Man is the measure of everything

 

Introduction

It is impossible to deal effectively with any scholarly pursuit without recourse to epistemology, the branch of philosophy concerned with the origin, nature and limits of human knowledge, and the methods of acquiring it or having acquired it. This is self-evident in relation to any academic topic, but in investigating the origins of the human model of reality, which I claim is the ultimate purpose of palaeoart or rock art studies, it is utterly indispensable.

The human being is an intelligent organism, the product of a long evolutionary process. Its continued existence is contingent upon its possession of several sensory faculties. It is on the basis of these faculties that we as a species map and comprehend physical reality, as if that were their role. This is the greatest misunderstanding in science or in the intellectual comprehension of reality, and in the human instance it is the basis of anthropocentrism (the interpretation of reality purely in terms of human values or experience, which is totally unscientific).

To view anthropocentrism with even a semblance of objectivity, which is by no means easy for us, it is useful to consider the role of human sensory faculties. These were certainly not selected on the basis of being the best possible combination of such abilities for the purpose of determining ‘objective reality’ (for the sake of the argument, let us assume that such a state can exist; I am not suggesting this to be the case). The principal criterion in their evolutionary ‘selection’ was that the sensory faculties of every organism in any planet’s global biotope must relate to the same physical reality as the rest of that planet’s biomass, often even to the same forms of perceptual manifestations of such reality. Evolutionary dynamics would not permit exceptions within such a system, and an organism not relating to it would not survive even if environmental conditions were perfectly suitable for it. The perceptive abilities of any species in the universe are perhaps best described as a compromise between the need to match those dominant in the rest of its particular biotope, and possessing enough variation relative to competitors to have an evolutionary edge over them.

However, the fact that in a particular biotope, any participatory organism from a microbe to a human relates in some fashion to a particular set of variables (e.g. spatial or temporal variables) is no proof that these are the only ones possible, or that they define some finite reality. Yet it is from this that anthropocentrism (in the case of humans) derives its confidence. The obvious explanation for our confidence in equating the reality we experience with ‘objective reality’ is that, provided we continue to experience it only within the cognitive framework we have traditionally used, it is not likely to be challenged. Much of what we call science is actually an exercise of systematically augmenting an anthropocentric framework, through the misapplication of empiricism. Valid empiricism is the principle that human sensory experience is the source of human knowledge, whereas if this view is corrupted to regarding human sensory experience as the sole measure of how things really are in the world it becomes a major falsity. However, much of what is regarded as ‘science’ falls into the latter category, which I shall call ‘low science’ here. ‘High science’ seems largely restricted to aspects of philosophy and theoretical physics.

Perception and cognition

A significant dilemma we are facing in the area of investigating how humans developed their models of reality is that all of the background information we can possibly muster comes from archaeology and palaeoanthropology, academic endeavours that certainly belong to the ‘low end’ of the scientific spectrum. It goes without saying that, in considering the origins of human ontologies itself – surely the most ambitious scientific endeavour possible to us – such a state of affairs is entirely inadequate. An alternative approach is not just preferable here; it is prerequisite.

Let us look at some basic propositions about perception generally. The possibility of perception is attributable to physical processes spreading out from centres and retaining certain characters. Without them it would be impossible for different percipients to perceive the same object or phenomenon from different points of view, and no intelligent organism, humans or ‘little green men’ from outer space alike, would have been able to discover that its individuals existed in a common world. A significant factor in our ‘perceptual confidence’ (by which I mean our confidence that our perceptions are ‘valid’ in the determination of their causations) is the similarity between the perceptions of different organisms in similar situations. Intelligence itself would have been impossible without the discovery of a common reality, hence intelligent reflection would not have occurred.

‘Awareness’, like ‘intentionality’, is a very rubbery concept, and I emphasise that I shall use it here only in the vaguest of meanings. Now, this awareness of a common reality experienced by most, but not all, humans (and the same, one presumes, applies to all intelligent beings in the universe, should any others exist, have existed, or will ever exist) is clearly attributable to perceptions. Perceptions are patterned responses to sensations caused by physical objects or their properties. For instance, an object might reflect light radiation in a particular way, so that certain wavelengths dominate. A visual system sensitive to this selective reflection of light will perceive a sensation we call colour vision, and the organism possessing such a system will infer the physical property of colour. While the perceived object in question no doubt possesses a large number of alternative properties, only very few of which a human may perceive (even with the help of the technological extensions of our sensory abilities, measuring instruments), there are good reasons why natural selection promoted certain sensory faculties in us and not others.

‘Low science’ as I define it here ignores our inability to explore Plato’s or Kant’s noumena (see below), which renders it as distant from ‘high science’ as any other metaphysical system invented to cater for human belief preferences (such as religions). At this point it needs to be emphasised that this differentiation does not, as it might appear to, allude to some elitist program – on the contrary: ‘high science’ involves the most profound humility, because it demands a recognition of human inadequacy that contrasts sharply with the confident brashness of naive empiricist or logical positivist ‘science’. In epistemologically rigorous science, our lack of major access to ‘truths’ is always accepted.

Why can we be so certain of the inadequacy of our cognition in questions of access to reality?

The sensory perceptions of any organism, including one possessing some level of ‘self-reflective’ intelligence as we define it, were presumably acquired through its evolution. They are then a rather haphazard collection of neural abilities in relating to particular physical processes in the world outside our bodies. The brain ‘knows’ enough to carry out its function of fabricating individual reality, and from this we construct consensus reality through social intercourse.

Two crucial points emerge from this: firstly, our knowledge of anything occurring outside our heads must be very precarious. Secondly, the dynamics governing the evolution-determined acquisition of sensory abilities cannot be assumed to be related to some design aiming to equip us with the ability to define ‘objective reality’; there was no survival benefit in such an ability, as I have emphasised on many occasions. Rather, one would suppose that these dynamics resulted from chance variation in the struggle for existence, so they would have been selected for their utility in survival. Survival, of course, is in no way related to objective reality, it merely reflects an ability to respond to the environment.

All our perceptions relate to events, to changes in the physical world; a steady-state reality would not be perceptible to us or to any other being. To perceive an event not taking place in the percipient’s body, there must be a physical process in the world, outside the reach of its hard-wired neural system, that produces a stimulus on the surface of our body (or a receptor such as the retina) that is neurally detectable. It is most reasonable to postulate that this rather tenuous link between our nervous system and the real world provides absolutely no justification for the fond delusion of humans that they have access to some significant reality. Of course they do not and it is salutary to remember that this was known to some Greek scholars millennia ago. In particular, the principle of anthropocentric perception of reality was beautifully captured in Plato’s simile of the cave, prompting some present-day commentators to say that the discipline of philosophy has managed to produce no more than mere footnotes on Plato’s insights. Nevertheless, collectively these ‘footnotes’ are still useful, and we shall consider a few of them here.

 

Some elementary philosophy

Emmanuel Kant, in his seminal Critique of pure reason (1781), developed Plato’s concept of a dichotomy between the knowable and the unknowable, and coined the concepts of a perceptual construct of reality (consisting of phenomena) and an objective reality consisting of noumena. While this distinction remains embedded in contemporary epistemology, there are significant problems with it. Basic to a Kantian model of the world is the assumption that the phenomenal reality is experienced uniformly by all humans, irrespective of their cultural conditioning. In the 18th century, this was certainly the expectation in European thought, which at that time was incapable of perceiving its own ethnocentrism. Even Ludwig Wittgenstein initially reaffirmed its basic validity with the aphorisms of his Tractatus Logico-Philosophicus (1921). But he contributed significantly to questioning the logical positivism developed on Kantian thought when he examined the role of language in concept formation and maintenance. In his early phase, Wittgenstein asserted that thought (and he referred to human thought no doubt) is the logical picture of the facts, which in turn are made up of ‘atomic facts’ (actually Bertrand Russell’s term; the 1922 English translation is a corrupted version of the German text). The thought is the significant proposition, and propositions are truth functions of elementary propositions. The purpose of language is to state the facts, which it does by picturing. Thus language seems to have a structural similarity to what it describes. Ethical or metaphysical statements can only be nonsensical violations of the legitimate application of language, and in this Wittgenstein includes his own utterances on the theory of language. He regarded his own metaphysics as useful or important nonsense, and philosophy as it is traditionally understood as rooted in linguistic confusion. His sentiments were later expressed in a different way by Richard Rorty in his The linguistic turn (1967) when he called for overthrowing the ‘spectatorial account of knowledge’: philosophers had never been able to establish that they were doing anything more than eternalise contingent prejudices.

In Wittgenstein’s later phase (commencing around 1918), in which he contradicts his own Tractatus, he rightly focuses on the influence of language. However, this seems to have had no effect on logical positivism – a point worthy of closer examination. The ‘facts’ positivist reality is made up from correspond to linguistic symbols, or ‘pictures’. These ‘facts’ are phenomenological facts, i.e. they relate to the internal and relativistic construct of the world. To present in (philosophical) language anything which contradicts logic is impossible, because thought is itself meant to be logical. Hence the only mode of constrained mental activity (which the positivist calls ‘thought’) intelligible to the human mind is logical thought as we define it. But since it is entirely couched in tautologies it cannot express anything of significance. So while logical positivism has to accept that nothing at all can be said about reality, it nevertheless pretends that its knowledge about the world, derived entirely from linguistic formulations of empiricist constructs, is valid. This is the greatest self-contradiction in what has come to be called philosophy of science, indicating that this philosophy, and the science based on it, is essentially a farce. The tyranny of empiricism which has become the hallmark of the 20th century scientism (as opposed to science) lacks the integrity of a Wittgenstein, who denounced his own work as a mistake and a self-contradiction.

It is therefore necessary to examine the influence language has on our preferred concept of reality. The communicative units of any language, verbal or otherwise, are of course symbols. Thus a symbol is a mediating tool the mind uses to represent the world. An intelligent species’ knowledge is mediated by symbols which represent abstracted components of species-centric (or, we should preferably say, culture-specific) reality, carved out from perceptions of the objective world in the analytical process of the mind as it builds its image of the world. The lingual structure of anthropocentric reality, including that created by science, is difficult to appreciate by the human mind, precisely because all it can know is predicated on its own symbolism. All conceptual beings seek validation through reference to an external standard of their conceptual standard from others. However, the need of external validation behaves inversely to the number of successful inductions the organism has experienced ontogenetically. Therein lies the reason, and the only reason, for ethnocentrism, and ultimately anthropocentrism.

 

Science and truth

To express this state of understanding more succinctly: humans are incapable of determining what is true. Science is not satisfied with this state of affairs, so it has found a way around this rather large problem. In the 1930s Karl Popper devised falsification as a means of separating science from non-science: propositions must be presented in such a way that they can be disproved by some conceivable spatio-temporally located event exemplifying a possibility which the proposition would exclude. In the second half of the 20th century, this burden of falsification was somewhat modified: we now speak of refutation instead. This is because the falsifying evidence may itself be misinterpreted, and a refuted proposition is not necessarily false. The refuting evidence may be subjected to further testing, and if itself found to be problematic (as is often the case in science) a refuted proposition may be reinstated.

Essentially this system of scientific testing through refutation has become so universally accepted in the sciences that refutability is now considered to be the principal hallmark of a scientific proposition, hypothesis or theory. To illustrate by example: the proposition that Homo erectus was in the Americas is not scientific. No amount of absence of such finds can conclusively prove that he was not present in those continents. However, the proposition that Homo erectus was NOT in the Americas is refutable, it is testable. Hence it is scientific. We may one day discover evidence of the hominid’s presence in the Americas, so the possibility of refutation is always given. The assumption of scientists is that, if a proposition has been tested thoroughly, and if we have failed to refute it, such an idea or model is considered to be strengthened, and continues to be strengthened by every refutation attempt it survives. However, at no stage will the scientist consider it to be ‘true’. He can never know that; that knowledge, or rather that lack of knowledge, is what makes him a scientist. So a scientist is not someone who knows something to be true, but someone who knows nothing to be true. ‘Truth’ can be found in religion, never in science. No real scientist has ever come across one finite, absolute truth.

These simple principles apply throughout science. Or perhaps we should say: areas of human knowledge-claims to which this principle of refutability cannot be applied are not scientific. This does not in any way suggest that they must be invalid, or that we ought to ignore them. They simply do not belong into the realm of science.

 

Science and rock art

We have thus arrived at rock art studies, and we can now consider the role of science in this field of research. Or more specifically: what comprises rock art science? In compliance with the criteria we have considered, rock art science has to consist of the pursuit of trying to learn about rock art by presenting and testing refutable propositions about it, and by retaining those for hypothesis building that seem to survive every refutation attempt. Typical refutable propositions about rock art would be the following examples:

1. A paint residue contains organic components, and it is proposed that they are of vegetable fibres, on the basis of microscopic examination. The proposition can be tested by various alternative methods, e.g. oxidation, combustion, spectroscopy, electron probe microanalysis, and continued failure to refute it would result in increased confidence.
2. A cupule A is claimed to be deeper than a cupule B. There are many methods available to us to measure the respective depths, e.g. slide calliper, photogrammetry, the use of a laser instrument. Hence we can test the proposition.
3. An engraving was made with a metal tool. Replication experiments, microscopic study, detection of metal traces would be some of the techniques one might use in attempts to refute this proposition.

However, the vast majority of propositions about rock art we have heard from rock art enthusiasts are not testable. By far the most common claims concern the objects supposedly depicted in rock art. This is the most primitive form of rock art commentary: the modern observer tells us what the rock art is intended to depict. He or she might even argue with other, equally subjective observers about the meaning of the figures. Animals are confidently identified in this manner, or weapons, utensils, musical instruments, or activities animate figures are engaged in. Obviously these claims are not refutable (unless the rock artists themselves can decide the issue), they depend entirely on the cognition, perception and conditioning (cultural, academic, religious, ideological etc.) of the alien observers: we have no way of invalidating them, other than to use the same criteria, i.e. our own subjective way of detecting iconographically diagnostic aspects of a picture. These claims are not testable, hence they are not scientific. They may well be true, this is not the issue here – science as we have seen is not concerned with truth. The issue is whether or not they can form part of a rock art science. Clearly they cannot, they belong into some other field.

To make matters worse, these idiosyncratic pronouncements about what the rock art depicts are often employed in the formulation of complex and far-reaching theories. For instance, an enthusiast might ‘determine’ the types of objects shown together with anthropomorphs (e.g. perceived artefacts), and from this construct a chronological framework for the art in question. Or he might identify animal species in the paintings, and from them deduce the kind of environment or climate at the time the art was produced. Quite apart from the obvious problem that these deductions are drawn from unscientific (non-refutable) claims, we have every reason to be concerned about them when we consider how wrong they have sometimes been shown to be. For instance, we know of examples where corpora of rock art had been attributed to hunter-forager societies on the basis of the animals or activities supposedly depicted in them, only to discover later that they must have been made by pastoral peoples (e.g. in eastern Spain and in the Sahara). Or style as perceived by the modern observer was used to attribute the art to societies that lived tens of thousands of years ago, only to discover that the art was very much younger (e.g. in the Coa valley of Portugal). The claims about the age of the art, at least, are scientific, because they are refutable. But the evidence they were based on was not, and it is usually more difficult to refute fervently held, subjective claims in this field, than those presented for refutation. While the refutation of scientific claims is a routine activity, the refutation of non-scientific claims often involves theories that are strongly and persistently defended, and this applies particularly in archaeology.

This alone is already a good reason why we should avoid non-scientific propositions about rock art, and yet more than ninety per cent of everything ever published about rock art has no scientific justification whatsoever. It consists largely, and often entirely, of non-refutable claims of alien researchers about graphic systems of people about whose iconographic concepts we know nothing because we have never made a determined effort to learn about them. Instead we, the ignorant outsiders, have told each other what we think about the art!

 

Conclusion

Why should it matter so much whether the interpretations we create about rock art, and about palaeoart generally, are valid? Many rock art connoisseurs seem to enjoy inventing meanings for the art, and why can ‘rock art research’ not be a pursuit based on essentially harmless fantasies?

I perceive three reasons for opposing this argument, and the ideology implicit in it. The first is that rock art enthusiasts are among the most serious threats to the continued survival of the rock art. Site visitation as well as misdirected attention of the general public generated by our activities all contribute to shortening the life expectancy of rock art, much of which is highly susceptible to preservation threats of various types. Rock art enthusiasts have a tendency to want to record the art, often by inappropriate methods, and irrespective of the fact that the same art may have already been recorded by dozens of others. (Most of the recording methods used in these endeavours involve degrees of interpretation, so the records produced often differ somewhat.) All of this detracts from the survival of the rock art in question.

The second reason for opposing the application of unscientific methods in rock art research is more important to me. I believe that palaeoart is the only major resource at the disposal of science that can tell us something about the processes by which hominids formed their constructs of reality. Earlier on in this paper we have seen that these constructs of all humans, including our own, are probably false. If human cultural evolution is an attempt to manipulate the environment, then the most important contribution was to create constructs of reality. Before humans set about altering the physical conditions of their world to create more favourable survival conditions for them, they created a cognitive artefact, a construct of reality. If we are to ever understand our own, present-day construct of reality we must learn how it was developed in the first place. For this we need factual information about palaeoart, not fantasies. And since this is the most difficult task science could ever set itself, it does not seem unreasonable to me to demand the highest possible standards of rigour in rock art science. They must be equal to those we have come to expect of the hard sciences, which are inevitably bound up in principles of refutability. If rock art research cannot at least match these standards then it is not worth doing. It would be a waste of resources and enthusiasm, not to mention the threat to rock art that any research of it involves.

Finally, the third reason for demanding that rock art studies be conducted in a rigorous scientific fashion is that the public has a reasonable expectation that the pronouncements we make about rock art are reliable and credible. Within a non-refutable system of knowledge-claims it is easy to invent interpretations and defend them by recourse to academic influence. Because they are not refutable they cannot be subjected to scientific testing. Yet by pretending that these interpretations are the result of scientific investigation we are not only using false pretences to bolster the credibility of our claims, we are also discrediting science. Our claims to know what was depicted in rock art are entirely unscientific, hence it is hypocritical to present them in the name of science.

Perhaps many rock art connoisseurs have not been aware in the past what science is, and how its simple principles relate to our studies. I have attempted to clarify here what seem to be the main issues. In rock art science, we examine rock art in order to produce data that can lead to making certain propositions about the art: how it was created, when, in what context, and so on. These propositions must be formulated in such a way that they could be refuted; if there is no way to do so, then they are of no interest to science. Another role of the rock art scientist is to test (i.e. attempt to refute) such propositions, and to invent ever better methods of achieving this. Those propositions that consistently withstand refutation attempts can then be used to construct tentative models about more complex interpretations, which in turn can guide data acquisition strategies and research priorities. But it must be emphasised that, as a scientific discipline, we are not sufficiently progressed to permit the formulation of major syntheses in just about any area of our field. Our discipline is only a few years old (having been founded in 1988), still finding its feet as it were, and the amount of data that have been subjected to sustained refutation attempts remains absolutely minuscule. At this stage, the discipline requires of us patience and restraint in speculative interpretation, the posing of bold testable hypotheses, and the collaboration of enthusiastic scientists with a penchant for becoming real pioneers in one of science’s most ambitious endeavours.

February 2003