The focal point of Hume’s legacy is his critical reassessment of causality. It is well-known that he vigorously opposed causality construed as the necessary linkage between events (Hume 1992; Sect. VII).² In his opinion, what we perceive as relations between causes and effects boils down to our mere habit. Since we observe regularities in the functioning of nature, we tend to treat them—with no sufficient reason—as if they were necessary. Yet in truth there is no objective link between them and only our remembrance of past event sequences leads us to consider them intertwined. This argumentation, recited incessantly by new generations of philosophy students, is surprisingly more inspiring than its reverent status would suggest. Under close scrutiny it reveals its further consequences. According to Hume, the idea of causality stems from our memory: we remember past accidents and infer from them how causal links will work in the future (Hume 1739, pp. 40–93). If so, the only truly grounded empirical data must come from the present. As a consequence, even the most reliably observed and logged experience that already belongs to the past cannot be fully trusted.³ Even the most inevitable experience that will happen in the future cannot be relied upon, since it does not happen here and now. In a nutshell, for Hume all the past is nothing but our recollection; all the future—nothing but our anticipation. Naturally, it does not mean that thinking of past and future events should be discarded altogether; yet they should be treated with scepticism grounded in the recognition of the primary role of the present.

The above formulated interpretation of Hume’s sceptical paradox, based on a literal reading, is definitely bold, but it leads to a dead end: this scepticism must set aside its reservations and settle for practical everyday life, in which we act without persistent conscience of the groundlessness of our actions. Here we encounter the first affinity with Kripkenstein’s paradox, whose practical consequences equally boil down to nothing. However, in order to explore further analogies between Hume and Kripkenstein, it is necessary to develop a more radical form of Humean scepticism. Let us then set aside references to Kripkenstein for a moment and try to change the perspective on Hume’s paradox.

This “new” Humean scepticism concerns the relation between time and validity of statements. If we attribute the privileged position to the present, as Hume did, time cannot be deemed continuous. There are myriads of moments which already happened, there are numerous moments which are yet to come, but there is always only one moment of the present. And only this one gives empirical support to our statements because even the best observed events from the past are now our recollection, which cannot be unconditionally trusted. But is it only this support that must necessarily pass, or also the statement itself? If I see a bird right now and I write down on a scrap of paper a statement: “I see a bird now”, does this statement—formulated, let’s say, at a moment t ₁—belong to the past at a moment t ₂? Does it pass? Or does it remain indefinitely valid? If the bird I referred to is still there at a moment t ₂ and I say to myself: “I see a bird now” again, is it the same statement? Or is it different? Did the first statement pass with the moment in which it was formulated?

We are accustomed to perceive our judgements as independent from their concrete temporal pronouncements and believe in their abstract, atemporal validity. But if we reconsider Hume’s paradox profoundly, we would have to conclude that there is nothing in language that would immunise it against the disastrous impact of time. Why would the empirical perceptions have to pass, whereas our statements—which do not exist without some empirical manifestation—would remain untouched and still valid? If only present perceptions are of any value in justifying our knowledge, as Hume assumed, then also past statements cannot be effectively reproduced in the present. If I read now my notes I jotted down an hour ago, I should not be misled that only the circumstances I referred to at that time are gone; on the contrary—the very statement passed. If so, what do I read now? Another statement, which I formulate currently, with only dim recollection of some past act of writing.

In this radicalisation of Humean scepticism, which challenges our approach to language, all statements belong to moments of their formulation and irrevocably pass with them. Statements are not only irretrievable, but there is no method of verifying their mutual conformity. I cannot say that some past sentences, for instance p and q, are identical, because by saying so I do not refer to them but to some p’ and q’ that I have just created. The moment I utter such a sentence, let’s say s, it also belongs to the past and cannot be retrieved. It seems as if all my knowledge were reconstituted in the present, maybe in identical sentences (which I cannot ascertain), but definitely not the same. Yet this scepticism goes even further. If only currently formulated statements are fully legitimate and only those which relate to present impressions, then all our statements on past and future events are worthless. They might be pronounced or even believed, but nothing guarantees their reliability. Present statements are not valid for other moments of time, at least in terms of philosophical grounding.

What emerges from these questions could be described metaphorically as “the solipsism of the present”. Only in the present and on the present could we formulate justified statements; speaking about the past or the future, although customary, would have no actual validity. Even though we tend to perceive time as a continuous sequence of equal units which we know from recollection or anticipation, only the shifting moment of the present would be the real moment of time.

After we have thus re-interpreted Hume, let us see whether a similar procedure might be applied to Wittgenstein in Kripke’s exposition (“Kripkenstein”).⁴ Kripkenstein might formulate an analogous paradox, although applied not to affirmative statements, but to normative ones.

Kripkenstein’s paradox, to recapitulate it very briefly, concerns the usage of a rule in two different moments (cf. Kripke 1982, pp. 7–54). Let us name them t ₁ and t ₂. How can I be sure, asks Kripke after Wittgenstein, that if I constructed my rule at the moment t ₁, its meaning is preserved at the moment t ₂, when I apply it to a new case? How can I pose a link between what I meant in the past—and the current usage of my rule? Apparently there is no “fact”, as Kripke holds, that guarantees this meaning. Consequently, even though we project the use of a rule onto all future moments, nothing can guarantee the identity of a rule in its future applications. It is almost as if the time elapsing between two moments—of laying down a rule and its application—exerted a disastrous impact on the continuity of this rule in time⁵. At the moment of their formulation rules seem unproblematic, but as soon as this moment passes, their meaning starts to lack stability. Do they not resemble Humean statements in their link with the time of their creation?

Kripke acknowledges an evident analogy between Hume and Wittgenstein, yet for some reason he does not follow the path that this remark might open (Kripke 1982, pp. 62–64; cf. also McGinn 1984, p. 86). Let me then radicalise this analogy and suggest a conviction that both philosophers seem to share. Humean conception of causality and of the human self (which allegedly is nothing but a stream of changing impressions), as well as the Wittgensteinian rule-following paradox might be reasonably associated with both philosophers’ views on relations between time and language. For both of them, time works against validity of statements and stability of meaning. Interestingly, this effect of time does not manifest itself clearly: we speak about the past and the future, we read past writings, we lay down rules that will be followed in the future—with little concern about the justification of such actions. However, philosophical analysis reveals that no argument can support our expectation about future causal relations. Analogously, no fact guarantees that the rule currently applied remains identical with the one we established in the past.

To sum up these intuitions in one formula, one could say that for Hume and Kripkenstein there is no philosophical grounding that would assure preservation of validity of statements in time, either affirmative or normative ones. As long as the sceptical approach is adopted, continuous validity of statements (and rules) remains inexplicable. Yet in practice we assume that time is continuous and statements conserve their meanings. Both Hume and Kripkenstein are then tempted to accept a similar sceptical dualism between common practice and theoretical lack of justification. Both resort to acknowledging “blind action” that philosophy has to settle for. Nevertheless, we are unable to select one single “fact” that would accompany two separate moments and account for causality or stability of meaning.

Kant clearly identified the underlying presupposition of Humean sceptical paradox, namely the problem of continuity of time. As I suggested above, the ultimate consequence of Hume’s critique of causality would be the following: only the present can be described in a justified manner, since only then do we perceive impressions which are substantially different from unreliable imagination or memory. Humean time lacks continuity, as the past and the future are only our recollections or anticipations in the present. Consequently, certainty concerns only present impressions and statements which are currently formulated on them.

In his “Transcendental Æsthetic”, the opening part of The Critique of Pure Reason, Kant argues against selecting the present as the only legitimate form of time (cf. Kant 1855, pp. 28–35). In short, he points to the fact that time constitutes the ultimate “container” for all events—that is the only dimension in which all events can be compared (in terms of anteriority and posteriority). The following example will elucidate this argument. Contradictory statements, such as “It is raining” and “It is not raining” cannot be pronounced simultaneously with full assertion. We need to choose either of them, depending on the weather outside. Consequently, I cannot say “It is raining. It is not raining” and claim I speak the truth. However, once we differentiate moments of pronouncements of these statements, they might stand within the same narrative without contradiction (p can be true in t ₁, whereas non-p in t ₂). For instance, if I look from my window at noon, I might say “It is raining” and 1 h later I might say “It is not raining”. Then the sentence “At noon I had said it had been raining and 1 h later I said it was not raining” is not contradictory. Thus differentiation in time allowed me to juxtapose in one narrative statements of apparent contradiction. Consequently, time provides the most basic framework organising our experience coherently.