The common denominator of Pyrrhonian scepticism and Wittgenstein’s anti-sceptical strategy (which will occupy us throughout the remainder of this study) is the connection between contextualism and naturalism. At this stage, it is worth giving a provisional, anticipatory sketch of these two positions.
(1) The contextualism common to both Pyrrhonian scepticism and Wittgenstein is based on the idea that we can motivate and formulate versions of Cartesian scepticism only if we ignore the context-sensitivity of knowledge ascriptions or, more generally, of cognitive projects that process information with objective purport. Cartesian scepticism seems to presuppose the possibility of the absolute certainty of our beliefs: if we know anything at all, we know it with absolute certainty. Yet the absolute certainty of my belief that, say, there is a table in front of me disappears in at least one context: namely, the supposedly context-free context of Cartesian scepticism’s motivation. The impression of a paradox derives from how the premises of Cartesian scepticism are valid in a certain context but lose their validity as soon as we find ourselves beyond its scope. It follows that our beliefs forfeit their absolute certainty as soon as they are placed in the supposedly context-free context of epistemology, revealing Cartesian scepticism as an epiphenomenon of the epistemological attitude. So, if we can show that that attitude is ungrounded or even unreasonable by its own lights, we will be in a position to liberate ourselves from Cartesian scepticism without having to dissolve its paradoxes.
Say someone is sitting on the New York subway. They know they are travelling on a line that runs to Columbia University, and they also know to state as much if challenged. Yet say this passenger is on their way to a Descartes seminar and is discussing Cartesian scepticism with a colleague: they suddenly lose their entitlement to their knowledge claim, because, under the conditions determined by this epistemological context, they do not even know whether the subway system exists. If knowing something always presupposes the ability to exclude anything that would make that knowledge impossible – specifically: that one can do so as soon as one becomes aware of the threat – they do not know whether they really know what they think they know. Now, it would clearly be absurd to take this as a basis for seriously denying that somebody knows something, and this suggests that something or other is awry with the epistemological context. Having an adequate entitlement to the belief that one is travelling on the subway line that leads to Columbia University in no way implies that one retains one’s entitlement in all contexts or given any systematic background whatsoever. If we want to describe knowledge, we have to be on our guard that we do not depart from the context in which a given knowledge ascription is valid. Otherwise, we will be guilty of the classic error of misidentifying object domains (μετάβασις εἰς ἄλλο γένος).
Contextualism is a variant of the age-old idea of relativism: it claims that some particular warrant is not valid in all contexts, because a ‘warrant’ is only ever a warrant relative to or within a given context. Searching for an absolute warrant is therefore a senseless enterprise. In Sextus’ words: ‘since everything is relative, we shall suspend judgement as to what things are independently [ἀπολύτως – i.e. literally, ab-solute] and in their nature [ὡς πρὸς τὴν φύσιν]’.5 Both Sextus and Wittgenstein exploit a variant of contextualism to refute the form of epistemological foundationalism according to which there is a range of absolutely certain beliefs to which all non-foundational and derived beliefs owe their certainty. And they thereby oppose the assumption of an epistemic asymmetry of mind and world. In light of the apologetic dimension of the concept of knowledge (see above, p. 95), we cannot maintain that there is a class of absolutely warranted beliefs; the putative class would not constitute knowledge if we could not defend its members under critical questioning, and items of knowledge that have to be defended are not absolutely warranted. Rather, since they are relative to a context, they possess the validity they do only against the background of that context. All justification makes use of operating conditions that cannot themselves all be justified in the process of justification. Every procedure for making the warrant of a process of justification explicit will draw on some conceptual resources or other that resist being made explicit through that particular procedure. In this regard, knowledge acquisition remains incomplete, which is why there is always a way in for Pyrrhonian scepticism.
(2) Naturalism is the (anti-)sceptical strategy deployed most famously by Hume in his Treatise of Human Nature.6 This strategy does not argue that Cartesian scepticism is untenable in virtue of some logical or dialectical error. Rather, Hume accepts that scepticism represents the only rational system. Yet, so it would seem, no one imports Cartesian scepticism into their everyday life – indeed, it would be impossible to do so, and, as Descartes already noted, it is this impossibility that distinguishes Cartesian scepticism from madness.7
Because of certain ‘very general facts of nature’,8 human beings are evidently so constituted that they develop a steadfast reliance on a certain range of assumptions, even though we cannot obtain any rational justification for them. Finite epistemic beings such as we are can formulate sceptical hypotheses and demonstrate how there can be no absolute justification for our fundamental beliefs (such as the uniformity of nature or the existence of an external world). At the same time, however, life compels us to decide against the force of our arguments and to hold onto more beliefs that we are really entitled to from the perspective of a strictly theoretical attitude. Some entitlement is epistemically blind, including the kind of warrant we acquire in epistemological reflection. Engaging in epistemology is just more human activity and does not transcend the limits of the everyday: there is no extraordinary, god-like point of view onto our epistemic practices. Epistemology cannot reboot our entire system of knowledge claims so as to implement an idealised operating system that is somehow immune to the Pyrrhonian halting problem. And naturalism claims that this does no harm to local, finite knowers once they realise that they are entitled to isolate their local claims from the misguided assumption that they ought to be better grounded than they ever possibly could be.
We can find naturalism in this sense both in Wittgenstein and, more prominently, in Sextus. Wittgenstein’s naturalism revolves around the ideas of a ‘natural history’, ‘life’ or ‘forms of life’. Sextus also speaks of ‘life’ (βίος) and of the ‘guidance of nature’ (ὑφήγησις φύσεως). One of the most famous naturalistic passages in Sextus deserves to be quoted in full on account of its programmatic clarity:
Thus, attending to what is apparent, we live in accordance with everyday observances, without holding opinion [ἀδοξάσως] – for we are not able to be utterly inactive. These everyday observances seem to be fourfold, and to consist in guidance by nature [ἐν ὑφήγησει φύσεως], necessitation by feelings [ἐν ἀνάγκῃ παθῶν], handing down of laws and customs [ἐν παραδόσει νόμων τε και ἐθῶν], and teaching kinds of expertise [ἐν διδακαλίᾳ τεχνῶν]. By nature’s guidance we are naturally capable of perceiving and thinking. By the necessitation of feelings, hunger conducts us to food and thirst to drink. By the handing down of laws and customs, we accept, from an everyday point of view, that piety is good and impiety bad. By teaching kinds of expertise, we are not inactive in those which we accept. All this we say without holding any opinions [ἀδοξάσως].9
Both contextualism and naturalism are results of a series of sceptical arguments, which receive particularly clear formulations in Sextus’ work. Strawson has underlined an important parallel between Humean and Wittgensteinian naturalism, while both Kripke and Fogelin have shown the extent to which we can read Wittgenstein’s reflections on rule-following as providing a sceptical solution to a sceptical argument.