[1] no directive discourse can be true or false

Another possible formulation might be:

[2] only indicative discourses can be true or false

Many legal philosophers would be quite prepared to subscribe also to [2] and would consider it to be, to all intents and purposes, the equivalent of [1]. For my part I prefer to keep the two theses distinct and use only [1] here; this will equally serve my purpose, and is likely to generate less complications because of its more limited scope. As we know, the distinction between directive language and indicative language is not considered to be exhaustive by everyone who draws it. In addition, there are those who hold that apophanticity (truth-value) is a characteristic not only of indicative discourses, but also of other types of discourse, e.g. performative discourses. For this reason, theses [1] and [2] cannot be treated as equivalent and for our purposes it is better to use the weak thesis [1], which is after all contained logically within [2].

That truth-values cannot be referred to directives is a recurrent theme in the analytical philosophical literature; I would go so far as to say that this is one of its identifying traits, in particular in its Italian version. The fundamental, undeniable principle of analytical philosophy is the Great Division between indicative and directive discourses. The division is great and impassable (a logical gap), precisely because only indicative discourses can be true or false and qualified as such with the aid of intersubjective procedures; conversely, directive discourses can be neither true nor false, but can at most be qualified on the basis of such different values, other than truth, as individuals ascribe to them freely and that are ultimately the result of subjective choices.

So thesis [1] constitutes the real basis of analytical philosophy - a premise so basic and accepted that it is not necessary even to discuss it.

So it comes as no surprise that the analytical response to the question about the truth of that part of the directive that is legal is uncompromisingly negative: since law consists of directives (legal norms) and since directives can be neither true nor false, it follows that legal norms can themselves be neither true nor false. For the analytical philosopher, truth may only be a predicate, at most, of metalinguistic discourses (obviously only those in indicative mode) about law, but never of law itself. To put it bluntly, the question is simply not considered open to discussion; if anything is open to discussion, it will be the consequences of the denial of such attributes to directive discourses, in particular with regard to the logic of norms. It follows as inevitable that analytical philosophers will exclude law from the domain of truth: this is the necessary consequence of the conjunction between their divisionist semiotics and their metaethical non-cognitivism and non-objectivism. In analytical legal philosophy, non-objectivism and ethical non-cognitivism combine naturally with the voluntarist tradition that runs through all modern legal culture, from Hobbes to Kelsen: auctoritas non veritas facit legem is Hobbes' celebrated motto, on which Kelsen appends his lapidary statement: "there can be no question of a 'truth' of law. For law, even just law, is norm, and - as norm - can be neither true nor false".

But if he is to be coherent with his starting premises, the analytical philosopher must treat his own conviction, that legal norms can be neither true nor false, as the result of a choice: what we could call an ultimate choice of a philosophical nature, which means that it is open neither to logical demonstration nor definitive empirical proof. I also believe that this is how things stand. This is because the question of the truth of law takes us into the terrain of ultimate philosophical premises, which means that they are not open to further argument. Nevertheless, I believe it would be worthwhile to ponder a while on the topic of the truth of law: in the usual analytical way, to define the parameters, to draw some distinctions that have been missing until now or that have been drawn in a rather unsatisfactory way. In our field, the radical nature of the positions adopted has, as I have already suggested, resulted in practically all discussion being outlawed, with the consequence that the main reasons both for and against the ascription of truth-value to norms have often been proposed or refuted apodictically. Precious little interest has been taken in and precious little attention has been paid to the philosophical ideas latent in the various conceptions concerned with the relationship between truth and law, including those of the analytical conception itself.

The analytical philosopher's job is therefore primarily one of self-reflection, not least for the excellent reason that his own ideas are perfected and can be better defended, the more he refines them by comparing them with different ideas. Secondly, its usefulness becomes apparent as soon as he realises that the terms of the problem are not as simple as his own formulation may suggest. Indeed, the question "can law be true (or false)?" is irremediably ambiguous, for the obvious reason that there are various meanings that can be attributed both to truth and to law. In the following pages, the concept of truth will be used, so to speak, as an independent variable: several conceptions of truth will be scrutinised together with the ways in which they inevitably reflect on law.