The principle of legal certainty in the case-law of the European Court of justice : from certainty to trust

The principle of legal certainty in the case-law of theEuropean Court of justice : from certainty to trust

« Much of the uncertainty of law is notan unfortunate accident, it is of immensevalue » (J. Frank)

The principle of legal certainty, inherent to any western legalsystem, appears in many different shapes. At the crossroads ofEuropean legal traditions, the case-law of the Court of Justiceof the European Union1 is testament of that fact. According tothe Court, the principle of legal certainty requires that rulesof law be known, clear, precise, stable, certain andpredictable as regards their effects. How can theserequirements be reconciled with the fact that this principle isprobably the most uncertain, ambiguous and unpredictable of allthe European norms ?

This paper will show that the uncertainties and incoherences ofthe Court’s judgments regarding the general principle of legalcertainty (almost 2.500 so far) are the result of unquestionedpostulates as to what the principle means and can be –at leastpartly- explained by the history of legal ideas.

In a first part, a historical survey shows that from ancientGreek and Roman law until today, with a very significant turnat the 17th century, four ‘logics’ have built up that animatethe modern principle of legal certainty. Two of these logicspropose alternative bases for the principle : the cartesian logicis grounded on the requirement for (absolute) certainty while thefiduciary logic is built on the notion of trust. The two remaininglogics are related to the beneficiaries of the principle : thesubjects of law (subjective logic) or the Prince (political logic).This last logic can be traced back to Antiquity while the threeother logics mostly arose during the 17th century, with thescientific revolution of Descartes and Newton (cartesianlogic), the Lockean theory of the social contract (fiduciary

1 The Court of Justice of the European Union (CJEU) consists of threecourts, being the Court of Justice (hereafter the ECJ), the General Court(hereafter the GC) and the Civil Service Tribunal, which specific case-lawwill not be studied. The CJEU interprets European Union law (EU law),reviews the legality of the acts of the institutions of the EU and ensuresthe respect, by the Member States, of their obligations under the Treaties.

1

logic) and the development of individual rights (subjectivelogic). This brief overview will also set out the evolution ofthe principle of legal certainty on American soil. While itspresence, which was established during the 19th century,suffered from the attack of Franck, the goal of many Americanlegal realists was the predictability of law. The second part looks at the sprawling, confused andcontradictory nature of the case-law of the European Court ofjustice regarding the principle which, according to the Court,covers or is at the origin of numerus core principles,including legitimate expectations, vested rights, nonretroactivity and res judicata.

Two simple ideas are proposed in a final section: the first isto abolish the (unrealistic) cartesian logic as a basis forlegal certainty in favour of the fiduciary logic. The secondargues that, as regards beneficiaries, legal certainty shouldoperate mainly in favour of the individual and not the powersthat be, such as the European Union.

1. Back to the past

A generally accepted notion among scholars is that theprinciple of legal certainty is a child of the 19th centuryGerman « Rechtsstaat » (a -more State-based- approximation ofthe Rule of law). However, this does not explain the problemsraised by the case-law of the Court of justice of the EuropeanUnion. This paper demonstrates that to understand the trueessence of the principle of legal certainty, one needs to gomuch further back in time.

A. In Europe

The first logic of legal certainty appearing in Europe is thepolitical logic, where the principle was mobilised to serve thepolitical order of the day, be it the Prince, King, Emperor orCity. From Antiquity to Middle Ages, the need for law to beaccessible and stable was mostly aimed at ensuring theeffectiveness of rules adopted by the authorities, and not atprotecting the individual. Some Greeks viewed statutes andtheir stability as being almost sacred. Demosthenes wrote that

2

Locrians were so fond of the immuability of law that « theyconsider it right to abide by their old established laws, topreserve hereditary institutions, and not to legislate for thegratification of caprice, or for the escape of crime ; and sofar do they carry this, that, if any one wishes to enact a newstatute, he proposes it with his neck in a noose, and if thestatute is judged to be good and useful, the proposer goes awayalive, but, if not, the noose is drawn and he dies »2.

Under the Roman empire, one of, if not the most significantcontribution to the cause of legal certainty was EmperorJustinian’s codification. He sought to compile « all ancientlaw applied since nearly fourteen hundred years » in fiftybooks that would act « like a fence beyond which there will benothing else to search » and that would be published to ensurethat laws would be known by their addressees3. Justinian’sdesire for legal certainty, via the publication of his Code forhis subjects to read, must be principally understood in lightof his political ambition. As Max Weber puts it, there isgenerally a very strong political interest in legal certaintybrought up by codifications, that usually follow closely thecreation of new political entities4.

The best illustration of the political logic of legal certaintycame a few centuries later when the Kings of France compiledthe governing rules in a written form to strenghten theirpower. One such example is the ordinance of Villers-Cotterêts,imposed by King François 1st in 1539, that required judgmentsto be written « so clearly that there would be neitherambiguïty, nor uncertainty, nor need for interpretation »5.

While a detailed discussion of this period is beyond the scopeof this paper, reference must be made to the Renaissance and2 DEMOSTHENES, « The Orations of Demosthenes Against Timocrates, Aristogiten,Aphobus, Onetor, Zenothemis, Apaturius, Phormio, Lacritus, Pantaenetus,Nausimachus, Boeotus, Spudias, Phaenippus, and for Phormio », in Works ofDemosthenes, transl. by Ch. Kennedy, H. G. Bohn, 1861, vol. IV, pp. 36-37.3 JUSTINIEN, « Première préface sur la composition du digeste adressée àTibonien », in Les cinquante livres du digeste ou des pandectes de l'empereur Justinien, op.cit., Livre 50, §1 ; 8 et 24 (free translation).4 M. WEBER, Sociologie du droit, transl. by J. Grosclaude, Paris, Quadrige,P.U.F., 2007, p. 198. 5 In old French : « (…) escriptz si clerement qu'il n'y puisse avoir aulcuneambiguite ou incertitude, ni lieu a en demander interpretacion ».

3

the Protestant Reform initiated by Luther and Calvin. For thesekey events stimulated the apparition of the two factors fromwhich arose the modern world : the secularisation of publiclife and the emancipation of the individual from spiritualauthority. The 16th century was a new cultural world in whichthe clergy had lost the monopoly of study and knowledge, forthe benefit of a new social class of enriched bourgeois andnobles freed from military obligations, eager to contribute tothe development of western thought.

The most important century for legal certainty is, however, the17th century when the three other logics took off, even iftheir roots can be traced back to antiquity.

The cartesian logic results from the impregnation of legaltheory by the scientific thinking of the day, mainlyillustrated by Descartes, Galileo and Newton. In a centurywhere reason became almost a new divinity, the convictionspread that a rational knowledge of causes would eventuallyenable humans to predict all consequences and events to come.The question of predictability became indispensable to define afuture no longer guaranteed by God. This credo soonconvinced Natural law thinkers like Grotius and Leibniz, andgave rise to the cartesian logic of legal certainty : masteringlegal causes should allow knowledge of legal consequences.

In parallel, the fiduciary and subjective logics grew with thedevelopment of the theories of social contract and individualrights. Against Hobbes’ absolutist conception of the socialcompact, Locke considered that the people retain a supremepower to remove the legislative power « when they find thelegislative act contrary to the trust reposed in them: for allpower given with trust for the attaining an end, being limitedby that end, whenever that end is manifestly neglected, oropposed, the trust must necessarily be forfeited, and the powerdevolve into the hands of those that gave it, who may place itanew where they shall think best for their safety andsecurity »6. Instead of the certainty that law can offer(cartesian logic), the fiduciary logic emphasizes the trust6 J. LOCKE, Second Treatise of Government, Cambridge, Hackett Publishing, 1980, p.73, §149. According to him, power « ought to be exercised by establishedand promulgated laws ; that (…) the people may know their duty, and be safeand secure within the limits of the law » (§137).

4

reposed in the authorities that the latter must respect. Fromthe angle of the beneficiary of legal certainty, Locke’sfiduciary logic promoted a subjective (and not a political) logic.More generally, the notion of individual right (« droitsubjectif »), attributed by Villey to Ockham7, started to playa significant role in legal theory and political philosophy.

In many respects, the two centuries following the 17th centurytook over and extended its legacy with the quest for certaintyand the central position of the individual peaking during theEnlightenment. The English lawyer Fortescue Aland expressed hisconviction that law was « capable of as great Certainty, as anyScience, or Profession whatsoever, (that noble, and perhapsonly Science, Mathematicks, excepted) »8.

Montesquieu also defended a cartesian logic of legal certainty,while connecting it with its subjective logic. His famoussentence that reduces judges to the mouth of the law should beread in light of the double postulate of the time. On the onehand, men entered into a social contract, under which specificpowers were entrusted to a legislative authority. On the other,it was thought possible to enact laws offering the samecertainty as scientific laws9. If the legislator had the solelegitimacy to create law, judicial intervention had to remainpurely mechanical. In line with Locke, Montesquieu advocatedthe necessity for judges not to surprise the addressees oftheir decisions, thereby promoting the subjective logic oflegal certainty. However, unlike the English philosopher, theFrench thinker substituted, to the association of its fiduciaryand subjective logics, an articulation of the latter with thecartesian logic. This articulation has had a lasting impact onthe concept, creating a confusion between the democraticlegitimacy and the idealised capacity of the legislative powerto foresee everything.

7 M. VILLEY, La formation de la pensée juridique moderne, op. cit., pp. 240-247). 8 J. FORTESCUE-ALAND, « The Preface », in J. FORTESCUE, The Difference Between anAbsolute and Limited Monarchy; As it more particularly regards the English Constitution, Londres,Ed. W. Bowyer, 1714, p. v (« Nor do I foresee any Absurdity in saying, thatLaw is capable of the best Certainty, even Demonstration it self »).9 See MONTESQUIEU, L’esprit des lois, Paris, Flammarion, 2008, Livre XI, ch. VI,pp. 244-256.

5

Montesquieu had an important influence on the Italian lawyerBeccaria, who proposed to apply the notion of « syllogism »10

to law. He argued that, « in every criminal cause, the judgeshould reason syllogistically. The major should be the generallaw -, the minor, the conformity of the action, or itsopposition to the laws ; the conclusion, liberty, orpunishment »11.

This conception of judging, still shared by many Europeancontinental lawyers today also had a major impact in England.Blackstone, who had read Montesquieu’s Esprit des Lois12, publishedhis famous Commentaries on the Laws of England around 1767. He alsodenied judges the power to create law :

« But here a very natural, and very material, question arises:how are these customs or maxims to be known, and by whom is theirvalidity to be determined? The answer is, by the judges in theseveral courts of justice. They are the depositaries of the laws;the living oracles, who must decide in all cases of doubt, andwho are bound by an oath to decide according to the law of theland. (…) For it is an established rule to abide by formerprecedents, where the same points come again in litigation; (…)Yet this rule admits of exception, where the former determinationis most evidently contrary to reason; much more if it be dearlycontrary to the divine law. But even in such cases the subsequentjudges do not pretend to make a new law, but to vindicate the oldone from misrepresentation. For if it be found that the formerdecision is manifestly absurd or unjust, it is declared, not thatsuch a sentence was bad law, but that it was not law; that is,that it is not the established custom of the realm, as has beenerroneously determined »13.

The quest for certainty continued throughout the 19th century,spurred on by the scientific revolution, as illustrated by thedeterminist views of Laplace14 and Comte’s conviction that10 Major : all men are mortal, minor : Socrates is a man, conclusion :Socrates is mortal.11 C. BECCARIA, An Essay on Crimes and Punishments, trad. by M. De Voltaire, Dublin,John Exshaw, 1768, p. 12. 12 D.W. CARRITHERS, M.A. MOSHER et P.A. RAHE, Montesquieu's Science of Politics: Essays onThe Spirit of Laws, Oxford, Rowmand and Littlefield Publishers, 2001, p. 2. 13 W. BLACKSTONE, Commentaries on Laws of England, 6ème éd., London, 1825, pp. 69-70. 14 P.-S. LAPLACE, Essai philosophique sur les probabilités, Paris, Ed. Bachelier, 1825,pp. 3-5.

6

social phenomena can be foreseen with the same precision as forall other phenomena15. However, the cartesian model alsostarted to wane, albeit discretely, with the rise of thehistorical question or under the attacks from the masters ofsuspicion (Freud, Marx and Nietzsche).

In the legal field, the fact that Natural law theories werecalled into question by theories of legal positivism did notdiminish the importance given to the certainty of law, on thecontrary. French and German civil codes gave flesh to thebelief in the capacity of law to be predetermined by abstractlydefined concepts, despite the antirationalist reaction of theGerman Historical School or the authority of Portalis16. InEngland, the plea of Bentham in favour of codification, whichturned out to be unsuccessful, seems to have made him defendthe cause of the cartesian logic of legal certainty. However,his conception of the principle, according to which law shouldprotect its addresses’ expectations17, represents but a timidand isolated sketch of fiduciary logic.

The 19th century also set the stage for the German« Rechtsstaat », the French « Etat de droit » and the theorization ofthe (older) English « Rule of law », which all share the ideaof a political power guaranteeing individual rights and beingsubject to legal principles such as legal certainty. Indeed,the notion of a certain and predictable law appears to havebeen essential in framing the substantial economic developmentcaused by European industrialization. The success of the Ruleof law contributed to the expansion of the subjective logiceven if its lack of justiciability had a paralyzing effect inpractice. Interestingly enough, while one could think that the

15 A. COMTE, Cours de philosophie positive, 5ème éd., Paris, Au siège de la sociétépositiviste, 1893, Vol. 4, 48ème leçon, pp. 249 et 253. 16 This father of the French civil code wrote : « to foresse everything is agoal that is impossible to reach » (« Discours préliminaire prononcé parPortalis, le 24 thermidor an 8, lors de la présentation du projet arrêtépar la commission du gouvernement », in P.A. FENET, Recueil complet des travauxpréparatoires du Code civil, Paris, 1827, pp. 467-477 (free translation)). On theGerman historical school, see inter alia J.M. KELLY, A Short History of Western LegalTheory, O.U.P., 1992, pp. 321-322.17 J. BENTHAM, Theory of Legislation, transl. by R. Hildreth, 2ème éd., London, Ed.Trübner, 1871, pp. 149-157 (« The legislator owes the greatest respect tothis expectation which he has himself produced (…). The goodness of thelaws depends upon their conformity to general expectation »).

7

very notion of « Rule of law » was forged to establish –orrestore- individual trust in the authority to respect rights,the fiduciary logic seems to have disappeared from legalthinking.

In a less obvious way, the German « Rechtsstaat » and the French« Etat de droit » allowed for a return of the political logic oflegal certainty, under a renewed form18. Behind its officialsubjective logic, legal certainty continued to serve theinterests of the Leviathan, even if Machiavel’s Prince wasreplaced by the Nation-State and its legal order. In a contextwhere Nation-States tried to establish their foundations, thearguments of the « Rechtsstaat » and of legal certainty – ideallyunder its cartesian form, which embody notions of politicalneutrality and relentless logic- represented fantastic (andartificial) instruments of legitimization.

While the search for certainty continued during the 20thcentury, some scientific discoveries (like Heisenberg’sprinciple of uncertainty), and new philosophical thoughts (likeHeidegger and Popper’s) undermined its claims in a more radicalway than ever before.

The State was increasinlgy expected to arbitrate numerous andgrowing conflicts of interests and law started to be used as anactive instrument for social change. Moreover, language becamethe centre of attention. After the linguistic turn, it was nolonger considered as the « garment of thought »19 but as itsinevitable –and sometimes ideological- medium.

From the point of view of legal certainty, the explosion ofindividual interests and the uncertainties of language drovemost legal theoreticians to accept that the indeterminacy oflaw was unavoidable. Almost all of them, however, could nothelp clinging to the need for certainty in law and to thepossibility of ensuring the predictability of case-law. Thistendency is very obvious with the American realists.

18 It is less the case for the Rule of law.19 Th. CARLYLE, Sartor Resartus, Boston, J. Munroe and company, 1837, p. 79.

8

B. In America

Blackstone’s « declaratory » theory had a massive influence inthe United Sates and the cartesian logic of legal certaintyplayed a largely undermined role in the 19th and early 20thcenturies20. It is likely that dominant American legal thoughtadhered broadly to legal positivism during the 19th century21.In that context, legal certainty was not unknown, its veryterms being used regularly by the courts22. In 1905, the FrenchAmerican lawyer Fr.R. Coudert published an article dedicated tojustice and certainty, in which he reported the discontent fromboth the New York bar and its clients caused by the growinguncertainty of case-law23. Described by J. Frank as one of themost influential American scholars24, professor J.H. Beale fromHarvard published in 1916 his Treatise on the Conflict of Laws, whichwas a plea for legal formalism and legal certainty :

20 C.S. ROGERS, « Perspectives on Prospective Overruling », U. Mo. Kan. City L.Rev., 1968, p. 41. B.H. Levy considers that Blackstone’s theory mirrors themechanical conception of his century (« If law, like nature, weremechanical, then it is only natural that judge should be an automaton »)and regrets the impact that this theory still had, in 1960, on Americanscholars : « The hold of an inherited legal philosophy controls our mindagainst our will and forces us into intellectual circumlocutions whichwould shame a medieval schoolman » (« Realist Jurisprudence and ProspectiveOverruling », Penn. L. Rev., 1960, pp. 2-4). In 1824, Justice John Marshallfrom the Supreme Court wrote : « Judicial power, as contradistinguishedfrom the power of the laws, has no existence. Courts are the mereinstruments of the law, and can will nothing. (…) Judicial power is neverexercised for the purpose of giving effect to the will of the judge, alwaysfor the purpose of giving effect to the will of the legislature; or, inother words, to the will of the law » (Osborn v. Bank of the United States, 22 U.S.9 Wheat. 738 (1824), p. 866).21 J. E. HERGET et S. WALLACE, « The German Free Law Movement as the Source ofAmerican Legal Realism », Virginia L. Rev., 1987. 22 Supreme Court, Humphrey Marshall and Wife vs. James Currie [février 1807], inReports of decisions in the Supreme Court of the United States, 5ème éd., Boston, Little,Brown & Cie, 1870, Vol. 2, p. 60 ; Barry v. Edmunds [avril 1886], Virginia LawJournal, Vol. 10, Issue 4 (April 1886), p. 206. At a state level, see Engels,Sheriff & Cie v. Day and others, Reports of cases in the Supreme Court of the State of Arkansas,1842, Vol. 3, p. 278. 23 That article was also published in Fr. R. COUDERT, Certainty and justice ; studies ofthe conflict between precedent and progress in the development of the law, New York, Appleton& Cie, 1913, p. 13. See also W.N. HOHFELD, « Some Fundamental LegalConceptions as Applied in Judicial Reasoning », Yale L.J., 1913, pp. 28-54. 24 J. FRANK, Law and the Modern Mind, New Brunswick, Transaction Publishers,2009, p. 53.

9

« The social need of continuity in law is most clearly feltbecause society needs to know the law in advance of judicial action uponit. In order that law may help rather than hinder the carryingon of the work of society it must be possible for every person, of his ownknowledge or by the help of others’ knowledge, to discover the application of the law toany contemplated act. He must be in some way secured against unexpected legalconsequences of his actions. Business could not go on, industry couldnot be maintained unless it were possible for the producer or themerchant to learn how he could conform his activities to the law.For this purpose it must be possible for one learned in the law tospeak with authority on the application of law to the proposedacts and to predict with reasonable degree of certainty the decision of courts incase the legality of the acts should be called in question »25.

Last but not least, came the phenomenon of « Treatises », whichinvolved the systematic compilation of case-law. In 1923, theAmerican Law Institute was founded and asked to edit Restatements inthe most important areas of case-law. Around the end of the1920s, these treatises could be considered as theimplementation of legal positivism, even if their limitedsuccess already signalled its decline, mainly promoted byAmerican realists26.

It must by underlined that, contrary to European thought,American attachment to legal certainty has always been framedby pragmatic philosophy and firmly based on the decisions ofcourts27. As early as 1892, Harvard’s professor Chipman Grayrefused Austin’s conception according to which legal scienceshould restrict itself to describing law, defined as theSovereign’s commands28. In The Nature and Sources of the Law, he wrotethat law was only judge-made rules and that laws were onlylegal sources29.

Despite Gray’s pioneering works, it is usually accepted thatthe « real » father of the realist movement was Justice Wendell25 J.H. BEALE, A Treatise on the Conflict of Laws, op. cit., pp. 45-46, §4.12 (theauthor’s emphasis).26 J. E. HERGET et S. WALLACE, « The German Free Law Movement (…) », op. cit.,pp. 429-430. 27 J.E. HERGET et S. WALLACE, « The German Free Law Movement (…) », op. cit., p.420. 28 J.C. GRAY, « Some Definitions and Questions in Jurisprudence », Harv. L.Rev., 1892, pp. 24-28..29 J.C. GRAY, The Nature and Sources of the Law, Aldershot, Ashgate Dartmouth, 1921,pp. 54 et 63 (« all the Law is judge-made law »).

10

Holmes Jr30. The famous judge, who used to say that « the lifeof the law has not been logic : it has been experience »31

broke definitely with formalist and positivist currents oflegal thinking. His relativist conceptions (« You can give anyconclusion a logical form ») brought him to consider that lawcould be reduced to « the prophecies of what the courts will doin fact, and nothing more pretentious »32.

The realist school boasted brilliant and influential minds likeCardozo33 or Llewellyn34. However, the following linesconcentrate on Frank’s Law and the Modern Mind as it can beconsidered as the official grave digger of the principle oflegal certainty, six feet under the American soil35. Hisstarting point was the myth, in which both lawyers and laymenstill believed, that law could be clear, certain andpredictable, while it would always be vague and variable asdealing with human relations : « Even in a relatively staticsociety, men have never been able to construct a comprehensive,eternized set of rules anticipating all possible legal disputes30 J. FRANK, Law and the Modern Mind, op. cit., pp. 132-133 ; J.M. KELLY, A ShortHistory of Western Legal Theory, op. cit., pp. 366-368. As writes Fuller : « Therealist himself likes to consider his movement as taking its origin in anaddress delivered by Justice Holmes in 1897 » (« American Legal Realism »,Penn. L. Rev., vol n°82, 1934, p. 429). See also H. PIHLAJAMAKI, « AgainstMetaphysics in Law : The Historical Background of American and ScandinavianLegal Realism Compared », Am. J. Comp. L., 2004, pp. 471 et 479-481 ; J. E.HERGET et S. WALLACE, « The German Free Law Movement (…) », op. cit., p. 401.These authors consider that the official beginning of the movement was aspeech given by professor Oliphan at the annual meeting of the Association ofAmerican Law Schools en 1927 (Ibid., p. 431).31 O.W. HOLMES Jr., Common Law, London, MacMillan, 1882, p. 1.32 O.W. HOLMES Jr., « The Path of the Law », Harv. L.R., 1897, pp. 459-461 et466.33 B.N. CARDOZO, The nature of the judicial process, New Haven, Yale University Press,p. 10 (« I take judge-made law as one of the existing realities of life »).34 K. LLEWELLYN, The Bramble Bush. The Classic Lectures on the Law and Law School, New York,O.U.P., 2008, pp. 15 et 116. According to him : « What these officials doabout disputes is, to my mind, the law itself (…) And rules, in all ofthis, are important to you so far as they help you see or predict whatjudges will do or so far as they help you to get judges to do something.That is their importance. That is all their importance, except as prettyplaythings » (Ibid., pp. 5 et 7). He also wrote : « the only test of whetherand how far a rule authoritatively prevails as a rule of Ought is : how far will courtsfollow it (…) » (Ibid., p. 82).35 On his impact on the discredit of the notion in the US, see J.R. MAXEINER,« Legal certainty : a European Alternative to American legal indeterminacy?», Tul. J. Int’L & Comp. L., 2006-2007, p. 544.

11

and settling them in advance »36. For Frank, there was nohypocrisy in this failure « to recognize fully the essentialplastic and mutable character of law »37, which is due to anunconscious childish need for certainty :

« That religion shows the effects of the childish desire torecapture a father-controlled world has been often observed. Butthe effect on the law of this childish desire has escaped attention.

And yet it is obvious enough : to the child, the father is theInfaillible Judge, the Maker of definite rules of conduct »38.

Frank’s definition of law was the following : « Law, then, asto any given situation is either (a) actual law, i.e., aspecific past decision, as to that situation, or (b) probablelaw, i.e., a guess as to a specific future decision »39. Hecompares the judicial process to the writer of a detectivenovel who comes up first with an idea of the end and thenbuilds backwards the reasoning to reach it : « Judicialjudgments, like other judgments, doubtless, in most cases, areworked out backward from conclusions tentativelyformulated »40. For him, the stimuli that bring a judge todecide a case in one sense or another are not only rules andprinciples but also « political, economic or moral biases »and, even more, individual factors41. He went as far as toexpect from judges detailed autobiographies to make themconscious of their preconceived ideas and determinisms42 andsuggest that law schools « become, in part, schools ofpsychology applied to law in all its phases »43.

An interesting realisation is that, despite his ferocious andefficient attack against legal certainty, Frank gave a36 J. FRANK, Law and the Modern Mind, New Brunswick, Transaction Publishers,2009, p. 6.37 Ibid., p. 11. 38 Ibid., p. 19 (« The essence of the basic legal myth or illusion is thatlaw can be entirely predictable. Back of this illusion is the childishdesire to have a fixed father-controlled universe, free of chance and errordue to human faillibility. In early stages of legal development this desirewas more intense than now (…) »).39 Ibid., pp. 50-51.40 Ibid., pp. 108-109.41 Ibid., pp. 113-114. 42 Ibid., p. 123 and 124 (« For in the last push, a judge’s decisions are theoutcome of his entire life-history »).43 Ibid., pp. 124 and 156.

12

significant value to predictability44 : « By abandoning aninfantile hope of absolute legal certainty we may augmentmarkedly the amount of actual legal certainty »45.

It appears that, unlike what one might think, the realistcurrent did not succeed in getting rid of legal certainty but,on the contrary, protected it or, at least, reduced legaluncertainty. Given the legislative lacunae, the undeterminedcharacter of precedents and the creative power of judges,realists thought that enhancing prediction of judicialdecisions could only be achieved through the adoption of ascientific method based on psychology, economy, sociolology orstatistics46. In a curious way, the realist school maintainedthe faith in some kind of cartesian logic of law, whichpostulates that judicial decision is, to a certain extent,predetermined47.

2. The case law of the European Courts of justice

In January 2013, the ECJ and the GC had rendered more than2.500 decisions containing the expression « legal certainty »,

44 Following Frank, everybody can make law, by making predictions on futuredecisions and « the more accurate, the more reliable, the forecastcontained in it, the more valid the rule » (J. FRANK, Law and the Modern Mind,op. cit., pp. 297-298).45 Ibid., p. 171. Maybe conscious of his own support to the « childishmyth », he wrote, in an addendum to the second printing : « That furtherobservation and description of what induces decisions will make futuredecisions markedly more predictable seems to the writer most improbable. Heinclines, tentatively, to the belief that most more accurate description ofthe judicial process will serve to show that efforts to procure suchpredictability (via anthropology, economics, sociology, statistics orotherwise) are doomed to failure » (Ibid., p. 399).46 For W.E. Rumble,  « what the legal realists rejected was not the goal ofincreasing legal certainty, but the particular means which havetraditionnaly been relied upon to achieve it » (« American Legal Realismand the Reduction of Uncertainty », J. Publ. L., 1964, p. 64). See also H.J.M.BOUKEMA, Judging. Towards a rational judicial process, W.E.J. Tjeenk Willink, 1980, p.26 ; L.L. FULLER, « American Legal Realism », op. cit., pp. 434-437. 47 On the aftermath of the realist school, see W.E. RUMBLE, « American LegalRealism and the Reduction of Uncertainty », J. Publ. L., 1964, pp. 61 et 72 ;J. E. HERGET et S. WALLACE, « The German Free Law Movement (…) », op. cit., p.436 ; J.M. KELLY, A Short History of Western Legal Theory, op. cit., p. 369. Forcriticisms of their positions, see inter alia H. KELSEN, Théorie pure du droit, 2d

ed., transl. by Ch. Eisenmann, Paris, L.G.D.J., 1999, pp. 95-97 and H.L.AHART, The concept of law, 2d ed., Oxford, Clarendon Press, 1994.

13

with the existence of the principle first « acknowledged » asearly as 196148. Over the years, its scope has undergone animpressive expansion and today encompasses several othergeneral principles of law (a). It is important to underlinethat general principles of European law have « constitutionalstatus »49, with equal ranking as the founding treaties (whichform EU’s « primary law »). This status gives them a(potentially) far-reaching impact, as they can be used tointerpret or annul secondary law (legal instruments based onthe Treaties, i.a. regulations, directives, decisions,conventions, etc.).

However, the « sprawling nature » of the principle of legalcertainty and its constitutional status contrast hugely withits limited effectivity which is scattered with confusions oreven contradictions (b).

a) A sprawling nature

According to the ECJ the principle of legal certainty requires,in a nutshell, « that rules of law be clear, precise andpredictable as regards their effects »50. The application ofthe rules of law must therefore « be foreseeable by thosesubject to them »51. It demands in particular that EU « rulesenable those concerned to know precisely the extent of theobligations which are imposed on them. Individuals must be ableto ascertain unequivocally what their rights and obligationsare and take steps accordingly »52. One should notice theabsolute terms used by the ECJ, that evoke the cartesian logicof legal certainty.

48 Case 42 et 49/59, S.N.U.P.A.T., 22 March 1961, Rec. pp. 103 et 159. TheECJ was created in 1952 and gave its first judgement in 1954. All thejudgments of the European court of justice are available on the website ofthe Court (www.curia.eu). A « C » before the reference or nothing (C-72/10or 28/74) refer to judgments of the Court itself (the ECJ), while a « T »before the reference (T-77/11) refers to the Court of First instance (theCFI).49 Case C-101/08, Audiolux, 15 Octobre 2009, §63 ; Case C-174/08, NCC Construction Danmark, 29 octobre 2009, §42.50 Case C-72/10-C-77/10, Marcello Costa, 16 February 2012, § 74. 51 Case C-201/08, Plantanol, 10 September 2009, §46.52 Case C-345/06, Heinrich, 10 March 2009, §44.

14

The principle of legal certainty has also been described as an« umbrella » principle53, as it covers or gathers several otherprinciples54.

The principle of legitimate expectations, which is said to be a« corollary » of the principe of legal certainty, can be reliedon by « any economic operator on whose part the nationalauthorities have promoted reasonable expectations »55. Vestedrights cannot be called into question as « (a)ny other solutionwould be contrary to the principle of legal certainty »56. Itis also settled case-law that « the principle of legalcertainty precludes a regulation from being appliedretroactively, irrespective of whether such application mightproduce favourable or unfavourable effects for the personconcerned »57.

For the Court, a « limitation period, by preventing situationswhich arose a long time previously from being indefinitelybrought into question, tends to strengthen legal certainty butcan also allow the acceptance of situations which at least inthe beginning were unlawful »58. However, « in order to fulfiltheir function, limitation periods must be fixed inadvance »59. The same principle of legal certainty requires,indeed, « that a provision laying down a preclusive period,particularly one which may have the effect of depriving aMember State of the payment of financial aid its applicationfor which has been approved and on the basis of which it has53 X. GROUSSOT, General principles of Community Law, Groeningen, Europa LawPublishing, 2006, p. 24.54 About general principles of EU law, see T. TRIDIMAS, The General Principles of EU Law, sd ed., Oxford European Community Law Library, 2006.

55 Case C-201/08, Plantanol, 10 September 2009, §53.56 Case C-337/07, Altun, 18 Decembre 2008, §59-60.57 Case C-154/05, Kersbergen-Lap e.a, 6 July 2006, §42.58 Case T-22/02, Sumitomo Chemical, 6 Octobre 2005, §82 (« The extent towhich provision is made for it is therefore the result of a choice betweenthe requirements of legal certainty and those of legality, on the basis ofthe historical and social circumstances prevailing in a society at a giventime. It is therefore a matter for the legislature alone to decide »). Morerecently, the Court confirmed that « (i)n general, limitation periodsfulfil the function of ensuring legal certainty » (Case C-367/09, SGSBelgium e.a., 28 Octobre 2010, §68).59 Case 52/69, Geigy, 14 July 1972, §21 ; Case C-74/00 P et C-75/00 P,Falck, 24 Septembre 2002, §139.

15

already incurred considerable expenditure, should be clearlyand precisely drafted so that the Member States may be madefully aware of the importance of their complying with the time-limit »60.

In the absence of any provision on the matter, « thefundamental requirement of legal certainty has the effect ofpreventing the Commission from indefinitely delaying theexercise of its power to impose fines (…) »61. In a famous caseRSV, the Court declared void a decision because of the« Commission's delay in giving the contested decision », which« could in the case in point establish a legitimate expectationon the applicant's part so as to prevent the Commission » fromrequiring the national authorities to order the refund of aState aid62.

The ECJ also established several rules of interpretationgrounded on the principle of legal certainty. According to theCourt, « where it is necessary to interpret a provision ofsecondary Community law, preference should as far as possiblebe given to the interpretation which renders the provisionconsistent (…) with the principle of legal certainty »63.Furthermore, it would « run counter to the principle of legalcertainty to interpret differently two provisions worded in anessentially identical manner and which, moreover, appear in thesame article of a Community regulation »64.

The multilingual nature of the EU can obviously cause problems.On one hand, « any divergent interpretation at national level »of obligations provided in a European directive would« undermine the objectives of the Community legislation andlegal certainty »65. On the other hand, the « elimination oflinguistic discrepancies by way of interpretation may in60 Case 44/81, Allemagne/Commission, 26 May 1982, §16.61 Case 52/69, Geigy, 14 July 1972, §21. The European Commission can bedefined as the executive body of the EU. It is supposed to represents theinterests of the EU, while the Council represents the interests of theStates and the European Parliament the interests of the citizens.62 Case 223/85, RSV, 24 Novembre 1987, §17-19.63 Case C-1/02, Borgmann, 1 April 2004, §30.64 Case C-401/99, Thomsen, 20 June 2002, §35.65 Case C-391/05, Jan De Nul, 1 March 2007, §23. A uniform interpretation ofEU law is required in « the interest of legal certainty » (Case C-275/05,Kibler, 26 Octobre 2006, §22).

16

certain circumstances run counter to the concern for legalcertainty, inasmuch as one or more of the texts involved mayhave to be interpreted in a manner at variance with the naturaland usual meaning of the words »66.

The theoretical importance of the principle of legal certaintyhas turned out to be deceptive in practice. There is indeed ahuge gap between the very generous principles described aboveand their very strict application.

b) Tangled tentacles flailing at the air

While the principle of legal certainty is still framed inabsolute terms (law must be certain, clear and precise so thatindividuals can ascertain unequivocally what their rights andobligations are), the ECJ decided in a case brought by theBelgian State, that « a degree of uncertainty regarding themeaning and scope of a rule of law » was « inherent in thatrule » and that the Court’s control can be « confined to thequestion (of) whether the legal measure at issue displays suchambiguity as to make it difficult for that Member State toresolve with sufficient certainty any doubts as to the scope ormeaning of the contested regulation »67. One cannot ignore thegap between this relative wording (much more realistic) and theabsolute wording that it still used when referring to legalcertainty.

The scope of the principle of legitimate expectations has been,over the years, narrowed down to the point of questioning its« general » nature. For example, this principle cannot beinvoked by « a prudent and circumspect economic operator » that« could have foreseen that the adoption of a measure is likelyto affect his interests »68 or « by an undertaking which hascommitted a manifest infringement of the rules in force »69.Economic operators are « not justified in having a legitimateexpectation that an existing situation which is capable ofbeing altered » by the national or European authorities « inthe exercise of their discretionary power will be

66 Case 80/76, North Kerry Milk Products, 3 March 1977, §11.67 Case C-110/03, Belgium/Commission, 14 April 2005, §31.68 Case C-201/08, Plantanol, 10 Septembre 2009, §53.69 Case C-65/02 P and C-73/02 P, ThyssenKrupp, 14 July 2005, §41.

17

maintained »70. Furthermore, « even if the European Communityhad first created a situation capable of giving rise tolegitimate expectations, an overriding public interest maypreclude transitional measures from being adopted in respect ofsituations which arose before the new rules came into force butwhich are still subject to change »71.

More examples could be given but it suffices to say thatjudgments where the Court found a violation of the principle oflegal certainty (or its « subprinciples ») are marginalcompared to the number of cases where it was invoked72.

More worrying is the fact that the ECJ’s case-law revealssignificant contradictions in its application of the principle.Let us focus on two examples, selected amongst many.

EU law had not foreseen any time limitation regardinglitigation beween the European Bank of Investment (EBI) and itsagents. Relying on the principle of legal certainty, the GCdecided that, by analogy with the period applicable to Europeancivil servants, a three-month time limitation was reasonable.Thus, in its Orlando judgment, the GC considered that an actionbrought by EBI staff members, that had arrived two minutesafter these three months, was not admissible73. However, thefact that a judge can decide, with retroactive effect, time-limits within which litigants must bring proceedings, istotally incompatible with the requirement, justified by thesame principle of legal certainty i.e. that it should be knownin advance74. Such a decision illustrates a conflict betweenthe political logic of legal certainty (time limitationintroduced to preserve the interests of the EU) and itssubjective logic (individuals should be able to rely on delaysthat are predefined by a statute)75.

70 Case C-201/08, Plantanol, 10 Septembre 2009, §53 ; Case C-335/09P,Commission/Poland, 26 juin 2012, §180.71 Case C-519/07 P, Commission/Koninklijke FrieslandCampina NV, 17 Septembre2009, §85. 72 J.-P. PUISSOCHET, « ‘Vous avez dit confiance légitime ?’ (le principe deconfiance légitime en droit communautaire) », in L'État de droit, Mélanges enl'honneur de Guy Braibant, 1996, p. 593.73 Case T-234/11 P, Orlando Arango Jaramillo e.a./BEI, 19 June 2012, §27.74 The judgement was overruled by the Court of justice but for differentreasons (see Case C-334/12 RX-II, 28 February 2013, Orlando).

18

As seen above, the same principle of legal certainty requiresan EU text to be interpreted both according to the natural andusual meaning of the words and in a way that suppresses anydivergence between the Member states. The first requirementclearly favours the subjective logic of legal certainty whilethe latter serves its political logic, that being the EU’sinterest in reaching a uniform interpretation of Europeannorms. A problem arises when the « usual meaning » differsamongst the Member states, as the EU has 24 official languagessince Croatia joined in July 2013. M. Röser was subjected, inGermany, to penal proceedings in accordance with a Europeanregulation that allowed for two interpretations under itsGerman version. After having realised that the English, Frenchand Italian versions were phrased without ambiguïty, the ECJdecided that the interpretation which was the least favourablefor M. Röser should prevail76. That choice granted more legalcertainty to the whole European legal system but ignored themeaning that M. Röser could legitimitaly rely on.

3. Two explanations and two proposals

a) History’s input

The author’s thesis is that the gap between the ECJ’s discourseregarding the principle of legal certainty and its applicationshows the survival of its cartesian logic. Indeed, the Court’sdefinition of legal certainty in absolute terms assumes theexistence of perfectly rational public authorities and legalsubjects, rules as precise as scientific laws and a neutralmathematical language. However, these assumptions have turnedout to be untenable, especially in the application of EU law.There are two possible explanations for the maintenance of thecartesian myth.

The first is that, following the scientific revolution, law andscience acted as a substitute for the certainty offered by Godwhich started to be questioned from the 17th century onwards.The effect of that process resulted in the inclusion of75 It is also goes against the CFI’s affirmation that, « as regardslimitation periods, legislative provisions unconnected with the case inpoint cannot be applied by analogy » (Case T-137/01, StadtsportverbandNeuss e.a., 17 septembre 2003, §123). 76 Case 238/84, Röser, 27 February 1986, § 22.

19

cartesian logic in the collective (legal) unconscious. A secondexplanation lies in the dual quest for legitimacy andeffectiveness carried out by the ECJ in its early years. Theacceptability of the new political order being a condition ofits success, the Court needed to convince the signatory Statesthat the treaties they had signed were not going to surprisethem. And what better way to reassure the founding members,heirs of Montesquieu, Beccaria and Savigny, than to propose amodel that saw judges as being merely the mouths of a law thatwas fruit of their fragile compromises? We observe again theinteresting and fictional combination of the cartesian andsubjective logics of legal certainty (legal subjects being herethe Member States of the Union), that contributed to establishthe legitimacy of Nation-States in the 19th century. However,and in the same way, it is mainly the political logic that thecartesian myth supports in order to strenghten theeffectiveness of the new legal order (the Prince being here theEU).

Our historical survey uncovers a second element that helps toclarify some of the contradictions of European case-law, whichis the conflict between the political and subjective logics ofthe principle of legal certainty. Just like the French kings, the new Community order needed toestablish the effectiveness of its rules while maintaining theachievements of the « Rechtsstaat » in terms of protectingindividuals. This could explain why the Court of justiceconsiders that the principle of legal certainty requires legaltexts to be interpreted in the natural and usual meaning of itsterms (subjective logic), while at the same time imposing, inthe interest of EU law, a uniform interpretation in case oflinguistic discrepancies (political logic, as in the Rösercase). With regard to time limitations within which appeals maybe lodged, legal certainty requires these limits to be known inadvance by legal subjects (subjective logic) but the sameprinciple is also used to justify denying individuals, withoutany legal basis, access to European justice (political logic,as in the Orlando case).

b) From certainty to trust

20

On the normative level, the author first proposes topermanently rid legal certainty of its cartesian logic andsecondly advocates a return of its fiduciary logic. The firstproposal is dictated by the nature of law, which defies allmechanistic reduction. Maintaining cartesian logic requiresmaintaining the illusion of a rational legislator who can - andwants- to speak in unambiguous language. It keeps the myth oflegislator’s omniscience alive and denies the very essence ofpolitical compromise. It also loses sight of the teachings ofhermeneutics, which shows that any text requiresinterpretation, which is never neutral but involves, as Gadamerwrote, a "sense of anticipation" in the head of theinterpreter77.

More broadly, the issue of interpretation entails the need toquestion the cartesian claims of judicial syllogism popularizedby Beccaria, a great admirer of Montesquieu. Nothing in thejudicial process can be reduced to purely logical reasoning.Almost every case involves a choice made by the judge betweendifferent standards and principles (major), a subjectiveselection of the « relevant » facts (minor) and acontextualisation of the decision (conclusion).

Not only does the European legal order not escape from thesesources of uncertainty, but on the contrary, intensifies them.The adoption of the European norm presupposes the political,linguistic and legal performance of bringing into a single textthe provisions negotiated by 28 States and intended to have aneffect in the same number of legal systems, in 24 officiallanguages. The same problems arise at judicial level, whetherthis is the ECJ or national judges submitting a request for apreliminary ruling. Finally, the general principles of law,which include legal certainty, can be seen as arms of massdestruction regarding their potential and unpredictable action.Their constitutional status allows them to declare void, intheory, any piece of secondary law.

These « barriers of uncertainty », that pave the path of thejudicial process, culminate in a death sentence for the

77 H.G. GADAMER, Vérité et méthode. Les grandes lignes d’une herméneutique philosophique,transl. by P. Fruchon, J. Grondin and G. Merlio, Paris, Ed. du Seuil, 1996,p. 351.

21

cartesian model of legal certainty. The European Courtproceedings should instead be considered in the configurationit presents : that of a potential "event" (« événement »). Thisnotion is borrowed – or rather diverted- from the Frenchphilosopher Alain Badiou and will be presented briefly, withall the shortcuts this entails. According to Badiou, an "event"is what comes to supplement a situation, which appears as its"outlaw", an irreductable singularity. The event, he writes, is"both located - it is an event of a particular situation- andsupplementary, hence absolutely detached or loosened, from allthe rules of the situation"78. Badiou gives, as examples ofsuch events, the French revolution, the Galilean creation ofphysics or the invention of classical music by Haydn79.

It might not be appropriate to compare -even landmark-judgments of the ECJ to these famous examples. The notionof « event » nonetheless provides stimulating food for thoughtregarding the judicial phenomenon. With each Court proceedingcomes the possibility of an irreductable singularity that wouldsupplement the original legal situation. It can be a departurefrom precedent or an unexpected fundamental decision. Bydefinition, it is impossible to predict with certainty whensuch an event will occur. What may seem to the modern (Europeancontinental) lawyer a bitter failure of the system wasconsidered as perfectly normal by Roman lawyers or Christiancasuists from the Middle Ages (and perhaps lawyers from commonlaw countries). The word "case" (« cas » in French) comes fromthe latin word « casus », derived from the verb « cadere »which means « to fall »80. It is therefore not surprising thatthe « case » sometimes « falls » beyond what the law meant tooriginally cover. Most problems that pervade the ECJ’sjurisprudence stem precisely from the random way in which casesarrive in its chambers and from the Court’s denial of theirsingularity.

78 A. BADIOU, L’Ethique. Essai sur la conscience du mal, Caen, Nous, 2003, pp. 62-65 and92 (free translation) ; A. BADIOU, Manifeste pour la philosophie, Paris, LesEditions du Seuil, 1989, pp. 16-17 and 89. 79 Ibid., p. 96. 80 S. BOARINI, « Collection, comparaison, concertation. Le traitement du cas,de la casuistique moderne aux conférences de consensus », in Penser par cas,J.-Cl. Passeron et J. Revel (dir.), Paris, Ed. de l’EHESS, 2005, p. 13.

22

It is therefore necessary for the ECJ to emphasize in itsdiscourse a relative conception of the requirements of legalcertainty (« some » clarity, a reasonable « predictability »),although one might wonder why bother – since its its applicationby the European courts frees the principle from its cartesianlogic. However, it is precisely to avoid a situation where theCourt, to save a rule, feels constrained to conclude that it isclear, precise and predictable when it is not, to the point ofblaming the legal subjects for its unpredictability. To use thesporting metaphor of the pole-vault, the bar of legalpredictability being set too high, confronts European judgeswith the following dilemma : either they decide that the legalpole (EU law) is too small, or they find that the vaulter (thelegal subject) does not do what it takes to clear the obstacle.Short of concluding that every EU rule of secondary law isunsufficiently clear, precise and predictable (and thereforeinvalid given the constitutional status of legal certainty),judges end up considering that its subjects have notdemonstrated due diligence in understanding the applicablerules. To avoid such an unfair and perverse effect, the authorconsiders the notion of « trust » to be a better and indeedmore honest basis for the principle than the cartesian« certainty ».

Which brings us to the second proposal, which is to do justiceto the fiduciary logic of legal certainty, which has been leftin the cord since the 18th century. In this regard, inspirationcan be fruitfully drawn from the works of Luhmann onexpectations and on Fuller’s interactional view of law.

According to Luhmann, the individual is a being that formsexpectations towards the highly complex world that surroundshim/her and who must face contingent and complex possibilitiesof action : « complexity in practice means compulsion toselect, contingency means danger of disappointment and thenecessity to take risks ». Even if « expectations can be formedand stabilised in relation to the environment », the problemcomes from the fact that other individuals also formexpectations. Everyone tries to guess the others’ expectations,which creates a « double contingency » : « the price to pay for

23

adopting stranger’s perspectives is unreliability »81. Theexistence of a rule can make « the orientation towardexpectations unnecessary » as it « further absorbs the risk oferror arising from expectation or at least reduces it ». Thanksto the rule, it can be said that the deviant was wrong :« Rules thus unload complexity and contingency fromconsciousness »82.

Luhmann makes a distinction between cognitive expectations (onelearns from his/her mistakes) and normative expectations (whichremain even if someone acts against them) : « Cognitiveexpectations, then, are characterised by a not necessarilyconscious preparedness to learn, whilst normative expectationssignify the determination not to learn from disappointments ».He thus defines norms as « counterfactually stabilised behaviouralexpectations », their validity being independent of their respector realisation83. He goes even further by writing that the« function of the legal system may be defined as producing andmaintaining counter-factual expectations in spite ofdisappointments »84. In a similar way, Luhmann argues thattrust is a mecanism for reducing social complexity, that can beillustrated by the simple example of the fantastic mutual trustthat is required to drive on a highway85.

Fuller criticized authors such as Austin, Hart and Kelsen forseeing in law a « unidirectional exercise of authority »instead of an « interactional process »86. The functioning of alegal system « depends upon a cooperative effort – an effectiveand responsible interaction - between lawgiver andsubject » and rests on the existence of a « relatively stablereciprocity of expectations between lawgiver and subject »87.Obedience to rules is meaningless if the authority will not payattention to them but, equally, the rulemaker will not have81 N. LUHMANN, A Sociological Theory of Law, London-Boston-Melbourne-Henley,Routledge, 1985, pp. 24-26.82 Ibid., p. 30.83 Ibid., p. 33.84 N. LUHMANN, « Operational closure and structural coupling: thedifferentiation of the legal system », Cardozo Law Review, 1991-1992, p. 1426.85 N. LUHMANN, La confiance. Un mécanisme de réduction de la complexité sociale, trad. parS. Bouchard, Paris, Ed. Economica, 2006.86 L.L. FULLER, The Morality of Law, 2d ed., New Haven and London, YaleUniversity Press, 1969, p. 223. 87 Ibid., pp. 209 and 219.

24

much incentive to accept the restrictions imposed by the Ruleof law if he knows that his subjects are neither ready orcapable of obedience88.

While Luhmann draws attention to the importance of expectationsfor human beings and on the role that law and trust can play inreducing social complexity, Fuller underlines the« interactional » nature of law. Once its is acknowledged thatthe law does not have the innate ability to determine itsapplication fully in advance, the recipients of the standardwill continue to obey as long as they have confidence in thefact that the authorities will respect their expectations basedon this standard. Trust plays - and should continue to play - adecisive role in any legal system.

It will come as no surprise that the author argues for legalcertainty to be mobilized as a priority for individuals(subjective logic) rather than in favor of the Union (politicallogic). Legislators are elected by the people, with a certainmandate and as such must take responsibility for the legalrules they enact. The State’s respect for the trust of itspeople represents the counterpart of its power to unilaterallydetermine the rules. It must be clearly emphasised that thisproposal is not about the substance of the judgments but aboutthe reasoning of the Court. In other words, it is suggestedthat when priority is given to the interests of the Union, itshould be supported by arguments other than legal certainty(either the effectiveness or the Rule of law of the Union).These arguments should have been used to justify the Orlandocase, outlined above, not legal certainty.

As regards the substance of the case, a combination of bothinterests (Union and individuals) where possible, could providea perfect compromise. If the Court adopts a time limitationwithout any legal basis, it could refuse to apply it in theindividual case before it and to the other procedures with thesame object and introduced on the day of its judgment (asolution already adopted when the Court limits the temporaleffects of its decisions). The same reasoning could be appliedto situations where the Court chooses a uniform interpretation

88 L.L. FULLER, The Morality of Law, 2ème éd., New Haven and London, YaleUniversity Press, 1969, p. 219.

25

of a regulation that suffers from linguistic divergences. Theunfortunate M. Röser, while facing criminal proceedings inGermany, was also subjected to an interpretation of a Europeanregulation according to its Italian translation. Would it nothave been possible for the Court to make an exception in hiscase, to protect his expectations, legitimately founded on atext in his native German ?

An argument for the principle of legal certainty to protect thesubjects of law in a system where the Rule of law applies, doesnot advocate that they take advantage of its protection withoutobligations on their part. Instead, subjects of law are in partresponsible for their own« legal certainty » and legalauthorities should expect that they have an awareness of andoverall respect for the rules that apply.

In this context, it is interesting to refocus on the famousadage that "no one should ignore the law". Traditionally it wasused to assert that the error of law was without effect. Inthis sense, it falls under the « political » logic, beingprimarily targeted at the effectiveness of measures taken bythe Prince. Whether or not the addresse of the norm actuallyknows it is immaterial. However, in an interactional conceptionof law, the adage acquires another meaning, which breaks theauthorities' obligations regarding legal certainty down intotwo elements : the first requires the authority to ensure that"everyone knows the law" by establishing the necessaryconditions for the recipients of the law to know about it. Thesecond – less consensual- element requires the authorities togive proper consideration to the legitimate expectations formedby the subjects of the rule. It requires not only congruencebetween law and its application (it becomes the responsibilityof the authorities not to "ignore" the law) but also that theauthorities take into consideration the expectations formed bythe subjects of law on the basis of what seems to them to bevalid law (if that perception is reasonable and caused by theauthorities’ behavior). It is important to underline that"proper consideration" does not automatically mean to follow upon these expectations, since other legitimate interests may beat stake and prevail. It is nonetheless argued that they shouldbe given the place they deserve in the judge’s reasoning.

26

The ECJ and GC should therefore build their reasoning aroundtwo distinct steps (which they often do albeit implicitly).They should first check whether the factual circumstances ofthe case justify application of the principle. This means, inbrief, to determine whether the « European social contract »has been disrespected from the point of view of the reliabilityof law. This could arise, for example, when the authoritiessuddenly modify a constant administrative practice, without anyprior warning. The second step comes into play, if there is aviolation of legal certainty, by requiring the ECJ to strike abalance of interests and assess whether the interest in legalcertainty must prevail over legally protected interests.

Contrary to what one might initially think, accepting ajudicial ‘balancing of interests’ in the examination of legalcertainty can be a way to restore its credibility orlegitimacy. Rather than find, almost invariably, that theprinciple was not violated because the alleged damage wassupposedly predictable, the Court may recognize the legitimacyof the defendant's claim but decide that, in the case at hand,it can not succeed.

Conclusion

To conclude, it is interesting to note the strong relationshipbetween time and legal certainty. Time is indeed at the heartof most requirements of legal certainty (predictability, non-retroactivity, transitional measures, etc.). If time is asocial construction89, it is still necessary to determine whichconstruction should prevail to apprehend the principle of legalcertainty.

One of the missions that the cartesian paradigm wished toentrust law with was to master Time, from the perspective oflegal determinism. The cartesian logic of legal certainty tendsto immobilize the past to secure the future. It prohibits anyretroactive effect, glorifies vested rights and promotesabsolute respect for precedent. It assumes a uniform andcontinuous temporality while the law must apprehend a realitythat is neither uniform nor continuous. The fiduciary logicinsists instead on an unfixed temporality that must be built.

89 F. OST, Le temps du droit, Paris, O. Jacob, 1999, pp. 12-13.

27

While the cartesian logic assumes the ability to guarantee thelegal consequences in advance of any future event, thefiduciary logic ensures that legitimate expectations formed apriori will be taken into account retrospectively. Such aperspective is, in this author’s view, both more realistic anddesirable. It is essential to preserve, or at least give achance to, human responsibility and human freedom.

28