2. The place and purpose of Article 38

An immediate suspicion should arise when one looks at the first paragraph of Article 38 as a whole, and not just at letter (b) on custom that forms its second heading.Footnote ²¹ The other three clauses obviously do not purport to be definitions. All they do is to point out the categories of rules and normative materials to be taken into consideration by the ICJ: conventions, general principles, judicial decisions, and teachings of publicists. These legal categories are simply enumerated with some specifications, but not properly defined. It would, therefore, be rather odd if the second clause stepped out of the row and were really intended to constitute a formal definition of international customary law instead of just to enunciate another set of relevant norms. While it may superficially look like a definition, it was not meant to be one.

On reading it without preconceptions one cannot evade the impression that the whole of Article 38 was never supposed to define the sources of international law at all, still less to ‘codify’ them as claimed by Lauterpacht, or to give a full list of them. Rather, its basic purpose is to circumscribe the Court’s latitude as to the rules it ‘shall apply’ when deciding ‘such disputes as are submitted to it’. The article can best be compared with the provisions on the applicable law as they are typically found in arbitral agreements or, for that matter, with Article 7 of the abortive Hague Convention XII of 1907 on an International Prize Court.Footnote ²² The Statute of the ICJ after all is a sort of permanent compromis, and Article 38 figures in what may be seen as its heartpiece, Chapter II, entitled ‘Competence of the Court’. Having first defined its competence ratione personae (Articles 34 and 35) and ratione materiae (Articles 36 and 37), it ends with what might be called the Court’s competence ratione adhibendi iuris, that is, its discretion in drawing the various normative registers that may come into play in performing its task. Such, then, would seem to be the meaning of Article 38 in its immediate context. It is amply confirmed if we widen the perspective and have a look back at its origins.

The debates surrounding these origins were not lofty considerations in an academic seminar on assessing the various sources of modern international law, but very practical, sometimes heated discussions in a politically loaded bargaining on which sources were to be applied by the international tribunal that was being set up. This is what appears when we step back in time, not just to 1945 when the present Statute was adopted, but by a whole century when its predecessor was elaborated by the Advisory Committee of Jurists in charge of drawing up the Statute of the PCIJ under the terms of Article 14 of the League of Nations Covenant.Footnote ²³

3. The Advisory Committee and the Descamps Proposal

The Committee held its sessions in Paris during five weeks in June and July 1920.Footnote ²⁴ It numbered ten jurists appointed by the Council of the League of Nations from the five principal victor nations (France, Great-Britain, Italy, Japan, United States of America) and from five other, mostly European countries (Belgium, Brazil, The Netherlands, Norway, Spain). While they were not technically representing their states as in a diplomatic conference, the members of the Committee were nonetheless representatives of their specific legal educations and cultures, with the corresponding outlook on international law and on the judicial function in general. Civil law systems were clearly prevalent in number, but they were easily matched by the prestigious delegates of the two great common law countries, Lord Walter Phillimore and Elihu Root (assisted by Dr James Brown Scott, who attended all the sessions of the Committee and acted at times almost as one of its members).Footnote ²⁵ The two Anglo-American members were to prove the main antagonists of the president of the Committee, Baron Edouard Eugène Descamps, of Belgium, a distinguished veteran professor, senator, and statesman who, for having attended both Peace Conferences at The Hague in 1899 and 1907, considered himself ideally suited to be in the chair, notwithstanding some polite dissents among his colleagues.

Descamps was the author of a Proposal on the applicable law, the ancestor of Article 38, which reads as follows in its original French text:Footnote ²⁶

1. 1. le droit international conventionnel, soit général, soit spécial, comme manifestation de règles expressément adoptées par les Etats;

2. 2. la coutume internationale, comme attestation d’une pratique commune des nations, acceptée par elles comme loi;

3. 3. les règles de droit international telles que les reconnaît la conscience juridique des peuples civilisés;

4. 4. la jurisprudence internationale, comme organe d’application et de développement du droit.Footnote ²⁷

What has been said of Article 38 becomes even more obvious in its precursor: as the opening sentence makes clear, Descamps’ purpose was not to define the sources of international law in general terms, but merely to enumerate the categories of relevant rules to be taken into account by the ‘judge’. These are presented as constituted, existing sets of rules, not as norm-producing, formal ‘sources’ of international law.

Apart from the genetic link with Article 38, however, there is a generic discrepancy between the two texts. By its overall features and spirit Descamps’ Proposal does not look at all like a draft article to be adopted as such after having been critically examined and discussed in the Committee, with possible amendments in its wording. The comparison with Article 7 of the ill-starred Hague Convention XII of 1907, which was in everybody’s mind, is in this respect quite revealing: it simply posits without more ado the rules to be applied by the international prize court.Footnote ²⁸ By contrast, Descamps accompanies each of his items with a few words of comment, which have an almost didactic air about them. The whole Proposal is not drawn up in the style one would expect from a true draft article, to begin with the introductory sentence mentioning an unspecified ‘judge’ instead of ‘the Court’, as would have been called for by Article 14 of the League Covenant. Rather than being a projet d’article, the Proposal seems to have been meant as a preliminary base de discussion, and it is indeed labelled as such in the Committee’s Procès-Verbaux.Footnote ²⁹ Its purpose, as suggested by the formulation of the clauses, was to channel the debates that would only thereafter lead to a proper draft article. In other words, Descamps had launched something like a test balloon; but it was a dirigible balloon, set on a clear course. Far from being a neutral sampling of possible types of rules to be admitted or rejected, the Proposal contained an implicit argumentation aiming at an integral approval of the whole scheme underlying its four clauses, that were to be applied ‘in successive order’.

The four of them have indeed a common pattern inasmuch as they all end with a short comment indicating the principle of validity of each category of rules concerned, and vindicating in the same breath their relevance for the ‘judge’. All of them state the reason why the said rules ‘are to be applied by the judge in the solution of international disputes’. Such is obviously the case, first, of conventional law, for the reason that it consists of ‘rules expressly adopted by the States’.Footnote ³⁰ Secondly, if there are no treaties applying to the case, international custom will quite as naturally have to step in, as ‘being [a common] practice between nations accepted by them as law’.Footnote ³¹ Thirdly, failing applicable treaties and customary law, there are yet other ‘rules of international law’ that could not be ignored inasmuch as they are ‘recognized by the legal conscience of civilized nations’.Footnote ³² Finally, ‘international jurisprudence’ is obliquely relevant ‘as a meansFootnote ³³ for the application and development of law’.

Such, then, seems to have been the original intent of the Proposal Descamps submitted to his colleagues on 1 July 1920. It was an exploratory document with an almost apologetic bent, commending a maximal normative latitude for the future court. It was not a catalogue of the sources of international law in general, nor did it define any one of them in particular. What looks like explanations clarifying the four sets of rules are in fact motivations for their inclusion into the spectrum of applicable law. The second clause makes no exception, in spite of a fugitive impression. It is not meant as a definition of customary law, nor as an account of its formation or a guideline for its identification. All it does is to bring home its relevance to the members of the Committee, and hence to justify its place in the court’s toolbox. Justifying something does of course imply some degree of explanation, but there is more rhetorics than theory in it.

Admittedly, the first two clauses of Descamps’ Proposal were in no need for such a commendation: treaties and custom had both long since been regarded as the undisputed basic constituents of positive international law. But this was much less so, for the two other items: ‘jurisprudence’ (that is, case law resulting from judicial or arbitral proceedings), was rather a source of legal wisdom and inspiration than a source of law, which is why Descamps presents it ‘comme organe d’application et de développement du droit’; and his ‘conscience juridique des peuples civilisés’, though not an uncommon phrase among Continental jurists in the decades preceding the First World War,Footnote ³⁴ was bound to resonate somewhat suspiciously to the ears of Anglo-Saxon practitioners. Here was obviously the main reason for the commendations appended to each clause; and it seemed somehow normal that all the four of them should be fitted out with such a motivation in order to ensure the overall acceptance of the Proposal.

4. The Proposal’s deep structure and practical thrust

Granting that Descamps’ intent was above all pragmatic does not preclude, however, that his legal panoply also hinted at various law-creating processes and spelled out a theoretical view of the sources of international law. His conception could roughly be termed ‘Grotian’ inasmuch as it rested on two pillars: positive law, on the one hand, as enacted by the will of the states, directly by conventions or indirectly through custom; and, on the other hand, ‘natural’ principles that exist ‘objectively’, by themselves as it were, beyond any human volition or doing, because they are inherent in human nature and society, and hence spontaneously ‘recognized by the legal conscience of civilized peoples’. This ‘recognition’ was moreover to be actively prompted by ‘international jurisprudence’ (that is, at the time, mainly case law of international claims-commissions and arbitral awards, as well as judgments of prize courts or other domestic jurisdictions of superior rank dealing with international matters)Footnote ³⁵ and great jurisconsultsFootnote ³⁶ (just as Grotius had adduced the sayings of ‘wise men’ from all ages, and preferably from classical Antiquity, to reveal the principles of natural law).

To be sure, while the dualistic structure of Descamps’ system was thus indeed ‘Grotian’ in spirit, he preferred to avoid the expression natural law, which had been thoroughly discredited by the onslaught of nineteenth century positivism. He would instead rather name it ‘objective justice’, in tune with the sociological trends in legal philosophy which were rife by the turn of the century, especially in France and Belgium. Yet, to all ends and purposes this allegedly objective law was in many respects but a rejuvenated avatar of classical natural law, that is, the ‘true’ law inherent in man as a social being, and hence in human society, apart from any positive enactment, based as it was on the key principle of social solidarity. As Descamps endeavoured to explain to his sceptical colleagues, such a ‘fundamental principle of justice’ was equally permeating the society of states, giving rise to ‘certain rules, necessary for the system of international relations’ (‘certaines règles nécessairement liées à l’économie essentielle des rapports sociaux dans la vie internationale’).Footnote ³⁷

While these theoretical considerations were firmly rooted in his mind, they were not the only reason for Descamps to extoll the rules of objective justice. It was above all for jurisdictional effectiveness that he advocated their inclusion into the future court’s normative arsenal. His main concern was to avoid that the court find itself at a loss for want of relevant conventional and customary law. It would then be constrained to pronounce a non liquet and leave the dispute unsettled, which in Descamps’ eyes would constitute a denial of justice (just as it would in municipal legal orders).

The latter point has to be seen against the background premise of the court’s jurisdiction being compulsory. This was not at first a matter of course to the members of the Committee, and, as we know, it was not in the end accepted by the governments that finalized the Court’s Statute in the League of Nations. For the time being, however, this point had been left open by Article 14 of the League Covenant, which formed the legal basis of the Advisory Committee’s deliberations. Was the Court’s jurisdiction (or ‘competence’, as it was named in civil law style, which partly survives in the English version of the Statute) to be subject to the parties’ consent in each case, or should it be compulsory, and if so, for what types of disputes? This fundamental choice, which in the latter option impinged on the jurisdiction ratione materiae, was the object of the three meetings immediately preceding the discussion of the applicable substantive law.Footnote ³⁸ The Committee had determined, despite some reservations of its Japanese member, Mineichirō Adachi, that the Court’s jurisdiction should be compulsory at least for the types of disputes enumerated in Article 13, paragraph 2, of the League Covenant.Footnote ³⁹ This was indeed what the members had mostly in mind when thinking of a genuine judicial court reaching beyond a merely arbitral tribunal.

So far so good, but as a corollary of this position the court would be under a strict duty to render a judgment in all cases submitted to it. It could not allow itself to get away with a non liquet. One cannot oblige the members of a legal community to submit themselves to jurisdictional proceedings instead of ‘taking the law into their own hands’ without securing them at the same time an effective decision. Pronouncing a non liquet (as the Roman judges were allowed to do in the formulary proceedings if the matter was ‘not clear’ to them)Footnote ⁴⁰ would amount in modern conceptions to a denial of justice. This is what Descamps was endeavouring to avoid through his objective justice: apart from theoretical preferences it was in line with the understanding that the court’s jurisdiction would be compulsory.

5. The Descamps Proposal challenged

Though Descamps’ colleagues mostly shared his concerns, they were doubtful as to the means he offered to allay them. There was indeed another aspect to compulsory jurisdiction which threatened to jeopardize the whole venture. While the principle was quite normal in domestic jurisdiction, it meant almost a revolution among states. If sovereign nations, and especially great powers, were expected to cope with compulsory jurisdiction, they had first to be reasonably certain not to run inconsiderate risks due to the law that would apply. The applicable rules had to be as ‘positive’ as possible, allowing for maximum control on their part: treaties only, if possible, recognized custom at most, if need be. Descamps’ objective justice was not a good bid in this perspective, however well-meant in other regards.

Unsurprisingly, Descamps’ main antagonists turned out to be the representatives of the two foremost powers at the time, Elihu Root, a former Secretary of State of the United States, discreetly backed by Lord Phillimore, a member of King George V’s Privy Council. Already toward the end of the discussion on the scope of the Court’s jurisdiction, even before the question of the applicable rules was broached, Root entered a caveat calling attention to the acceptability of an all-too comprehensive jurisdiction: ‘The States’, he admonished, ‘would not accept a Court which had the right to settle disputes in accordance with rules established by the Court itself, and by the interpretation of more or less vague principles.’Footnote ⁴¹ He recalled that Article 7 of Hague Convention XII of 1907 had become the major stumbling-block for the institution of the international prize court precisely owing to the vagueness of the principles it was to apply.Footnote ⁴² ‘Nations will submit to positive law’, he warned, ‘but will not submit to such principles as have not been developed into positive rules supported by an accord between all States’.Footnote ⁴³

When Root made this pragmatic statement, probably thinking of his own country that was drifting into isolationism, he was already acquainted with Descamps’ Proposal, which had been circulated beforehand; he was anticipating the tug of war that was going to oppose him to the Baron, with whom he felt but little affinity anyway. Conversely, Descamps, who fully reciprocated these feelings, must have been aware, when drawing up his Proposal, that he had to tread lightly in order to get his way. This is doubtless what explains the apologetic turn of his preliminary document, and especially the justifications accompanying each of its four clauses.

As we have seen, it was mainly the third clause, calling on ‘the legal conscience of civilized peoples’, that needed such an advocacy. And this was indeed the clause which Root pounced upon as soon as the discussion on the Proposal was opened. The fourth clause on ‘international jurisprudence’ fared hardly better. In line with his previous observations, Root considered that ‘[t]hese two clauses constituted an enlargement of the jurisdiction of the Court which threatened to destroy it’.Footnote ⁴⁴ One could not, in his opinion, ‘compel nations to submit their disputes to a Court which would administer not merely law, but also what it deems to be the conscience of civilized peoples’.Footnote ⁴⁵ What appeared to Descamps as the ‘objective law of justice’ illuminating the world must have frightened Root as a kind of legal blackhole. Lord Phillimore for his part, in good common law spirit, thought that clauses 3 and 4 could in any event (and should indeed) be expunged and subsumed under international custom.Footnote ⁴⁶ But Descamps, in a memorable statement, passionately held his ground.Footnote ⁴⁷ After a lengthy discussion, Hagerup, the Norwegian member, diplomatically suggested that Root might rephrase clause 3 to make it easier to swallow for confirmed, impenitent positivists.Footnote ⁴⁸ Together with Phillimore, Root thereupon brought Descamps’ ‘legal conscience of civilized peoples’ under control by relabelling the clause ‘general principles of law recognized by civilized nations’,Footnote ⁴⁹ which was in fact the formula that would be retained in the end. Descamps meekly gave in to this reformulation, declaring it sufficient to drive off the spectre of non liquet.Footnote ⁵⁰ By the same occasion clause 4 was agreed upon in its equally modified version, referring henceforth not only to ‘the authority of judicial decisions’ but, in addition, to ‘the opinions of writers’, which after some debates became ‘the doctrines of the best qualified writers of the various nations’,Footnote ⁵¹ both sources of authority being at this point still thought of, in Descamps’ terms, as ‘means for the application and development of law’.Footnote ⁵²

6. From basis of discussion to draft article

The clash of views between the President and his colleagues went on for two whole meetings. While clause 3 was clearly the main battleground, there were some skirmishes also on clause 4 and on the ‘successive order’ prescribed by the preamble for the application of the four clauses. The first two of them, by contrast, did not raise any objections and were therefore hardly mentioned during the whole debate.

The controversy on clause 3 had, however, an unintended and unnoticed, yet crucial, side-effect: it entirely changed the status of the Baron’s whole negotiating document. What he had launched as an exploratory base de discussion to disentangle the subject matter moved at once up to a formal projet d’article, and it was from the beginning perceived and worked upon in its entirety as a draft article. This surreptitious change was an indirect consequence of the particular configuration of clause 3. As observed earlier, the four clauses initially had a common structure: each category of rules listed was followed by a short comment as to why the court should be enabled to draw on it. These comments were intended for internal use during the preliminary discussion; they were directed at the members of the Committee, and at them only. Once they had produced their suasive effect for the several clauses, they were to vanish (just as the male of the praying mantis is said to be utterly devoured by his mate after fulfilling his debitum coniugale). They had no place in the final article, which could ideally limit itself to a bare mention of the classes of rules retained (except possibly for clause 4, to stress its subsidiary function).

If such had indeed been the President’s intent when he jotted down his basis of discussion – and in view of the available materials this cannot be more than a plausible conjecture – it harboured a snag of his own doing. For he had, maybe on purpose, refrained from providing the third category of rules with a specific label such as ‘the law of objective justice’ (as he was to do, the day after presenting the Proposal, in his statement defending the clause against its detractorsFootnote ⁵³). Instead, he had just tagged it ‘the rules of international law’, a blank label which did no more than to suggest an ostensibly well-defined and recognized category of legal norms applying between states. Could it be that this colourless designation was to act as a reassuring camouflage to make the clause less obnoxious to the staunch positivists in the Committee? At any rate, these ‘rules’ were specified only by the second component of the clause, which commended them ‘as recognized by the legal conscience of civilized peoples’. This specification was thus a necessary constituent of the clause, it could not simply be dropped as long as the first component had not received a proper denomination. But as the second component raised objections of principle as to the substance of the ‘rules’ in question, the nature and wording of the whole clause was immediately discussed, already at this preliminary stage, in terms of a true draft article. This is how Descamps’ ‘rules of international law’ turned into ‘general principles of law’, which for safety had to be nailed down as being positively ‘recognized by civilized nations’, with no ‘legal conscience’ any more allowed to intrude. Clause 4 having also attracted Root’s thunderbolts was in its turn drawn into the same dynamics and was equally discussed in that same spirit.

Reformulating clauses 3 and 4 thus turned them implicitly into parts of a proper draft article; and this is how the whole of the Proposal was inadvertently perceived from the beginning. It was only natural therefore that its first two clauses would be seen in the same light with their hortatory end components. Instead of being deleted as presumably anticipated by Descamps, the appended comments stayed in place, subject only to drafting changes, although they would serve no purpose any more. Why indeed tell judges or the litigants that international conventions establish ‘rules expressly adopted by States’? This is a truism – and was obviously meant as such by Descamps – no less than reminding them that international custom amounts to ‘practice between nations accepted by them as law’. Judges ‘of recognized competence in international law’ are in no need of such information, and the same can be expected from persons otherwise qualified ‘for appointment to the highest judicial offices’ in their country.Footnote ⁵⁴ As it came, however, these merely incentive components of Descamps’ basis of discussion were dealt with as integral parts of what had become a fully-fledged draft article, and instead of being dropped they subsisted as such, except for some adjustments, in the final Article.

7. The clause on customary law and its conventional law twin

The moment has come, however, to turn more particularly to the second clause of the Descamps Proposal. What should be avoided here (for our purpose at least) is to consider it in isolation (as would otherwise be normal). In order to capture its initial meaning it has to be read just as we left it a moment ago, in conjunction with the first clause. Born together like twins, they form a pair, encompassing the entire field of positive international law, as against the rules of objective justice of clause 3; yet they are dissimilar twins, which accounts for some distinct traits in spite of their fundamental kinship.

The common legal quality of the two classes of rules is due to their shared principle of validity: both were considered by definition to proceed from state will, which was but the obverse of their essence as positive law. In this guise treaties and custom had indeed by the end of the nineteenth century become the generally recognized basic sources of positive international law, itself being more and more looked upon as international law tout court, with positivism largely reduced to legal voluntarism.

Covering the whole of positive international law, the first two clauses of Descamps’ Proposal were therefore seen by him as strictly complementary, in the manner of yin and yang in the circular tai-ji symbol.Footnote ⁵⁵ They have to be read one in the light of the other. This is what gives its meaning to the phrase describing international custom as ‘a common practice among nations accepted by them as law’.Footnote ⁵⁶ It implicitly refers to the first clause: the fact for custom to consist of ‘practice accepted as law’ by the generality of nations is the counterpart of conventions forming ‘rules expressly adopted’ by the various states. Customary law, just as conventional law, is supposed to derive from volition, though instead of being formally laid down and ‘expressly adopted’ in writing, it is tacitly ‘accepted as law’ by being informally practiced as such among the whole community of nations. In terms of voluntaristic positiveness this may be second best if compared with contractual engagements (or, for that matter, with legislative enactments in domestic law), but it was nonetheless construed as resulting from an implied meeting of wills. Custom has indeed since Roman times been regarded as implicitly ‘willed’ by the people acting rebus ipsis et factis instead of declaring its will by formal suffrage,Footnote ⁵⁷ and this traditional explanation of customary law as based on ‘general assent’ or ‘tacit consent’ (tacitus consensus omnium)Footnote ⁵⁸ was still a commonplace in Descamps’ time, be it in municipal or international law. Owing to an exacerbated conception of state sovereignty, these Romanistic formulas were taken at face value by some (especially German and Italian) positivists – Arturo Ricci-Busatti, the Italian member of the Committee, being among them, in full complicity with its Secretary general, the arch-positivist Dionisio Anzilotti. Ricci-Busatti did in fact criticize on this ground Descamps’ differentiated wordings in the first two clauses of his Proposal. To be validly invoked in court, customary rules, instead of merely being ‘accepted as law’ by the nations in general, had in his opinion to be specifically adopted by the litigant states themselves: ‘Custom, like any convention applicable to a case, must be in force between the parties in dispute’, he asserted dogmatically, but without being followed by anyone from among his colleagues.Footnote ⁵⁹

Descamps in particular would hardly have agreed to reduce international customary law purely and simply to a kind of contractual web; which brings us, after their basic homogeneity, to the specific difference between his two clauses. The principle of will and consent underlying them both found in custom a sort of muffled expression as implicit, in fact largely constructive recognition of a body of rules applying indistinctly to the whole community of nations, even without being ‘expressly adopted’ by all its individual members. The two yin and yang components encapsulating positive international law were thus complementing each other, but also contrasting by their legal fabric. Such a disparity is already perceptible in the surviving fragments of the above-mentioned Roman jurisconsults, despite their contractual language in matters of longa consuetudo. They did indeed base it on a presumed tacita civium conventio,Footnote ⁶⁰ but it was supposed to be an equivalent of legislation: long-standing usage approved by the consent of those sharing in it was declared to ‘imitate the law’ by emperor Justinian himself.Footnote ⁶¹ It was understood that this ius ex non scripto Footnote ⁶² could only be of subsidiary importance in Roman law, just as in any other legal order governed by a genuine, comprehensive legislation. Such is not the case by definition among sovereign states in the international community, which is therefore notoriously craving for a substitute, for a ‘general international law’ in twentieth century parlance, fulfilling the function of an absent legislation.

8. Custom as the common law of nations

It was precisely in this quasi-legislative capacity that Descamps put forth international custom as being ‘une pratique commune des nations, acceptée par elles comme loi’. The key words are ‘commune’ and ‘loi’. The adjective ‘commune’ (omitted in the English translation) discloses the normative potential of the practice, involving by itself its acceptance ‘as law’ by the whole community or, in other words, its status as the common law among nations. It was indeed as such that Lord Phillimore, back in London, was to present the clause to the members of the recently founded Grotius SocietyFootnote ⁶³ (although the adjective ‘common’ had in the end been altered to ‘general’, which does not convey the same comprehensive, all-embracing momentum). In Descamps’ mind the common practice, made up of the cumulative actions and positions of all the nations, was deemed of itself to be accepted by them as their common law.

This is also what appears in the President’s emotional speech of 2 July defending his Proposal against the objections it had incurred the day before, especially with regard to its third clause. Before getting to this chief bone of contention he briefly touched upon the first two clauses, although they had raised no objections, as if to gather strength and speed for the decisive clash of arms ahead. Nobody would deny, he argued on entering the lists, that, ‘when a rule is expressly laid down by a general or special treatyFootnote ⁶⁴ between the parties, the primary duty of the judge is to apply it’.Footnote ⁶⁵ The same holds good, he continued, ‘when the judge finds himself in the presence of a clearly defined custom, of a rule established by the constant, general practice of the nations, having thereby obtained force of law among them’ (‘une règle établie par la pratique constante, générale, des nations, devenue, à ce titre, une loi pour elles’).Footnote ⁶⁶ The very nature of the practice, its generality and consistency, seems by itself to vouch for its legislative potential. Descamps might have stopped there, but to clinch his point he added some sweeping observations on the historical importance of custom for the rise and progress of the law of nations. Given ‘the absence of legislation in this field and the limited development of conventional law between the States’, he explained, custom is indeed the most natural means for its growth, ‘since it is but the expression of the nations’ legal convictions and social needs, as disclosed in their common and constant ways of acting’.Footnote ⁶⁷ The legislative function of customary law as a whole is thus implicitly reasserted, although the common practice appears less as the determining factor than as an expression of deeper social needs and moral forces, in a manner recalling Savigny’s historical school or the more recent sociological conceptions.

Such was Descamps’ presumable idea of custom and its place among the sources of international law. Yet, we should remember that, whatever doctrinal implications one may perceive in these incidental observations, they were not really meant as such; their purpose was mainly instrumental, they had to back up Descamps’ argumentation in this particular context. He was not lecturing on the nature and functions of customary law but affirming its place in the list of applicable rules (although in this case, as for conventions, it was shadow-boxing or just warming up for the next two clauses, since the admission of customary law had not been challenged). Descamps’ point, and the underlying message to his colleagues, was that the states, as potential litigants, could by no means take exception at least to these types of positive rules being applied to their future disputes, since it simply amounted to taking them at their will, actual or constructive. This is what their respective characteristics were initially expected to convey as a matter of course to the members of the Committee. But Descamps’ message did not stop there: while he took both categories of positive law for granted, he brought out in the same movement their insufficiency by confronting them at once with the ‘objective’ rules of the third clause, which were closest to his heart but most questionable to his colleagues. All this was clearly spelled out in his speech of 2 July; it was latent already in his sketchy Proposal, and especially in its second clause, which has to be read in this light. That was its initial purpose, and nothing more.

9. The Baron’s unnoticed blunder

This leads us now to consider the twin-clauses in a more formal perspective. While they were cognate by their substance, they also had similar structures.Footnote ⁶⁸ In Descamps’ initial wording both sets of rules were at first named (‘le droit international conventionnel’/‘la coutume internationale’) and both were then shortly characterized (‘règles expressément adoptées par les Etats’/‘une pratique commune des nations, acceptée par elles comme loi’). As previously observed, the function of these end limbs was ostensibly to set out the nature, in fact to confirm the inclusion of the category of rules concerned among the applicable legal norms. This is why they were in each clause linked to the corresponding head part by a locution spelling out their suasive function (‘comme manifestation de’/‘comme attestation d[e]’). It is because conventional law is made up of ‘rules expressly adopted by the States’ that it has its place in the list; likewise, it is because it consists of a ‘common practice between nations accepted by them as law’ that custom has to figure on it. This is what the two copulative locutions were supposed to express in Descamps’ Proposal, as has been amply argued earlier on.

However, by labelling the two locutions as he did, Descamps – not quite a Descartes in spite of his philosophical leaningsFootnote ⁶⁹ – had let another snag slip into his Proposal. The two locutions obviously were interchangeable in his mind, and this is how they were understood by his colleagues. In the English translation they were indeed both rendered by the simple participle ‘being’. This is what Descamps had actually meant to say, and he could easily have uttered it by using twice some low-profile expression such as ‘en tant que’ or ‘comme étant’. Instead, he preferred the more highbrow ‘comme manifestation de’ in the first clause; and as he probably had been taught in his school-days (as we all were) not to repeat the same expression at too short an interval, he decided to match it with what seemed an equivalent, ‘comme attestation d’une’ in the second clause. In fact, despite a superficial impression, the two locutions run as it were in opposite directions: ‘manifesting’ something is not the same as ‘attesting’ or ‘bearing witness’ to something. A correct equivalent would have been a phrase like ‘comme expression d’une’. But nobody noticed the discordance (which incidentally shows how little attention these first two clauses received); and, far from being checked in the subsequent revisions, it was even aggravated and introduced into the English version. Descamps’ initial blunder marks thus the remote origin of the awkward formulation of the clause on customary law, which is the theme of the present note. This evolution can be neatly retraced in its successive rewordings, as we shall do in the next section, before turning to the parallel development of the clause on conventions.

10. Tidying up the clause on customary law

The English text, which at first had stayed in the shadow of the French original, obtained pride of place in this process, the prime movers of which were indeed the Anglo-Americans in the Committee. Elihu Root – the only one of the company not to know French – was the driving force, with Lord Phillimore on the back-seat, and the eloquent James Brown Scott in the side-car presumably doing the main work. Root and Phillimore had never met before, but they were immediately in unison,Footnote ⁷⁰ and somewhat at loggerheads with the President. As will be remembered, Root had undertaken to bring Descamps’ ‘legal conscience’ in the third clause under control, to make it more predictable for potential litigant states. By the same token, however, the two accomplices wrought several changes in the other clauses, and these were in turn to affect the French text. Here was indeed the point where the first two clauses of Descamps’ exploratory document began for their part to be taken ‘seriously’ as elements of a proper draft article, which brought with it the need to harmonize the two versions.

Nowhere else was this need more palpable than in the clause on customary law. Its English translation was pretty approximate, though rather felicitous inasmuch as it rendered the connecting locution ‘comme attestation de’ simply by the participle ‘being’. Had this been maintained (or even improved to ‘as being’), and moreover transferred to the French text by some expression like ‘comme étant’ or ‘en tant que’ or ‘en sa qualité de’, it would have saved the Committee its subsequent errings and their dubious outcome. Instead, no effort was spared to find an English equivalent for the inadequate ‘attestation’, which stayed substantially untouched throughout the whole process; it was changed to ‘preuve’ under the influence of the English ‘evidence’ only at the very end, after the Draft Statute had been handed on to the Council of the League of Nations. Before getting there, however, we have to retrace the evolution of the clause in the Committee.

A first tentative change was effected by Root on 3 July in his response to Hagerup’s conciliating suggestion of the day before.Footnote ⁷¹ While the main target was clause 3, the clause on custom was also slightly modified so as to read:

The participle ‘recognized’, rather than duplicating the ‘accepted’ figuring later in the sentence, was very probably added to get closer to the French ‘comme attestation d’une pratique’. If so, however, it clearly missed the mark (which itself was of course beside the real point). The term recurred a fortnight later as a substantive, making the clause read ‘international custom, being the recognition of a general practice, accepted as law’,Footnote ⁷³ which was a trifle less inadequate. In any event, this course was not pursued any further.

More important and decisive was another proposition presented the same day as the Root proposal seemingly together by Descamps and Phillimore, with amendments by Ricci-Busatti; it read as follows with regard to the second clause:

The words ‘among said States’ (‘des dits Etats’), instead of the former less determinate ‘between nations’ (‘des nations’), were certainly of Ricci-Busatti’s make; they went in fact well beyond a simple change in wording. As noted above, the Italian positivist regarded custom strictly as a tacit agreement, and this is what his amendment wanted to convey, with a clear reference to the first clause, which in the version redrafted by Root mentioned not just ‘rules expressly adopted by the States’ as in Descamps’ initial Proposal, but rules specifically adopted ‘by the States which are parties to a dispute’.Footnote ⁷⁵ Both clauses would thereby have become even more perfect twins than in Descamps’ Proposal, but, as we have seen, Ricci-Busatti was not followed on this point by his colleagues. Yet his amendment, though abortive, remains at least indirectly significant with respect to the Committee’s conception of custom.

By contrast, the words ‘as evidence of common practice’ were to prove in part a lasting presence in the clause on custom. They aimed at replacing Root’s ‘being recognized practice’ in order to match Descamps’ ‘comme attestation d’une pratique commune’. The crucially important adjective ‘commune’ was thereby made to appear (at least temporarily) in the English version. Much less fortunately, though, Descamps’ misguided ‘comme attestation de’ was even invigorated by ‘as evidence of’. Not only was the basic idea of ‘attesting’, ‘testifying’ or ‘witnessing to’ maintained, but it even gained some additional drive, the ball flying all the further the wrong way. Instead of Descamps’ slip being corrected by something like ‘comme traduisant une’, which would at once have impressed the right turn on the French text, the latter’s bend sinister was thus inflicted on the English translation.

At the same time the English wording added something to the initial phrase by altering the relationship of its terms. Descamps’ ‘attestation’ had simply meant to equate the two components of the clause, and this equivalence was perfectly captured by the English participle ‘being’ (which came in fact nearer to Descamps’ true intention than his own deceptive wording). In his original understanding customary law simply was ‘a common practice between nations, accepted by them as law’, whereas making the one ‘evidence’ of the other amounts to drawing apart the terms of the equation, which boils down to a merely contingent relationship between them. This was to allow in its time for a renewed reading of the clause, which shall be examined later on.

For the time being it is enough to note that it was at this juncture that the fateful expression ‘as evidence of’ made its entry – and was to stay in place unflinchingly until the final adoption of the Statute. Who was its precise author is not clear from the Procès-Verbaux, but it probably originated in the Anglo-American camp, under the possible impression of two celebrated, by then relatively recent cases.Footnote ⁷⁶ The semantic alteration caused by the new locution was at any rate too inconspicuous to be noted by anyone in the Committee. The rendering of the French ‘attestation’ by ‘evidence’ was apparently regarded as satisfactory, and it was certainly taken up as such by Root and Phillimore in their reformulated draft of 14 July, which ran as follows:

In the wake of the former amendment, this wording clang even more to Descamps’ original clause. By then the text, hitherto considered in isolation, had been integrated as Article 31 into a comprehensive overall draft of the future Statute. This Root–Phillimore plan, as it was dubbed, was superseded on 19 July by a Draft Scheme prepared by the Drafting Committee, in which the second clause of Article 31 became:

This version marks the first appearance of the adjective ‘general’, replacing ‘common’ in the English text, and the final disappearance of the (‘practicing’ and ‘accepting’) ‘nations’ in both texts. A possible reason for this change may have been the difficulty of rendering Descamps’ ‘une pratique commune des nations acceptée par elles comme loi’ in palatable English, which had led to the clumsy phrase in the Root–Phillimore version of 14 July; hence probably the decision to simplify it at one stroke by obliterating the ‘nations’ and all that went with them by the single adjective ‘general’, the authors of the practice being thus left to recede into a sort of legal limbo.Footnote ⁷⁹ The French text, which up to then had remained basically unchanged despite Ricci-Busatti’s attempted remodelling, was also brought to slim down a little, the relative clause at its end shrinking to a three words’ apposition:

Both texts were adopted at second reading on 22 July, with the remaining disparities, as Article 35 of the final Draft Scheme of the Statute.Footnote ⁸¹

In the meantime, Albert de Lapradelle as rapporteur had hurriedly prepared a report on the whole draft in the shape of a running commentary, which he read out to the Committee on 23 and 24 July. His comment on Article 35 is very short: it starts with a doff to Article 7 of the Prize Court Convention of 1907, before succinctly paraphrasing the four clauses. What he says on the second clause deserves to be quoted in all its brevity, first in the original version: ‘la coutume internationale dont la continuité atteste une pratique commune’, which is rendered in English as: ‘international custom in so far as its continuity proves a common usage’.Footnote ⁸² It would be hard in so few words to be more beside the question. Descamps’ queer logic is pushed even further and reaches a new climax: the ‘continuity’ is indeed said to attach to customary law as such, which is in turn supposed to ‘prove’ the ‘common usage’. This is truly putting things head to tail since, if anything, it is of course the continuous practice which, by its very consistency, is deemed to bring forth and to ‘prove’ the various rules making up the whole bulk of customary law.Footnote ⁸³ But Lapradelle’s well-sounding phraseology seems to have charmed asleep whatever critical sense and juristic sensibility may have survived after a five-weeks march through more than 30 sessions of exacting work, not to speak of the growing haste in which the Committee found itself during the last days of its meetings.

This was not quite the end of the drafting process, but it stopped there for the Advisory Committee. Before moving further to its last stage in the League of Nations, we have, true to our approach, briefly to look at what happened to the companion clause on conventional law.

11. Tidying up the clause on conventional law

Whereas, regarding the clause on custom, the Committee largely focussed on attuning its French and English versions, this intent was less prominent though present also with respect to the first clause. But the main concern here was to confer some reasonable function on its end part – ‘comme manifestation de règles expressément adoptées par les Etats’ – which had become a redundant platitude once it had served its hortatory purpose in the Committee. This is what appears from the clause’s progress through the successive stations of its pilgrimage.

In the amended text submitted by Root (with Phillimore looking over his shoulders) on 3 July, both these intents had some part. On the one hand, Descamps’ stilted ‘comme manifestation de’ was felicitously replaced by ‘comme constituant des’,Footnote ⁸⁴ which was a perfect pendant to the rather plain though quite adequate ‘being’ in the English translation. On the other hand, instead of just stating in the abstract that conventions consist of ‘rules expressly adopted by the States’, the revised text expected them to be adopted ‘by the States which are parties to a dispute’ (‘par les Etats parties au litige’, as the better French version runs).Footnote ⁸⁵ Although the judges do not need to be reminded that only treaties in force between the litigants are applicable eo nomine, the clause thereby gained at least a semblance of function (pleonasms being current and most normal in legal language, as observed on another point, tongue in cheek, by Rafael Altamira, the Spanish member of the Committee).Footnote ⁸⁶

The functional logic thus conferred on the clause was emphasized in the Descamps–Phillimore proposal of the same day, as amended by Ricci-Busatti, which basically was a variation on the Root text. While Root had maintained Descamps’ collective designation ‘le droit international conventionnnel’, with its English counterpart ‘conventional international law’, the amended text adopted a more ‘nominalist’ caption breaking it down to ‘les conventions internationales’ and ‘international conventions’, respectively. In addition, the participle ‘being’ was replaced by ‘as constituting’, in order to bring it into conformity with the French ‘comme constituant’.Footnote ⁸⁷ But ere long the English ‘as constituting’ was amended to ‘forming the rules’ in the Root–Phillimore plan of 14 July, which kept the initial ‘being’ still faintly ringing.Footnote ⁸⁸ Five days later this was again altered to ‘establishing rules’ in the Draft Scheme submitted by the Drafting Committee, probably under the influence of the parallel change wrought in the French version.Footnote ⁸⁹ The latter had in fact been rephrased to ‘posant des règles’, in an attempt to match the English ‘forming the rules’, only to be altered, at the end of this chassé-croisé, to ‘établissant des règles’.Footnote ⁹⁰ It was also at this stage that Descamps’ ‘règles expressément adoptées’ was replaced by ‘règles expressément reconnues’ and ‘rules expressly recognized’.Footnote ⁹¹

As we know, this was to be the definitive wording in both languages. It was about the best that could be done in doctoring up the clause to make it operational instead of just descriptive. As it stands, it does indeed provide that treaties have to be taken into consideration by the Court (only) inasmuch as they are in force and prima facie relevant between the parties to the dispute. This goes without saying, but in saying it the clause was ‘saved’ from sheer redundance by fulfilling, at least ostensibly, a true function.

And yet, this bricolage was bound to remain an unsatisfactory half-way house insofar as it maintained the former suasive intent along with the new practical finality. The ‘rules expressly recognized’ were indeed wholly part of the initial purpose of the Descamps Proposal, implicitly echoed by the ‘practice accepted as law’ of the second clause. Not only has it become superfluous in the final article, but it presents the drawback of not being quite foolproof; for it could be read as distinguishing between conventions that do, and others that do not establish ‘expressly recognized rules’. However absurd such an almost scholastic quibbling may seem, it did find some adepts.Footnote ⁹² If at all, the clause should have been radically redrafted and simplified into something like ‘conventions in force between the parties’.Footnote ⁹³

12. The final harmonizing in the League of Nations

Let us now return to the works of the Advisory Committee where we left them on 24 July. Lapradelle’s report was voted upon as soon as he had finished reading it out, and it was unanimously adopted by the Committee.Footnote ⁹⁴ Moreover, three propositions moved by Root were carried, the third one stating ‘that the Committee have no wish to advocate or argue its conclusions’.Footnote ⁹⁵ The Draft Scheme was thereby irrevocably released from the orbit of the Committee. In fact, though, several members were to have some share in its further elaboration within the League of Nations, now as representatives of their governments, pending the final adoption of the Statute by the end of the year.

First it was up to the Council of the League to examine the draft, in August and October 1920. As a result, most importantly, the principle of compulsory jurisdiction was abandoned, only later to be revived in some measure by the optional clause of Article 36 of the final Statute. This had no direct incidence on the latter’s Article 38, which was still numbered Article 35 in the draft. But it did open the way to a renewed discussion of its third clause during the subsequent deliberations in the First Assembly of the League of Nations in November and December 1920. The Assembly put the Draft Statute on the agenda of its Third Committee, which included five members of the former Advisory Committee.Footnote ⁹⁶ These were in turn all again members of the Sub-Committee that was to impart the finishing touch to the draft. Here it was that the discussion on the third clause with its naturalistic flavour was shortly resumed; but instead of a change in the clause it produced the additional second paragraph empowering the Court ‘to decide a case ex aequo et bono’, at least ‘if the parties agree thereto’. This was the only important amendment, although its practical consequences were to prove rather insignificant.

The initial portion of the draft article, forming henceforth its first paragraph, basically remained in place as it had been devised at The Hague. Of its first two clauses, the one on conventional law was left untouched, whereas the clause on customary law received its definitive shape only at this ultimate stage. There is no record of the debates leading to that final wording; the changes were apparently considered merely cosmetic (although in fact they were by no means negligible), aiming in the main as they did at eliminating the divergences between the two versions of the clause (both having by then become equally official texts). This fine-tuning was immaterial for the English version, in which only the relative clause at the end was in its turn reduced to the apposition ‘accepted as law’, as it had been in the French text before. Conversely, however, the French ‘acceptée comme loi’ became ‘acceptée comme étant le droit’. This change caused a new divergence, but it is easily understood since the term ‘loi’ is in French (as generally in civil law conceptions) closely linked to a national lawgiver; law on the international plane would normally be called ‘droit international’ or ‘droit des gens’, just as custom, whether international or domestic, would be termed ‘droit coutumier’. This change goes, however, at the price of blurring Descamps’ idea of custom as the common law among nations.Footnote ⁹⁷ The same is of course true of the change from ‘pratique commune’ to ‘pratique générale’, now in conformity with the English text as revised beforehand in the Advisory Committee.Footnote ⁹⁸ Here was also the moment when the English ‘evidence’ leapt over as ‘preuve’ to the French text, replacing Descamps’ ‘attestation’; whereby his initial blunder, progressively developed in the Advisory Committee, had run its course full circle.

This was the end of the drafting process. A general glance at Article 38, as it resulted from the vote on the Statute in the League of Nations Assembly on 13 and signature on 16 December 1920, is now in place before having a last look at the clause on custom in its final formulation.

13. Reading the clause creatively

The lineal descent of Article 38 from the Descamps Proposal remains clearly discernible, at least for what now forms its first paragraph; and even the additional second paragraph allowing for an equitable settlement could be seen as an offspring of the Baron’s ‘legal conscience’ witnessing to ‘objective justice’ after it was ousted from the third clause by Messrs Root and Phillimore.Footnote ⁹⁹ Their redrafting of that clause was in fact the most incisive change brought about in the initial project. As a result, Descamps’ ‘Grotian’ dualism, setting the two paired types of positive rules, custom and treaties, against a rival category of ‘objective’ legal principles, was broken up. Instead of being coupled with conventions, customary law now appears sandwiched between just two other kinds of ‘recognized’ positive rules, judicial decisions catering to the three of them, together with the authoritative teachings of eminent publicists. Yet these various changes do not obliterate the genetic link with Descamps’ sketch.

While that lineage is distinctly traceable, however, we should not lose sight of the generic difference between the two texts. Descamps’ Proposal was not a draft article but only a basis of discussion, and the comments adorning each clause were merely temporary incentives advocating their adoption. Their purpose being fulfilled, they were either to disappear or to be reformulated. In fact the four of them were maintained in one form or another. The first clause was rendered at least ostensibly operative by a slight specification at its end. The third clause was from the start drastically readjusted by positively ascertainable ‘general principles’ taking the place of a rather elusive ‘legal conscience of civilized peoples’ spelling out a would-be objective law. Much less was needed in the fourth clause to make it viable as a legal provision: the subsidiary position of case law was already endorsed in the Descamps Proposal, and the same was of course true of the teachings of publicists that were added to it. In all three cases the Baron’s suasive comments had thus given way to a properly normative function.

Nothing of the sort appears at first sight to have happened with Descamps’ second clause. The alterations it underwent in July and December 1920 were in the main directed at cutting it down to size and harmonizing its two versions linguistically. In the initial phase of its drafting it was mainly the English text that was adapted to the French original, in the final stage the movement was reversed. While several terms were changed and others omitted in the process, all these piecemeal amendments do not seem to have aimed at altering the basic texture of the clause and impressing it with a fundamentally new thrust. The step from incentive description to normative prescription was not taken. Instead of being simply dropped as probably anticipated by Descamps, his motivating comment just stayed in place like an erratic block, slightly rechiseled but otherwise unchanged. Apparently, its original function went unheeded, possibly owing to a portion of vis inertiae coupled with a quantum of textual fetishism, not uncommon among jurists.

But jurists also have a knack for infusing legal texts with unexpected meanings by reading them somewhat creatively, following the interpretative maxim of effectiveness (‘effet utile’: ut res magis valeat quam pereat). In spite of its puzzling formulation, it is indeed difficult to accept that the clause be merely descriptive and devoid of any practical function beyond Descamps’ initial intent.

What this creative reading has become in the course of the century since the making of the Statute is well known: it is none else than the one adopted by the ILC almost as a matter of course in its recent work on the identification of customary international law. As has been observed at the outset of this note, the core part of the clause – ‘a general practice accepted as law’ – is widely perceived as a working definition of customary law. This quasi-official status of the formula as a sesame opening up the whole field of customary international law hinges on the presumption that it embodies the two-elements conception. The question therefore arises whether it was already understood in that way by its authors in 1920 (which is surely the general assumption); or to what degree it is not rather the result of a subsequent reading that has itself become customary with plenty of paper practice and opinio doctorum to its credit.

There was certainly no such implication in Descamps’ initial formula. His above-mentioned explanations amply show that it was entirely geared to making sure that custom be incorporated, along with conventions, in the positive law to be applied by the future court. That is what mattered in its description as ‘a common practice among nations accepted by them as law’. The formula boils down to stating the ground of validity of customary law as a whole and its relevance as a consolidated, undisputed body of rules, with an implicit reference to the first clause, which symmetrically asserts the relevance of the whole bulk of conventional law on account of its being ‘expressly adopted’. The thrust of both clauses was rhetorical rather than analytical: they meant to convince rather than to teach. They were in fact devoid of any doctrinal implications. Any resemblance between the second clause and some particular conception of customary law is purely accidental. Such is the case especially of the two-elements doctrine which had been consecrated a couple of decades earlier in private law by François Gény; the more so as the opinio iuris mania had not yet invaded international law circles by 1920; it was to become virulent only after the next world war. Not that the two-elements construct was incompatible with the clause: again, as pointed out already, the ‘material’ element could easily be made out in the ‘common practice’; and being ‘accepted as law’ can quite as well stand for its ‘subjective’ counterpart. But this is of no avail with regard to what the Baron really intended. Neither did he preach the two-elements gospel, nor did he hand out a corresponding catechism to the future judges instructing them as to the formation and identification of customary rules or, for that matter, as to distinguishing between genuine customary rules and simple usages. All this was beyond the scope of his clause which merely presented custom in a nutshell as the common law of nations. The ‘common practice accepted as law’ formed, if anything, one single, undifferentiated ‘element’ in the general fabric of international law. Whereas Gény’s two-elements test aimed at the judicial ascertainment of individual customary rules of doubtful legal status, Descamps’ formula posited customary law en bloc as an ‘accepted’, established normative corpus to be drawn upon by the Court in the absence of more specific conventional stipulations.

So much for the original Descamps formula. But what about its final avatar issuing five months later from the drafting process as Article 38 of the Statute? While the amendments were on the face of it essentially in wording, as shown above, they could still conceivably have been meant, beneath the surface, to favour the now dominant reading of the clause as a definition encapsulating the formative process of customary law, and hence as a kind of touchstone (or at least a rule of thumb) for testing the legal soundness of individual customary rules. It could thus be read in parallel with the first clause, which in its revised form may be seen as fulfilling a similar role for conventions by specifying ex abundanti cautela that, in order to apply, they have to be in force between the parties to the dispute.

The initial complementarity between the two clauses remains indeed perceptible, although its systemic implications are gone. The ‘general practice accepted as law’ of the one corresponds to the ‘rules expressly recognized’ of the other. While the ‘international custom’ of the second clause basically still represents the corpus of customary law as a compact whole, it could therefore also be seen as the aggregate set of individual customary rules, by analogy with the sum total of ‘international conventions’ of the first clause. Rather than comprehensively denoting customary law as the common law of nations, the ‘general practice’ could thus also apply to single customary rules. This move away from the original meaning in the Descamps Proposal may have been favoured by the Baron’s ‘pratique commune des nations’ becoming ‘a general practice’ ostensibly cut off from its roots in the community of nations. There is nothing stringent about this alternative reading – which in any event does not preclude the original one – but it is in fact how the clause is now mostly understood, not without a shade of ambiguity stealing over it.

Could it be that this ‘touchstone’ reading of the clause, which gives it the appearance of a heuristic and hermeneutic device, was at least subliminally present already among the members of the Advisory Committee? Their misreading the clause (with the rest of the Descamps Proposal) as a genuine draft article, could have induced them to see it in this light; which in turn might explain their substituting the term ‘evidence’ for the Baron’s initial ‘attestation’. That term was up to now left aside in the present discussion, but we have met with it earlier on, noting the inconspicuous but significant shift it brought about in the inner economy of the clause.Footnote ¹⁰⁰ It does indeed suggest of itself a heuristic function, just as its French equivalent ‘preuve’. The only hitch with this reading is – once more – that it works the wrong way: instead of a general practice being evidence of a customary rule, it is custom that is declared evidence of a general practice; which is not really quod est demonstrandum. It is almost insulting to charge the eminent men in the Advisory Committee with such a gross paralogy (although that is precisely what Lapradelle stumbled into in his final report on the Draft Statute, with nobody objecting). One should rather charitably think of them as wholly bent on harmonizing both versions of the clause, to the point of losing sight of what it (erroneously) said – like a team steadily rowing away but with no helmsman to steer the boat.Footnote ¹⁰¹ Had they really thought of rendering the clause operational as a divining rod for detecting customary rules, they could not but have reversed its polarity by changing the locution ‘as evidence of’ to something like ‘as evidenced by’ or ‘as evinced by’. Instead, they let its birth-defect untouched, perpetuating thereby, nay enhancing Descamps’ initial inadvertence.

We can therefore safely conclude that the clause had not in the end obtained a truly new meaning removed from the original one in the Descamps Proposal. It did not (chiming with Prospero’s airy sprite) ‘suffer a sea-change into something rich and strange’. The changes it did undergo, in the Advisory Committee or later in the League of Nations, were only in formulation, without impinging on its substance. The Baron’s initial suasive and descriptive intent was not superseded by a properly normative and operative function (be it heuristic or otherwise). The clause thus remained somehow stuck in the air, upside down, with no apparent purpose, apart from stating the obvious and stating it the wrong way around.

This, then, was the presumable situation in 1920, when the Statute came to life and made its entry on the international scene. It was only a matter of time, however, for the customary law clause to be instilled with a seemingly practical meaning, like the other clauses of Article 38. Such an outcome was obviously favoured by the fact that the whole of Article 38 was from the start generally perceived not just as a direction to the Court but as a repository of the sources of international law in general. This reading appeared all the more natural as the provision had in fact been dealt with in such terms by some members of the Advisory Committee in 1920.Footnote ¹⁰² Its second clause, moreover, seemed by its very texture to be a definition of customary law. From that point there was only a small step to identifying its core part with the two-elements conception. This step was taken at the latest after the Second World War.Footnote ¹⁰³

Yet it was precisely this reading of the clause that rendered it problematic. While it nicely tied in with its core part, there are still the words ‘as evidence of’. These are the real crux of the matter, giving rise to the various interpretative gesticulations mentioned at the beginning of this note. Usually, of course, the difficulty is either wilfully overlooked (with possibly a discreet hiccup) or it is set aside as a casual slip of the pen. It is not recognized for what it really was, an initial slip in the Baron’s mind worsened by subsequent, carefully misguided amendments. These errings are obviously due in the first place to Descamps’ undetected blunder. More fundamentally they are due to his Proposal being taken for a proper draft text instead just for an informal basis of discussion with a hortatory intent.

It may be asked why, then, instead of himself making this clear, Descamps let things take the erratic course just described. This can be plausibly explained by his paradoxically isolated, at times almost embattled position in the very Committee he was chairing.Footnote ¹⁰⁴ He was at once fully caught in the dynamics unleashed by the passionate clash on the third clause of his Proposal and was probably all too glad that at least the first two clauses went down smoothly, without eliciting any notable objections. In fact, he took hardly any part in their elaboration, the main initiative having passed to his Anglo-American colleagues.Footnote ¹⁰⁵

Another point has to be remembered in this regard. During its genesis, Article 38 was far from enjoying the eminent position it has (however undeservedly) acquired since as a widely beaming beacon, visible well beyond the (purely conventional) groundwork of the Statute on which it rests. It was not seen by its drafters as strategically pivotal, unlike the provisions relating to the ambit and nature of the Court’s jurisdiction or to the nomination of its judges. It would even seem that some of the members of the Advisory Committee, not only among the Anglo-Saxons, could have done without it.Footnote ¹⁰⁶ Article 38 originated in civil law conceptions, and it was above all Descamps’ baby. Had it not been for his idiosyncratic Proposal with its ‘Grotian’ underpinnings, the point would have been settled in no time with hardly any debate.

These collateral observations may add to the plausibility of the above findings which, based only on internal, essentially textual and contextual evidence, partly gathered from reading between the lines, cannot claim to be more than plausible indeed. But that is probably the risk any archaeological venture is bound to run, even if, instead of excavating spectacular relics from bygone civilizations, it only amounts to reading again and interpreting anew, beyond entrenched preconceptions, some perfectly familiar vestiges of a drafting process rather hurriedly recorded a century ago. The main difficulty resides in capturing the initial purport of the Descamps Proposal and in evaluating the import of the ensuing labours in the Advisory Committee. The attempted reconstitution of the impalpable meanderings of the minds remains at least partly hypothetical. So much is fairly certain, however, that the far-famed formula here revisited was not intended to be a definition of custom as a source of international law and still less an endorsement of the two-elements approach for its identification. And perhaps the closer reading endeavoured here of the genesis of Article 38 as a whole may contribute to lay to rest at least some of the debates that have hitherto needlessly beclouded it. If so, the present archaeological look may also prove to be a fresh look at an old acquaintance, having just crossed the meridian of its centenary.