It is commonplace sorting out two main kinds of argumentation in law: evidentiary argumentation and interpretive argumentation. The former concerns statements about the relevant facts to a lawsuit (like, e.g., “John Smith stabbed Henry Doe to death”), and is typically deployed for claiming such statements to be “true”, “false”, “established beyond any reasonable doubt”, “utterly (un)sound”, etc. The latter concerns interpretive sentences (like, e.g., “Section Y of the Traffic code expresses the norm N_i”, “The term ‘T’ in section Z of the Civil code refers to C₁ … C_n”), and is typically deployed for claiming such sentences to be “correct”, “right”, “wrong”, “true”, “false”, etc.

As soon as we move from the needs of ordinary law-jobs to the more demanding requirements of legal theory, however, the very notion and scope of interpretive argumentation appear saddled with uncertainty: What exactly does a piece of interpretive argumentation amount to? How can we tell genuine interpretive argumentation tokens from spurious ones? When is a purported instance of interpretive argumentation really “interpretive”?

Clearly, all such questions point to a boundary problem involving momentous jurisprudential issues like the proper concept and theory of legal interpretation, and the proper theory of legal norms, in so far as it intertwines with the former.

My purpose in this paper is describing and comparing two theories, both within the province of analytical jurisprudence (they have in fact very easily recognizable “godfathers” in such a line of inquiry), that provide alternative views about legal interpretation, written-law norms and interpretive argumentation. They are the frames of interpretations theory, on the one hand, and the container-retrieval theory (in its conventional linguistic meaning variety), on the other.¹

Supporters of the container-retrieval theory suggest their view cannot be reasonably resisted by supporters of the frames of interpretations theory.² By a critical survey of both theories, I will bring to the fore a few arguments for considering the resistance to the container-retrieval theory as being by all means reasonable: indeed, a theoretical “must”.

The frames of interpretations theory (hereafter, for brevity sake: “frames theory”) is an interpretive (or “interpretivist”) legal theory within analytical Kelsenian realism. It considers interpretation to be a key activity that intervenes at all crucial points in the working of “our” legal systems. It claims, accordingly, that the terminology and conceptual apparatus of a useful legal theory should accommodate to the pivotal role interpretation plays in legal experiences; it suggests that theoretical concepts, to be adequate, must be either interpretation-dependent (they must so far as possible bear a conceptual connection to interpretation), or salva interpretatione (they must openly rule out such a connection to some valuable theoretical purpose at hand).

The following components have to be considered in order to provide an account of the frames theory suitable to a comparison with the container-retrieval view: (1) the distinction between authoritative legal sentences, explicit norms, and implicit norms; (2) the distinction between interpretation and integration (“juristic construction”, “juristic law-finding”, “juristic law-making”); (3) the argumentative conception of legal interpretation and integration; (4) the conception of interpretive and integrative argumentation as institutional games; (5) the distinction between practical and cognitive interpretation; (6) the twin claims of universal methodological ambiguity and potential axiological ambiguity; (7) the idea of general norms of written-law as frames of interpretations, with the related minimalist conception of “written law”.

The activity of interpretation – the matter of which are legal clauses and the outputs of which are explicit norms – as it is usually performed by judges, other legal officials, jurists and lawyers at large is interpretation proper to practical purpose (practical interpretation proper): it determines the “correct” (“proper”, “true”, “right”) meaning of a legal clause; it translates a legal clause into the “correct” (“proper”, “true”, “right”) explicit norm, either in view of deciding a lawsuit (judicial interpretation), or in view of affecting such a decision (forensic interpretation), or else in view of providing the “right” solution to an abstract quaestio juris (juristic interpretation). It is, in any case, a will-geared, decision-making activity: it decides for a meaning as “the correct” meaning of a legal clause, at the same time tacitly or expressly ruling out alternative possible meanings as “uncorrect”.

By contrast whenever, for instance, a judge identifies an implicit norm in view of deciding a lawsuit, by definition such an activity of hers is not (practical) interpretation (proper): it is rather a piece of law integration; indeed, she determines the “proper” (“correct”, “true”) implicit norm to be applied to a case at hand, presupposing explicit norms have run out.

Like remarks apply to integration, the results of which are (liable to be) supported by some set of integration arguments (like, e.g., integrative uses of analogical, a contrario, a fortiori or general principle reasoning).

Interpretive arguments are built upon interpretive directives: for instance, any literal argument concerning a piece of legislation has its ground and starting point in a directive like “Statutory clauses should be interpreted according to their literal meaning”.³ The discrete set of interpretive directives licensed players select and employ to present a norm as “the correct” meaning of a legal clause is an interpretive code.

The basic rule of the interpretive argumentation game runs, roughly, as follows: “Employ the interpretive code that enables you to deploy the best set of arguments in view of presenting the interpretive output you set forth as the only correct one for the case at hand”.

An interpretive code, as a tool for justifying interpretive outcomes, is typically made of three sorts of interpretive directives: primary translation directives, secondary procedural directives, and secondary preferential directives.⁴

Primary translation directives are instructions pointing to resources (empirical data, actual or imaginary pieces of information, actual or conjectural ingredients of the legal order and legal experience) by means of which a legal clause should be translated into some explicit norm. We may single out five different types of translation directives: (1) directives of linguistic interpretation (e.g., “Statutory clauses should be interpreted according to the ordinary linguistic meaning of their expressions at the time of their enactment”); (2) directives of intentional or genetic interpretation (e.g., “Statutory clauses should be interpreted according to the original semantic intention of the historical legislator”; “Statutory clauses should be interpreted according to the counter-factual semantic intention of the historical legislator”); (3) directives of teleological interpretation (e.g., “Statutory clauses should be given the meaning pointed out for them by the objective purpose they serve”); (4) directives of authoritative interpretation (e.g., “Statutory clauses should be given the meaning established for them by the Supreme Court”); (5) directives of substantive, normative-ethics, interpretation (e.g., “Statutory clauses should be given the meaning pointed out for them by the critical morality they refer to”).

Secondary procedural directives are instructions pointing to the order that should be followed while deploying arguments from primary translation directives (e.g., “Deploy first an argument from primary directive PD1, and then from primary directive PD2, and then …”).

Secondary preferential directives, finally, are instructions pointing to the criteria that should be adopted to justify: (a) the ruling out of a given interpretive outcome (proposed, for instance, by the counsel of the plaintiff), as being “incorrect” in itself or by comparison with a different, “more correct” outcome (proposed, say, by the counsel of the defendant or by the judge); (b) the acceptance of a given interpretive outcome as all-things considered “correct”. Secondary preferential directive typically include so-called “systematic (interpretive) arguments” (like, e.g., the argument from coherence and the argument from completeness: “Statutory clauses should not be given any meaning logically incompatible with constitutional principles”, “Statutory clauses should not be given any meaning teleologically incompatible with the fundamental principles of the legal system”, “Statutory clauses should be given the meaning, among the several ones identified by means of primary directives PD₁ … PD_n, that is most instrumentally in tune with the requirements of fundamental principles”, “Statutory clauses should not be given any meaning showing the law to be incomplete as to the case at hand”, etc.).

One last remark seems in order, before proceeding. The methodological tradition (and the legal clauses that sometimes purport somehow to “approve” it) usually provide lawyers with a disordered set of indeterminate interpretive directives. Accordingly, when lawyers use interpretive codes to justify some interpretive outcome (i.e., some explicit norm), such codes are often made of directives they themselves have somehow shaped, sharpened, made more precise as to the interpretive resources to be used, and put in the “proper” order. Fatally, such a shaping, sharpening, resource-selecting and ordering is likely to mirror, and be affected by, each lawyer’s own methodological attitude, ideological stance and material interests. This suggests the following conclusion: the selection and use of an interpretive code by a lawyer interpreting a legal clause, i.e., playing the interpretive argumentation game, is a discretionary, value-laden, activity. This is one of the reasons – perhaps, the main reason – why, as I said before, practical interpretation proper (“textual interpretation”, “adjudicative interpretation”) is, according to the frames theory, a will-geared, decision-making activity. The same remarks apply to the game of law integration.

So far, we have considered the practical variety of interpretation proper: i.e., interpretation to practical purpose. However there are at least two further varieties of interpretation proper, not to practical purpose (at least: not directly and immediately so), but to theoretical or cognitive purpose (cognitive interpretation proper). These are the varieties of conjectural interpretation and creative interpretation.

While dealing with conjectural interpretation it is worthwhile distinguishing, in turn, two (sub)varieties: methodological conjectural interpretation and axiological conjectural interpretation.⁵

Methodological conjectural interpretation consists in laying bare, as to a given moment t’, the meanings that can be ascribed to a legal clause (say, LC_i), on the ground of the interpretive directives the legal culture considers as “required” or “approved” “by the law”.⁶ In so doing, the interpreter must avoid qualifying any of such meanings as “the only correct” (“true”, “right”, “just”) meaning of the legal clause. Rather, she should limit herself to working out a dispassionate inventory of meanings. This can be a minimal or a maximal inventory: in the latter case, she will claim it to account for (almost) all the methodologically viable meanings of legal clause LC_i, to exhaust the hermeneutical potentialities of LC_i, having reached, so to speak, the ultimate frontier of its possible meanings. The process of methodological conjectural interpretation may be recounted, tentatively, as including four stages.

In the first stage, the interpreter must identify the set of allowed interpretive directives (techniques, methods, criterions, rules) that may be considered as required or allowed by the law and/or the legal culture, in view of interpreting LC_i [SAID_t’ = ID₁, ID₂ … ID_n].

In the second stage, the interpreter must identify the set of possible combinations of interpretive directives, i.e., the set of allowed interpretive codes [SAIC_t’ = IC₁, IC₂ … IC_r].

In the third stage, the interpreter must identify, for each of the several interpretive codes previously singled out [SAIC_t’ = IC₁, IC₂ … IC_r], the related set(s) of interpretive resources [SR₁, SR₂ … SR_p].

In the fourth, and last, stage, the interpreter must conjecture (calculate) the set of meanings that can be ascribed to the legal clause LC_i, from the standpoint of the several combinations of allowed interpretive codes and related sets of interpretive resources [SMLC_i = EN₁ [f (IC₁, SR₁)], EN₂ [ƒ (IC₂, SR_2’)] … EN_r [ƒ (IC_r, SR_r)]. The set of alternative meanings so identified for the same legal clause LC_i – or, in other words, the set of alternative explicit norms (EN₁ … EN_i) into which LC_i can be translated – makes up the “frame” of possible interpretations of LC_i. Each of those meanings is a methodologically correct meaning: i.e., it is correct, by hypothesis, from a purely methodological point of view.

Axiological conjectural interpretation represents a variety of methodological conjectural interpretation. Here, the interpreter aims at identifying, not just the methodological interpretive frame of a legal clause (say, LC_i), but its axiological interpretive frame. Such a frame depends not only on the interpretive methods and resources available, but also on social axiology: more precisely, on the ethical views prevailing, or in any case recordable as being influential, in the society. These views may make a methodologically viable interpretive outcome unviable, for reasons having to do with the prevailing negative substantive social value of such an outcome.⁷ Accordingly, the scope of the axiological frame is likely to be narrower than the scope of the methodological frame.⁸

Creative interpretation consists, finally, in the identification of one or more meanings for a given legal clause (say LC_i