OPINION OF MR ADVOCATE GENERAL CAPOTORTI

DELIVERED ON 13 JULY 1982 ( 1 )

Mr President,

Members of the Court,

The request for a preliminary ruling now before the Court concerns one of the provisions of the EEC Treaty relating to the powers of the Court, namely the third paragraph of Article 177. The Italian Corte Suprema di Cassazione [Supreme Court of Cassation] wishes to ascertain whether that provision lays down an obligation to submit a case to the Court of Justice which precludes the national court from determining whether the question raised is justified or whether, and if so within what limits, it makes that obligation conditional on the prior finding of a reasonable interpretative doubt.

I shall briefly summarize the facts of the case. In September 1974 a large number of Italian textile firms, including CILFIT and Lanificio di Gavardo, brought an action before the Tribunale [District Court], Rome, against the Italian Ministry of Health for the recovery of sums which had been paid — wrongly, in their submission — by way of health inspection levies on imported wools. Those sums had been paid while Law No 30 of 30 January 1968 was in force, which fixed those levies at LIT 700 per quintal of imported wool, although the sum to be paid had been drastically reduced by Law No 1239 of 30 December 1970 to only LIT 70 per quintal.

Having failed at first instance and on appeal, the plaintiffs appealed to the Corte Suprema di Cassazione on the ground, inter alia, that the inspection levy should not have been collected since it was contrary to Regulation (EEC) No 827/68 of the Council on the common organization of the market in certain products listed in Annex II to the Treaty, including the “animal products” referred to in heading 05.15 of the Common Customs Tariff. The Ministry of Health's counterargument was that wool was not included in Annex II to the EEC Treaty and did not therefore fall within the scope of the aforesaid regulation.

According to the Ministry, the scope of Regulation No 827/68 was quite unequivocal on that point and therefore precluded any need to make a reference to the Court of Justice for a preliminary ruling.

In those circumstances, the Corte Suprema di Cassazione stayed the proceedings by order of 27 May 1981 and referred the aforementioned question to the Court for a preliminary ruling.

2. 

It is well known that, under the third paragraph of Article 177, where a question concerning the interpretation of the Treaty, the validity and interpretation of measures adopted by the Community institutions or the interpretation of the statutes of bodies established by the Council “is raised in a case pending before a court or tribunal of a Member State against whose decisions there is no judicial remedy under national law, that court or tribunal shall bring the matter before the Court of Justice”. According to the Corte Suprema di Cassazione, which has sought to clarify a lively controversy in progress amongst legal writers and discernible in decisions of the national courts, the aforementioned provision is open to two different interpretations. It may be regarded as imposing a strict obligation on courts of last instance in the Member States to refer questions to the Court of Justice for a preliminary ruling without allowing them any scope to consider such questions or to determine the extent to which they are capable of being answered in more than one way. Alternatively, it may be argued that the third paragraph of Article 177 permits national courts to carry out a preliminary examination in order to determine, within their discretion, whether a reasonable interpretative doubt exists in a specific case; that would remove any need for a reference in all cases in which no such doubt is discerned.

It should be pointed out at this stage, before the problem raised is considered, that the Corte Suprema di Cassazione, in spite of the broad terms in which its question is couched, has taken into account only references for a preliminary ruling on the interpretation of a provision of the Treaty or of a measure of secondary legislation, thereby disregarding the issues raised by references for a ruling on the validity of Community measures. That seems to be borne out by the fact that the Corte di Cassazione, in the final part of the question, expressly mentions the argument based on the existence of a “reasonable interpretative doubt”, and also by the fact that the action brought before that court is concerned precisely with the interpretation of provisions of secondary Community legislation (as is apparent from the order making the reference).

3. 

It is appropriate to endeavour to establish in the first place whether any guidance may be derived from the wording of Article 177 which may help to resolve the issue raised.

The use in the third, as in the second, paragraph of that article of the word “question” to indicate the subject-matter of the preliminary ruling procedure has been regarded by certain writers as confirmation of the argument that the scope of the obligation imposed by the third paragraph is restricted solely to cases in which genuine problems of interpretation arise, that is to say, where there are difficulties of interpretation. Furthermore, since the second and third paragraphs refer to “questions ... raised” in a case pending before a national court or tribunal, it has been argued that the provision was thus intended to cover cases in which one of the parties to the proceedings (which might include an officer of the State in legal systems which provide for such intervention) takes the initiative in raising before the court the existence of a problem involving the interpretation of Community law. It is commonly agreed however that a preliminary question may also be formulated by a court of its own motion, as the Court recently stated in its judgment of 16 June 1981 in Case 126/80 (Salonia v Poidomani and Baglieri, née Giglio [1981] ECR 1563, in paragraph 7 of the decision). When that happens, it is contended that the power of discretion vested in the court, even a court of last instance, cannot be doubted, and the conclusion is drawn that it would be illogical to recognize the court's power of discretion in relation to questions raised by it of its own motion and to deny it that power in the case of questions raised by the parties.

In my opinion, that reasoning leads to unsound results. I would observe in that connection that the word “question” — or rather the expression “such a question” — is used in Article 177 in relation to the three areas (corresponding to subparagraphs a, b and c of the first paragraph) in which the Court is empowered to give a preliminary ruling. In the second paragraph the words “question/questione” and “point/punto” are used as synonyms (at least in the French and Italian versions of the Treaty respectively ( 2 )). Therefore I believe it is inappropriate to give a slant to the word “question” by adding to it the idea that it involves a genuine doubt, a difficulty or some element of choice; it is natural that, with regard to certain aspects of a dispute, a decision must be taken “on the question” and a ruling on interpretation in always intended to dispose of an objective state of uncertainty.

As regards the inferences which, it is argued, may be drawn from the requirement that a question must be “raised”. I believe that in the aforesaid Salonia judgment that word was rightly interpreted as referring both to the parties and to the court. It is incongruous to regard the wording of the article as being consistent only with a situation in which the initiative is taken by the parties, whilst recognizing that the court is entitled to request a preliminary ruling of its own motion. In all cases it is Article 177 which allows the parties and the court to submit the question. If that is so, the obligation imposed by the third paragraph of that article likewise applies in all cases and not merely in those in which the initiative is taken by the parties. Moreover, there is no reason why the court's power to submit a question for a preliminary ruling of its own motion should be confused with its discretion to determine whether it is appropriate to submit such a question. That broad discretion is undeniably vested in courts other than those of last instance, whilst in the case of the latter it is quite reasonable that their powers should be confined to determining whether a preliminary ruling is necessary to enable them to give judgment and that whenever that necessity is recognized a reference to the Court of Justice should be mandatory.

In conclusion, the only unequivocal indication which may be derived from the text of Article 177 is the difference between the provisions of the second paragraph and those of the third. The courts referred to in the second paragraph may request the Court of Justice to give a ruling on a question concerning the interpretation of Community law, whereas those specified in the third paragraph must do so. I shall reconsider that simple fact in detail later on, after the problem has been reduced to its essential components.

4. 

As far as general principles are concerned, reference has frequently been made, whenever the problem in question has been considered, to the theory of the acte clair, the concise meaning of which is that if a provision is unequivocal there is no need to interpret it.

That theory has emerged within the French legal system, which entrusts the interpretation of international treaties exclusively to the executive (in particular, to the Ministry for Foreign Affairs) and merely allows the courts to apply them. In that context, in order to restrict the rôle of the executive and its interference with judicial activity, the courts developed the aforementioned theory, thus retaining the power to determine whether or not genuine difficulties of interpretation exist and thereby recovering a broad margin of discretion. Subsequently, the French Conseil d'Etat — and to a much lesser extent the Cour de Cassation — took the view that they could use that theory to curtail the scope of the obligation laid down by the third paragraph of Article 177 of the EEC Treaty.

That theory was echoed in the opinion delivered by Mr Advocate General Lagrange in connection with Joined Cases 28 to 30/62 (Da Costa and Schaake v Nederlandse Belastingadministratie [1963] ECR 40, at p. 45). There it was stated inter alia that “if the provision is perfectly clear, there is no longer any need for interpretation but only for application, which belongs to the jurisdiction of the national court whose very task it is to apply the law”. However, it would be wrong to lift those remarks out of their context and use them as an argument in support of the view that a national court is empowered to determine whether a preliminary question raised by the parties is justified. In fact it is clear from the context of that opinion that the sole purpose which my illustrious predecessor had in mind was to demonstrate that a fresh interpretation by the Court of Justice was superfluous if the same question had already been resolved by the Court in a previous decision. Moreover, that approach, that is to say recognition of the “authority of the interpretation” provided by the Court, was upheld by the judgment in the Da Costa and Schaake case ([1963] ECR 31), which I shall refer to again in due course.

In my opinion, the theory of the acte clair is of no assistance for the solution of the problem under consideration. If its origin and function are considered, it is easy to see that it was intended to rectify a situation obtaining in a specific Member State which cannot be compared with the situation under discussion. The distinction, with regard to provisions of international treaties, between the application of such provisions, which is a matter for the courts and the interpretation thereof, which is the prerogative of the Ministry for Foreign Affairs, is made in France but not in other Member States. Moreover, the claim for wider powers for the judiciary in relation to certain prerogatives of the executive is quite distinct from the division between the tasks of interpretation entrusted to the national courts of last instance, on the one hand, and those entrusted to the Court of Justice of the European Communities, on the other.

Secondly, the basic concept of the acte clair theory does not seem to stand up to closer scrutiny. Before a provision can be applied to a specific case, it is always necessary, from a logical and practical point of view, to determine its meaning and scope, failing which it is impossible to establish whether it is applicable to the case in question or to infer from its terms all the implications for that case. It may tentatively be stated that when a provision is applied its interpretation and application are interwoven and merge, but it is inconceivable for a provision to be applied without there being any need to interpret it, unless the meaning of the word “interpretation” is distorted in such a way as to suggest that some difficulty is necessarily involved. In the final analysis, the oft-repeated latin maxim “in claris not fit interpretatio” should be abandoned, since it is through the interpretation of a provision that it is possible to ascertain whether its meaning is clear or obscure. Those considerations carry even more weight in a system whose provisions all exhibit the technical difficulty of being drafted in several languages and the inherent difficulty that they affect a state of affairs which is already governed by ten national legal systems.

Finally, the evidence of the facts should not be ignored. The facts show that the acte clair theory, implemented in connection with Article 177, has been applied in a manner which I would not hesitate to describe as anomalous. The French Conseil d'Etat, the principal body to have applied that theory, ventured to affirm as early as 1967 that the concept of measures having an effect equivalent to quantitative restrictions on imports, within the meaning of Article 30 of the EEC Treaty — one of the most vexed questions arising in the Treaty, as is borne out by the decisions of the Court of Justice — did not require any interpretation (judgment of 27. 1. 1967, Syndicat National des Importateurs Français en produits Laitiers, Recueil Lebon, 1967, p. 41). Later, the Conseil d'État did not hesitate to interpret the following provisions: Articles 7 and 37 of the EEC Treaty (judgment of 27. 7. 1979, Syndicat National des Fabricants de Spiritueux Consommés à l'Eau, Recueil Lebon, 1979, p. 335); Article 113 of the EEC Treaty and Council Decision No 72/455 (judgment of 12. 10. 1979, Syndicat des Importateurs de Vêtements et Produits Artisanaux, Recueil Lebon, 1979, p. 373); Regulation No 950/68 of the Council Regulations Nos 3321/75 and 1541/76 of the Commission (judgment of 2. 10. 1981, Groupement d'Intérêt Économique Vipal, Recueil Dalloz-Sirey, 1982, Jurisprudence, p. 209); and Articles 34 and 37 of the Euratom Treaty (judgment of 23. 12. 1981, Commune de Thionville, Recueil Lebon, 1981, p. 484). Mention should also be made, in particular, of the judgment of 22 December 1978 in the Cohn-Bendit case (Recueil Lebon, 1978, p. 524), in which the Conseil d'État, interpreting Article 189 of the EEC Treaty, ruled out any possibility that directives might have direct effect (the case concerned the question whether a private individual might rely on a directive relating to freedom of movement for persons) in sharp contrast to the well-established case-law of the Court.

That shows, in my opinion, that the acte clair theory has far-reaching consequences: its effect is, in substance, to deprive the third paragraph of Article 177 of any meaning. Accordingly, there is no possibility of giving a correct answer on the basis of that unfounded and ambiguous theory to the question submitted by the Corte Suprema di Cassazione.

5. 

Counsel for the Italian Government, which has submitted observations in this case, considers that certain aspects of the Italian system whereby the courts refer questions of constitutional legitimacy to the Corte Constituzionale [Constitutional Court] may be of assistance for the interpretation of Article 177. Under Article 23 (2) of Law No 87 of 11 March 1953, the Italian courts are under an obligation to refer to the Corte Constituzionale all questions of constitutionality which are not “manifestly unfounded”. Those courts thus conduct a preliminary investigation to determine whether a question is justified. However, a mere doubt that a question is not manifestly unfounded is sufficient to give rise to the obligation to refer the matter to the Corte Constituzionale. Should the attitude of national courts be governed by analogous criteria when a case involves Community law?

In my opinion, that question must be answered in the negative for various reasons. Clearly, the review of the constitutionality of laws is quite distinct from the machinery designed to ensure the uniform interpretation of Community law. In order to achieve the first objective, every court acts as a filter to ascertain whether a question is justified, in accordance with the system in force in Italy. In relation to the second objective, however, the Treaties have laid down that, whereas certain courts are entirely at liberty to refer or to refrain from referring questions of interpretation to the Court of Justice, others are under an obligation to do so. Therefore, to authorize the lower courts to act as a filter would be superfluous, whilst to confer such a function on the courts of last instance would detract from the obligation laid down by the last paragraph of Article 177. It would still be necessary to clarify on what basis they would perform that function in view of the fact that the content of the provision is quite clear and that a principle in force in the law of a Member State may not of course be transposed into Community law (even though the principle may be adapted in accordance with the requirements of the Community legal order). In fact, the power and duty of a court to determine whether a question is well founded (or rather whether it is not manifestly unfounded) is based on a specific provision of Italian law, whilst no equivalent provision was inserted in the Treaty of Rome. That omission strikes me as significant —as regards negating the existence of any analogous power or duty in connection with Article 177 — since the Italian procedure for referring matters to the Corte Constituzionale was certainly not unknown at the time when Article 177 was drafted.

6. 

In support of the view which seeks to restrict the scope of the obligation on the highest courts to refer to the Court of Justice questions concerning the interpretation of Community law, certain arguments have been put forward which may be described as arguments based on expediency. It is contended that the effect of such an interpretation of Article 177 is, in the first place, to prevent the Court of Justice from having to contend with an excessive number of references for a preliminary ruling which might compromise its proper functioning and, secondly, to prevent delays or increased costs in national proceedings as a result of the submission of preliminary questions which are unfounded. Moreover, it is argued that the aforesaid view, inasmuch as it recognizes that national courts have a margin of discretion, is the most likely to ensure that the specific rôle of such courts is safeguarded.

That kind of reasoning is not, in my opinion, conclusive. It might be sufficient to object that the meaning of a provision cannot depend on reasons of expediency. However, the reasons which militate in favour of the opposite view should also be borne in mind. The requirement that courts or tribunals of last instance must always refer questions to the Court of Justice for a preliminary ruling is supported by the specific technical and formal characteristics of Community law to which I have already referred (different language versions; novelty of the content and terminology of Community law). It should be added that there are inevitably differences between the methods of interpretation adopted by the Court of Justice and those on which national courts rely, stemming from the differences between the legal spheres in which the former and the latter operate.

7. 

On the few occasions on which the Court has expressed its views on the third paragraph of Article 177, it has reaffirmed the mandatory nature of that provision, without making any allusion whatsoever to the possibility of leaving a margin of discretion to the higher courts. I have the following judgments in mind: 27 March 1963 in Joined Cases 28 to 30/63 Da Costa en Schaake, loc.cit.; 18 February 1964 in Joined Cases 73 and 74/63, Internationale Crediet v Minister van Landbouw en Visserij [1964] ECR 1;15 July 1964 in Case 6/64 Costa v ENEL [1964] ECR 585; 4 February 1965 in Case 20/64 Albatros v Sopéco [1965] ECR 29; 24 May 1977 in Case 107/76 Hoffmann-La Roche v Centrafarm [1977] ECR 957. I would observe that, whilst in the Costa v ENEL judgment the Court confined itself to paraphrasing the provision in question and in the Albatros judgment it merely alluded to “the power or obligation, as the case may be” to have recourse to the procedure laid down by Article 1 77, in the Da Costa en Schaake judgment (confirmed in substance by the Internationale Crediet judgment) it laid emphasis on two points: the distinction “between the obligation imposed by the third paragraph of Article 177 upon national courts or tribunals of last instance and the power granted by the second paragraph of Article 177 to every national court or tribunal to refer to the Court of the Communities a question on the interpretation of the Treaty” and the finding that the last paragraph of Article 177 “unreservedly requires courts or tribunals of a Member State against whose decisions there is no judicial remedy under national law ... to refer to the Court every question of interpretation raised before them”. However, that obligation, according to the Da Costa en Schaake judgment, may be deprived of its purpose and substance if the question raised is “materially identical with a question which has already been the subject of a preliminary ruling in a similar case”. The Court thus recognized only one exception to the obligation laid down by the aforesaid provision, namely the possibility of relying on an earlier preliminary ruling given by the Court relating to the same question.

Of particular interest, in my opinion, is the Hoffmann-La Roche judgment of 24 May 1977, because it gives prominence, in paragraph 5 of the decision, to the objective of Article 177 (“to ensure that Community law is interpreted and applied in a uniform manner in all the Member States”) and to the specific purpose of the third paragraph (“to prevent a body of national case-law not in accord with the rules of Community law from coming into existence in any Member State”). It is those two points which must be taken as a basis for answering the question submitted by the Corte Suprema di Cassazione.

Lastly, I should like to reaffirm the views which I had occasion to express in the opinion which I delivered in the Hoffmann-La Roche case: “since what is involved is the interpretation of a provision, such as that in the third paragraph of Article 177, which is essentially procedural in character, every effort must, in my view, be made to define its scope on the basis of objective and specific criteria which leave the courts which have to apply it with no margin of discretion.” Clearly, acceptance of the idea that the obligation to refer a matter to the Court exists only where a reasonable interpretative doubt has arisen would lead to the introduction of a subjective and uncertain factor and might prevent the procedure in Article 177 from attaining its objective, which is (as I stated in my opinion in that case) to ensure certainty and uniformity in the application of Community law.

8. 

In relation to that objective the second and third paragraphs of Article 177 obviously have different functions. The second paragraph enables national courts other than those of last instance to avail themselves of the Court's cooperation whenever they consider it advisable to entrust the Court with the interpretation of a point of Community law. Consequently, uniformity and certainty of interpretation are only partially achieved: that is to say, to the extent to which national courts decide to take advantage of the opportunity to refer questions to the Court for a preliminary ruling. However, the third paragraph makes such references mandatory and the purpose is clearly to ensure that they take place regularly in the ordinary course of events, since that is the only way in which uniformity and certainty of interpretation at Community level can be achieved entirely. Furthermore, the reason for the difference between the two paragraphs is well known: courts of last instance give final decisions which cannot be amended and which are capable of influencing trends in the lower courts of the same country. In other words, the “hard core” of national case-law consists of judgments delivered by courts of last instance. Clearly, the intention of the authors of the Treaty was to avoid any risk of distortions at that level by entrusting to the Court the main burden of creating a body of case-law on questions concerning the interpretation of provisions of Community law so as to avoid inconsistencies, differences of opinion and the resultant uncertainties.

If that is the rationale of Article 177, it seems undeniable in my view that the third paragraph of that article should be understood in the sense which is most likely to ensure that Community law is uniformly interpreted. That gives rise to four consequences:

(a)

the existence of a question of interpretation must be recognized whenever an aspect of, or an issue in, a case is governed by provisions of Community law (irrespective of the seriousness of the doubts to which that aspect or issue may give rise) and the court of last instance must rule thereon in order to give judgment;

(b)

it makes no difference whether the question is raised by the parties or is discerned by the court and the attitude of the parties (their agreement or disagreement as regards the point at issue) is irrelevant;

(c)

the court of last instance has no discretion to determine whether a question raised by the parties is well founded or whether the point of Community law which is relevant for the purposes of the decision must be assessed by itself or by the Court of Justice;

(d)

the obligation to refer a question to the Court of Justice for a preliminary ruling ceases to exist only when a preliminary ruling has already been given by the Court on the same question, although there is nothing of course to prevent the national court from approaching the Court of Justice once again in order to seek either a different interpretation of the provision of Community law in question or clarification of the interpretation already given.

One of the criticisms levelled against the view which I have put forward consists in the objection that the preliminary ruling procedure is thus conceived as an automatic mechanism, and, to make matters worse, according to the critics, that occurs precisely when it is for the supreme courts, which are usually jealous of their status in their respective judicial systems, to set that procedure in motion. However, the fact should not be ignored that it is in any event for the national court, whether of first or of last instance, to assess the relevance of question, that is to say to establish whether the interpretation of a provision of Community law is really necessary to enable it to give judgment. That type of control is accompanied by a wide margin of discretion and the Court of Justice has always acknowledged that it falls within the exclusive jurisdiction of the national court. The following judgments may be cited: 14 February 1980 in Case 53/79 Office National des Pensions pour Travailleurs Salariés v Damiani [1980] ECR 273; 29 November 1978 in Case 83/78 Pigs Marketing Board v Redmond [1978] ECR 2347; and 30 November 1977 in Case 52/77 Cayrol v Rivoira [1977] ECR 2261.

Finally, I should like to point out that an eminent national court of last instance had recourse to the procedure laid down by Article 177 in a case in which there appeared to be no doubt about the meaning of the rule of Community law to be applied (Article 119 of the EEC Treaty) but no consistent line of reasoning could yet be discerned in the relevant case-law of the Court of Justice. I am referring to the decisions of the House of Lords in Garland v British Rail Engineering Ltd and, in particular, to the order for reference of 19 January 1981 and to the judgment of Lord Diplock delivered on 22 April 1982 [1982] CMLR 179.

9. 

The Commission, which has, as usual, submitted observations in these proceedings, expressed support for the view that the obligation to refer a matter to the Court for a preliminary ruling comes into existence, in the case of courts of last instance, only where there is an interpretative doubt, but then sought to identify a series of objective circumstances which, if they were to arise, would make it impossible to deny the existence of such a doubt and a reference to the Court of Justice would have to be regarded as mandatory. Although that line of reasoning seeks to confine within very narrow limits the national court's discretion and thus to ensure in the vast majority of cases the intervention of the Court of Justice, it is an argument which I cannot share. I consider that that argument only appears to found the obligation on objective criteria (for example, on the existence of a conflict between the courts of first and second instance concerning the interpretation of a rule of Community law) and that it in fact confers a margin of discretion on the national court to determine whether or not the question of Community law which it must decide is well founded. In my opinion, recognition of that discretion clashes with the function of Article 177. A uniform interpretation of Community law by the Court is objectively in the public interest, which may not be subordinated to the existence or otherwise of agreement between the national courts in the previous stages of an action or to the assent or dissent of the parties. It must be borne in mind that Article 177 is capable of providing an interpretation of a provision of Community law which is of use to the entire Community — hence the Court was right to recognize as early as 1963 in the aforesaid Da Costa en Schaake judgment the value of precedents set by itself — but that is precisely the reason why the course of a specific action before a national court cannot affect the scope of the obligation laid down by the third paragraph of Article 177. Apart from that, the Commission seems to have arrived at its position by the wrong path. There is nothing, either in the text or in the function of Article 177, to justify proceeding upon a restrictive interpretation of the obligation laid down by the third paragraph of that article; on the other hand, a strict interpretation does not prevent the obligation from being attenuated in the manner emphasized in the Da Costa en Schaake judgment.

10. 

In conclusion, I am of the opinion that the Court should give the following answer to the question referred to it by the Corte Suprema di Cassazione by order of 27 May 1981 issued in the case of Srl CILFIT and Lanificio Gavardo v Ministry of Health.

The third paragraph of Article 177 of the EEC Treaty must be interpreted as meaning that a court or tribunal of a Member State against whose decisions there is no judicial remedy under national law is under an obligation to seek a ruling from the Court of Justice on the interpretation of primary or secondary Community law whenever it must decide a question of Community law raised either by the parties or by the court of its own motion to enable it to give judgment.

Even where a national court or tribunal considers that the question of Community law which it must decide is free from obscurity and ambiguity, and accordingly entertains no doubts regarding its interpretation, it is still under an obligation to seek a preliminary ruling from the Court of Justice unless the same question has already been the subject-matter of an interpretation by the Court.


( 1 ) Translated from the Italian.

( 2 ) Translator's note: In the English version the word “question” is used in both places.