Now that I have established the justification of sceptical arguments, it is possible to move on to the much more interesting business of their implications. Scepticism removes certainty, and without certainty we have, instead, provisionality. Since we can no longer justify holding any beliefs absolutely, we can only justify holding them more lightly.

The idea that beliefs should be held with a degree of conviction that is proportional to the evidence is one that goes back to Hume[1]. The implications of this, however, are more profound than Hume seems to recognise. Philosophical justification by itself does not offer an account of what it would be like to hold a belief provisionally and to a certain degree, because a priori reasoning can only tell us about the contradictory nature of absolute beliefs, not the positive calibration of the non-absolute beliefs we might use to replace them. Instead, the justification of specific non-absolute beliefs depends on a complex interplay of different factors of context, character, language and intention.  A priori reason deals only in invariable generalities, so using it to describe variables is like using an unreliable nuclear weapon to make a minor adjustment to your bicycle gears: the nuclear weapon either destroys everything or leaves it exactly as it was, when what is needed is gradual and subtle change. Only experience itself tells us what provisional beliefs are like, and any description of them is a psychological description, not ultimately a philosophical one.

That psychological description could be based on a theoretical model of the relationships between parts of the psyche that explains how provisionality works. This is what I will attempt in detail in IV.2. There I argue that provisionality is distinguished by offering optionality: that is, at the moment we make a judgement, there are alternative possibilities available to us. This forms a contrast with the repression of alternatives created by absolutisation, in which the supposed complete justification of one belief blocks all alternatives. This optionality also creates antifragility[2] – that is, the ability to benefit rather than suffer from a range of unexpected conditions. To maintain optionality, we also need what Daniel Kahneman calls ‘slow thinking’[3] at the right moments when circumstances make it possible to use it. In this way we avoid the cognitive errors that follow from rigid patterns of ‘fast thinking’ that may be adapted to some situations but are not sufficiently flexible in the wider range we are likely to encounter.

Alternatively, provisionality can be explained using a physiological model in terms of the relationships between the two hemispheres of the brain. Since it is the left hemisphere of the brain that maintains all our linguistic representations of the world and a sense of certainty about those representations[4], provisionality can be seen as the maintaining of sufficient awareness of the perspective of the right side of the brain to limit that left-brain certainty. The right hemisphere provides all contextuality in our awareness[5], and it is contextuality that is required for an awareness of fallibility. It is only if our beliefs exist in a functional vacuum, with other possibilities not being actively considered, that we can start to attach certainty to them.

Sceptical arguments, then, can act as a prompt to remind us that our brains have two hemispheres, and that if we have slipped into the mode of left-hemisphere dominance, this is not even justifiable in its own terms. We may need to take active steps to balance our mental states by taking enough account of the right hemisphere, in order for our judgements to become more objective and justified. There will be much further discussion of such steps later in this book and later in the series.

For the moment, however, I am going to focus not on the positive psychological explanation of provisionality, so much as the negative philosophical explanation of processes that interfere with provisionality. Philosophy cannot give a precise enough description to help us make subtle adjustments in our experience, but rather than working only with the destructiveness of a nuclear weapon, it can also work as an air-raid siren to warn us when a nuclear weapon is on its way, or as a bomb disposal expert to render the nuclear weapon useless. These kinds of operations are crude but vital. They provide us with a starting point, a clear field of security in which more subtle psychological work can begin to take place. In the middle of a nuclear war, we are not likely to be too troubled about the exact state of our bicycle gears.

The nuclear weapons in this analogy are metaphysical beliefs. It is metaphysical beliefs that fall foul of sceptical argument and are incompatible with provisionality. They are also purely the products of our left brain hemispheres, which maintain a self-referent certainty about them that repels any interference by the right hemisphere. The term ‘metaphysics’ is used in a variety of ways, but I am using it here to mean absolute beliefs that can only be dogmatically asserted, rather than provisionally asserted in a way that can be justifiably falsified through experience. As I will argue in more detail in 1.f, these kinds of beliefs function psychologically as apparently invulnerable rallying points for our egoistic identification – but the invulnerability is deceptive. A full survey of different types of metaphysical belief is also made in IV.3 & 4.

The ideas of verification and falsification have been another of the many philosophical battlegrounds in which philosophers have tried to reach absolute beliefs about the world of experience – and have failed to do so[6]. If we could manage absolute verification or absolute falsification through experience, then we could achieve certainty, but, as I have already established, we cannot do that. The distinction between metaphysics and general theory is, most importantly, that metaphysics pretends to that certainty and fails to entertain possible alternatives, whilst general theory is justified to a degree and set beside alternatives.

Metaphysical beliefs can often be identified from their necessary source of justification (i.e. the source without appeal to which they could not be justified). If the source of justification is absolute, such as a revelation from God, an absolute law of nature, or a purely a priori framework (or the direct denial of these things – see 1.e) then their justification cannot be provisional. Even if the belief in question claims to be a scientific theory, if it makes absolute claims this will fall foul of the problem of induction unless the theory is made provisional (and also remains consistently provisional in the further implications it is judged to have).

Even if the belief is claimed to merely provide a priori condition for other widespread beliefs, if it is claimed to be the only possible a priori framework that could be used it becomes metaphysical. So, for example, the observation that we generally rely on an a priori framework of space and time is not metaphysical, but the absolute assertion that all experience must use a framework of space and time is a metaphysical assertion (because without evidence of the absence of other possible frameworks, it must dogmatically appeal to the uniqueness of the one we have).

Another way that metaphysical beliefs can often be identified a priori is through their dualism.  Metaphysical beliefs tend to come in opposed pairs, where the same claim is either affirmed or denied, and dualism is the tendency to be restricted to these opposed pairs without awareness either of further alternatives, or of the equal justification that can be given to the opposing claim. For example, theism and atheism, realism and idealism, determinism and freewill (or determinism and indeterminism, depending on which features of determinism you highlight), and mind and body are each opposed pairs of metaphysical dualisms. If the application of a different explanatory framework makes a denial just as likely or unlikely to be correct as an affirmation, and this point is completely unrecognised by the person making the claim, then we are dealing with a metaphysical belief. For example, a theistic set of beliefs may superficially appear convincing even in the face of evil if we accept that only God knows his own justification for allowing evil events, but if we don’t accept this framework of explanation we could just as easily conclude that God does not exist because his goodness and omnipotence (which are usually seen as essential features of God) are contradicted by him permitting evil. Someone who puts forward this defence for believing or disbelieving in God without recognising the equal plausibility of the alternative is doing metaphysics. Another incidental feature of such dualism is that the third alternative – agnosticism – is not taken seriously, but for more on this see 1.e.

Metaphysical claims can often be readily seen as incompatible with scepticism because of their dogmatic justification and dualistic form, whereas provisional beliefs without these features are just as clearly not threatened by scepticism – not because they are proof against it but because they take it into account. If I claim that I have seen a blackbird in the garden several times, I rely not on a dogmatic source of justification, but only on my own observation on several occasions, and my memory of those occasions. I may be wrong (I may have misidentified the species, or misremembered the previous occasions, for example), but I could freely admit to discovering that I was wrong without it having the further implications that would undermine an absolute source of information. Similarly, I could confidently assert that I saw a blackbird several times without needing to defend this claim against a sceptical threat. However, if alternatively I asserted that the blackbird exists, I would have to assert that it did so regardless of my experience, and in the process repress alternatives. To distinguish that the blackbird really existed as opposed to the fact that I observed it, I would need to dogmatically assert that it could not have been an illusory blackbird, even though this would be just as coherent an account of the event.

It might be objected here that there are perhaps an infinite number of possible explanations of any experience, so that it is impossible to distinguish experiences with multiple possible explanations as metaphysical as opposed to those without such, for this would be a false distinction. It is not the possibility of multiple interpretations that is the point that makes the difference, but our awareness of them. When I thought I saw a blackbird, I could possibly have seen a disguised alien spacecraft, a hologram, or a hallucination. However, a self-consciously limited claim allows for these alternatives, whilst still confidently asserting that, as far as I can tell on the evidence available to me, I saw a blackbird, not a hologram. It might need further discussion to establish explicitly that my claim is intended to be provisional, because it takes into account these other possibilities without refusing to make a claim at all. If I go on to make the metaphysical claim that the blackbird really existed, however, my claim does not allow for these alternatives, but is arrogantly extended so as to actively deny any contesting claim that the blackbird did not exist.

This self-conscious limitation (an aspect of justification that I call agnostic foundationalism) will be discussed further in 1.e., and it is a relatively easy way in which provisional claims can usually be distinguished from metaphysical ones. Of course, we have to investigate closely the issue of meaning here (discussed in volume 3) and take into account the intentions which would enable us to distinguish claims with attached metaphysical assumptions from those without them. We cannot just pounce on every use of the word ‘exists’ and assume it is metaphysical, or conversely assume that every statement apparently just describing or generalising from observation is not. Claims accepted from the testimony of others also may or may not be accepted for dogmatic reasons. If the claim is made philosophically explicit, though, or if other philosophically explicit metaphysical claims are deduced from it, we can become clearer about whether it is metaphysical or not. To use a different analogy, metaphysical claims are like old oak trees in a storm, that are so stiff they cannot bend to the wind. They either stand or they break. Provisional claims, however, are like young oaks, flexible enough to change when the conditions change, and thus always able to stay adequate to those conditions. The difference in flexibility between provisional and metaphysical assertions is determined psychologically by the awareness of the possibility of different conditions, which is only generally rather than precisely reflected by the terminology used.

---

[1] Hume (1975) pp 110-111

[2] A term coined by Taleb (2012)

[3] Kahneman (2011)

[4] McGilchrist (2009) p.70

[5] Ibid. P.80

[6] For example, in Ayer (1946) – the verificationist approach, and Popper (1959) – the falsificationist approach