1 Introduction

Syntax of texts of EU law has attracted little research attention. So far, the most extensive analysis of syntax-related issues in texts of law (in general) was presented in the works representing or reflecting on structural semiotics. Jackson [22], evidently inspired by theories presented by Greimas [17] and Carrión-Wam [10] argued that any legal discourse (including law itself) was based on a cognitive model which made recourse to syntactic, deep semantic and pragmatic structures specific only to law. At the syntactic and semantic level, these structures were conceived to make possible legal propositions which involved the use of a small, yet distinctive, group of modalities (such as English should, may, must) which give law-expressing sentences their significant pragmatic value. These studies found out that the bottom-line factor which made it possible for such propositions to express the law was a non-language-originated factor—the authority of their originator (the legislator) or their primary addressees (such as, especially, judges) to, respectively, claim that their utterances are representations of legal norms, and verify, that these utterances are indeed the law.

In the existing body of literature, legal syntax is conceived to be significantly narrow form of general syntax. Greimas argument related to syntax is somewhat similar to that presented by Carrión-Wam, as it also emphasized the value of a narrow form of legal syntax in coding a given text as a text of the law—the one apt to produce legislator-intended effects (in the process referred to as “production juridique”). Yet, for the two researchers, syntax does not have any predictive value in judging whether a given text is indeed the text of the law. The decisive factor is only the fact that the text considered has originated from someone authorized to make law, again being a non-text related consideration.

Some worthwhile investigations, albeit rather rudimentary, have been undertaken within the plain language movement, which sought to the language of law a standard which would be sufficiently clear to ease (and measure) compliance (e.g. Adler [1], Felsenfeld [12]). For quite obvious reasons, these studies notice the value of syntax in achieving subjective or objective law clarity. The former form of clarity reflects the communication standards between respective, concrete parties involved in the communication process; the latter one is concerned with an “objective” measure of clarity of that communication (measured by reference to some uniform benchmarks).

Limited relevance of a general benchmarking issue for understanding and practicing law clarity was attempted to be overcome by the normative concept of cataphaticity of the law. This concept was about achieving the quality standard of communication of the law significant for the production of an adequate level of accessibility for its primary recipient(s). Thus, it made possible to depart evident contingency of subjective setting of communication as well as from the orthodoxy of the objective concept of clarity [32]. The core concept of cataphaticity—accessibility of texts of law to its addressees, was defined by reference to Higgins’ argumentation [20] which emphasised that communication would become understandable only if it has identified communication competences of the counterparties involved in it and achieved the required match between these competences. Relatively extensive studies on accessibility barriers of legal texts appear to fall within this area of study. They emphasize that syntax is an important contributor to such barriers whenever complex sentences (nominal and/or hypotactic structures of sentences) are used in a text created for special purposes (such as texts of law are). These studies acknowledged that such a complexity required an unrealistically inflated cognitive processing skills of the people involved in the communication process [34]. Importantly, they also noted that such skills were difficult to achieve because text precision (so much needed in texts of law) was highly context-dependent [4] and that textual complexity also necessitated language economisation and deagentization [13].

Some arguments worthwhile for syntax analysis were formulated within the studies of language policy where status, corpus and acquisition practices are investigated (see e.g. Hornberger [21], Tognini-Bonelli [38]). Within this field of study, status is meant to reflect social functions attributed to the language in its various social contexts, whereas corpus can be construed as the legitimate scale and scope of changeability of language as a factor of adjustment to social context in which language communication occurs. Acquisition is defined as “the allocation of users or the distribution of languages/literacies, by means of creating or improving opportunity or incentive to learn them, or both” [20]. Corpus studies contribute to the studies of syntax of texts of law when it comes to distributional analysis of words combination. They fall short of effectiveness when it comes to the analysis of syntax considered as a potential signalling system of legal genre or the mechanism of creating norms of behaviour (see e g. a representative example of Gożdż-Roszkowski [16]).

Regardless of uncontested merits of these studies, they were not that much concerned with syntax or—even less so—with style of texts of the law as with syntax-semantic quality of utterances (i.e. legal provisions) meant to effectively convey their primary content, i.e. norms of the law. If at all, syntax issues transpired in this research in a rather secondary, albeit important, recommendation to make sentences of texts of the law as short as possible and thus—presumably—understandable to people to which they are addressed [12].

The style as such has never attracted such an attention of legal researchers which would expand beyond their general interest in achieving something which they would consider “a good style” of legislation. Such a standard is most commonly defined in rather semantic terms as something which is a result of clarity, precision and unambiguity of expression of norms of behaviour (see a review of older concepts in Mellinkoff [31]). Notwithstanding, these studies raised an argument important for this study, that the “style” was a result of specific syntax and semantic choices in a production of a given text meant to produce a text of good quality, i.e. the text which meets functionality, reliability and even aesthetic standards. Functionality standard will be met whenever the utterance of law has pragmatic impact adequate to the legislator–actant’s intentions; reliability can be defined as consistency of that impact in various situations where the utterance of law can be applied; the aesthetic standard will be met when the utterance of the law complies with literary concepts of an acceptable formal construction; in a more sophisticated form of the latter claim, legal text (as any other text) will be deemed to meet aesthetics standard, when it smoothly represents the thought without attracting much attention to itself, or—in terms of semantic transparency concept of language morphology—when it makes it possible to easily foresee the meaning of utterances by referring it to a larger language structure relevant to the text considered [6].

The syntax-semantic choice is made within the boundaries of the general language system, i.e. Saussurian langue—which denotes operational paradigm of a given language system [36]. In the case of law, that language system is limited as the creators of texts of the law cannot use all the vocabulary and syntax available for any assertion in a given language, but rather have to use only a limited number of language arrangements. In other words, they have to significantly reduce semantic and syntactic choices to meet the rather rigid standard of legislation-making. Such constraints do not only arise within just one piece of legislation, but emerge in the entire legal system. Thus, from this point of view, language of the law should be considered anaphoric, i.e. constantly making recourse to the already existing body of the law. This systemic feature of legislation has an important bearing on the style (and, consequentially, syntax) of the law as any previous legislative choices contain the scope of legitimate legislative action when it comes to its syntax and semantics aspects. From the point of view of text studies, this argument can be re-formulated to mean that the existing body of law is a “material” for any new systemic elements. Following Kayser [24], the “material” (der Stoff) can be defined as any element referred to the relevant text (i.e. serving as a signifying point of reference) yet existing outside of that text, and having its own tradition as well as its own fixed mode of processing, its own timing and space location.

There is an important reason for the desperate failure of fully-fledged investigations concerned with syntax and style of law. Above all, what counts in any text of law is its meaning or sense. This contention should indeed drive any analysis more towards semantics than towards syntax of legal texts. Yet, adequate comprehension of the language comes only together with its grammar structure, thus compelling syntax investigation to be performed on equal footing with semantics.

This study has two main goals:

  1. (a)

    to investigate whether there is one model of syntax used in the multilingual (expressed in 24 official languages) EU law;

  2. (b)

    to verify whether syntax contributes to the uniformity of EU law.

The latter goal is indeed the one which immediately calls for an identification of how syntax stands in the overall model of achieving EU law unity.

2 Basic Propositions About the Legal Text

In this study we make an important choice of a theoretical convention initiated by Wróblewski [43] and reflected in more recent literature (e.g. Kurzon [26]; Trosberg [39]) distinguishing the language of the law from the legal language. According to this convention, the former is the language in which legal texts are expressed: consequentially, this language originates only from the—conceived in abstract, theoretical terms—legislative authority. In contrast, the legal language is a meta-language of the language of law, i.e. the language in which assertions (written or spoken) about the law (i.e. about the sources of law) are made by those who reflect on any text of the law the Consequentially to this contention, only the language of the law is genuinely and directly performative, i.e. it is intended (and understood to intend) to cause specific behaviour of those to whom it has been addressed. In contrast, the legal language is used in any form of language not containing the law but rather reflecting on it, e.g. with the intent to inform about the law or to persuade to accept some opinions about the law.

It is astonishing that all concepts of “text of the law” are either largely intuitive when it comes to the identification that any “text” (considered as it is experienced, often with no clear referral to an act of its promulgation) indeed contains “the law”. Yet, it is important to note that the text informs its readers about its specific law content because of its specific vocabulary and specific syntax.

Yet, there is no much agreement about what text really is in the academic community. According to Eco [11]:

“The so called signifying chain produces texts which carry with them the recollection of the intertextuality which nourishes them. Texts generate, or are capable of generating, multiple (and ultimately infinite) readings and interpretations (…) Signification is to be located exclusively in the text. The text is the locus where meaning is produced and becomes productive (signifying practice). With its texture, the signs of the dictionary (as codifying equivalences) can emerge only by a rigidification and death of sense (…) A text is not simply a communication apparatus. It is a device which questions the previous signifying systems often renews them and sometimes destroys them.”

Any text, including texts of the law, is a discursive or linguistic practice deploying some vocabulary [8]. As such, it is composed of “performances that have the pragmatic significance of assertions, which on the syntactic side are utterances of declarative sentences, and whose semantic content consists of propositions” [8]. Only with the established relationship to one another, they assume sense which can be conveyed in the process of communication. In what Brandom calls “the iron triangle of discursiveness”, the pragmatic realm of a discourse is about asserting, whereas the semantic realm is about proposition contexts and the syntactic realm about declarative sentences [8]. The necessary pragmatic significance of assertions must include practices of asking for their reasons and about explaining these reasons by counterparties involved in the communication. Thus, the process of communication always involve those who—by virtue of an applicable social convention—are obliged to give a reason for their assertions (or utterances—conceived for the sake of this article to have a broad meaning encompassing also writing representations of thoughts) and those who, on the other hand, are entitled to request the explanation of reasons.

The decisive aspect of any textual utterance containing the text of the law is about the nature of apocatastasis, i.e. the notional, yet reconstructible “presence” of the law-utterance originator in any situation where his/her uttered text is referred to in order to eventually understand its content (in some contingencies leading to its application). Apocatastatis situation is usually quite diligently designed by the utterance originator with respect to its primary intended addressees. Thus, with respect to the law, the legislator usually makes every possible effort to set the utterance semantic-syntax structure in the manner making it possible for the primary addressees to properly (i.e. in line with the legislator’s intentions) conceive it as a text of law, interpret it, and apply it. Moreover, apocatastatis situation conceived (and designed) by the legislator is usually “informed” about the law-relevant scope of application of the piece of law concerned so that it is delineated apart from non-law relevant situations, such as e.g. the ones where the text of law is being learned (e.g. by anybody interested in the general content of the given text of law or by students of law for an exam) with no immediate intention of its application.

In the law-relevant setting, apocatastasis situation must consider the text of law to be a piece of stabilised in a written form strategic communication having specific pragmatic value, i.e. intended to achieve some social goals and/or objectives normally not achievable (or achievable, yet at a much higher social cost) within any other social system except law. This aspect of texts of law is also meant to contain the number of their acceptable interpretations—quite contrary to what Eco maintained with respect to texts of general character.

In other words, apocatastatic situation sets the framework for a specific Sprachspiel in the sense assigned to it by Wittgenstein [42] (see also [35]). Namely, socially meaningful confrontation with the text involves significant presuppositions determined by the language, communication actions, and a general social context involved in this situation. Such presuppositions are available only for people who already know the relevant elements (language and non-language) of the social code (Sprachregel) employed in it and defining “life-form” (Lebensform) of the social practice in which this code is meant to be used. Thus, Sprachregel is a logical link between the utterance (in any socially meaningful form, including text with its specific syntax) and its understanding in a context—considered in general social terms as well as in terms of a given situation where it occurs. As a result, it is concrete and pragmatic category which is stabilised as to its conduct and outcome by widely accepted and known institutions, practices and techniques of behaviour (including, obviously rules pertaining to syntax and assigning logical value to it) which—in the same time—are suitable to verify the utterance considered as understandable (at the least) or conveying socially-significant content (at the most).

According to Levinson [27] any socially meaningful interpretation (decoding) occurs only with regard to something which meets the following conditions:

  1. (a)

    it has been communicated intentionally;

  2. (b)

    by force of a broader (than that “something”) social convention, it has a specific form adequately associated with the content of the communication;

  3. (c)

    that form draws from a set of features granting it contrastivity in relationship to other communications;

  4. (d)

    in the same time, that form follows systemic grammar.

Lyons [29] argues that the communicative context is a fully realised element of communication. That realisation encompasses identification of the following aspects of an utterance:

  1. (a)

    its social role and structure;

  2. (b)

    its setting in time and space;

  3. (c)

    the degree of its official status;

  4. (d)

    the means of its conveyance;

  5. (e)

    its content:

  6. (f)

    its realm which determines its register.

Thus, in the realm of pragmatics studies represented by Levinson and Lyons, the form of utterance is of high importance as an element of sense-decoding (interpretation). Yet, the two researchers emphasize that the form cannot suffice in this process and requires some presupposition, it is not convincing that the infra-textual or extra-textual elements identified by the two outstanding pragmatics researchers suffice to explain these presuppositions. Especially, they miss an important point that the reason or objective of the reader of a text has important bearing on how it is interpreted as it represents an important setting for other legitimate presuppositions the reader can have with respect to that text. Yet, the arguments of Levinson and Lyons highlight an important function of the form—i.e. that, itself, it conveys an important massage to the reader about the intended by its originator function of the text. For Lyons these there are two formal elements of that nature: the degree of the official status of the text and its register—which represents a rather defective concept as—in my opinion—register is enough to grant some, selected by the originator, degree of officialness. Regardless of this issue, register of any text of law is produced by what can be otherwise perceived as “legal style” with its distinctive “legal” syntax.

To explain this point one should bear in mind that text register represents a functional relationship between the utterance and its concrete situational context deliberately intended by the communication originator. In very elegant terms used by Halliday [18] we should talk here about an instantiation of a language system (i.e. the language potential as a meaning-making resource) in the form of text (understood to be a specimen of intentionally reorganised language structure or an artefact having a material counterpart to what it represents in abstract). Instantiation can be defined as a practical realisation of the relationship between the text and the language system. That realisation occurs through the cline of instantiation being a continuum between the overall potential and the particular instance of the text. Depending on the intended function of the text (and thus its formal pattern) and on the reason granted to the text by the reader, each text has its own possible models of instantiation which can be referred to as its register. Thus—as Halliday argues, “registers can be represented as a particular setting of systemic probabilities” [18].

Halliday’s concept brings us even closer to the identification of what text style is when it highlights that register can be interpreted as a characteristics of the language used in a given text, differing with regard to its field, mode and tenor [19]. The field is delineated (or granted) by the main content of the text determined also by the intended and realised social setting in which it functions as well as the actions it originates; the mode is the function of the text in that setting determined by the choice of the method of communication, the communication channel, the type of expression, as well as its rhetoric colour; tenor reflects the type of social interactions relevant for the text. As a result, it is possible to say that the register reveals the manner in which the communication originator (the actant) intends to treat the persons to which his communication is addressed.

Many researchers, especially Biber [7] argued that register was not necessarily the defining feature of a genre of text. In contrast, Kurzon [26] held that the text genre reflected a set of expectations the reader has about the text considered. Yet, this does not necessarily mean that linguistic aspects of utterance, such as especially lexis and syntax (which contribute to register) do not grant the text its recognizable character. Kurzon’s analysis indicated that legal genres might be clearly differentiated according to purpose, and—what is worth emphasizing—that legal register was also distinguishable and granted the legal text its genre. Yet, Kurzon also emphasized that register was the least powerful factor determining the legal genre—more powerful were the contextual factors: the situation in which the text is referred to and the purposes for which it is referred to. Such an argumentation was also strongly supported by the studies presented by Rassmussen and Engberg [34] who conceived legal text to be a strategic social occurrence of communication with strictly defined roles distributed among those involved in it where contextual factors play significant role in reading the text as the “text of law”.

Style should be somewhat related to text register as it should reflect the register field, mode and especially, tenor. Voermans [40] insists that style is a significant component of the mode aspect of register. He also argues that style of any text of law can, therefore, be considered a function of:

  1. (a)

    the wording, i.e. the content of the text of law produced by means of a concrete use of language, including any definition conventions adopted in that utterance;

  2. (b)

    the structure, i.e. any internal organisation of that content and mutual references respective parts of that content make with each other;

  3. (c)

    the superstructure, i.e. the reference the text consider makes to other texts of the same kind (genre) and the hierarchical organisation of all these texts of law;

  4. (d)

    legal-cultural identity of that text resulting from its reference to the sources of law in a given legal system.

Setting a text as a text of law is granted to be an intentional activity of its originator (the actant). Levinson considers such an operation to occur at two levels: one level is metasemantic and is concerned with the placement of a given utterance in a much broader communication system, thus resulting in what Levinson refers to as genre location (effected as a result of the realisation of what Levinson calls categorical intentions); another one is about giving the utterance its logical sense (effected as a result of semantic intentions) [27].

The reader knows of any legal character of the text in front of him also because of contextual references he or she is able to make to other prima facie relevant social factors which he or she is aware as pertaining to law (legal norms). These factors, if at the end of the day, they are to be verified as relevant, have to be somehow qualified as indicatory (or, quite the opposite—non-indicatory) that the text indeed contains specific content of law and not any other content (i.e. the content expressing statements or rules of a different than law legal system). Kaplan [23] argued that at least some elements of any utterance contained elements which are insensitive to any change in the context. They are referred to as “indexicals” and they are used to make anaphoric argumentation within a distinctive social system of reference i.e. the argumentation which makes its complete sense by a reference to the already existing, characteristic codes of a given discursive area, such as e.g. the law. The law is especially pervasive in this function because it requires any of its new narrative (or provision) to be formulated in a specific language emulating the already existing elements of the broader discourse narrative. Such a coerced feature of law narrative is indeed necessary to prevent systemic entropy of law, i.e. its potential indeterminancy as a system which may especially result from any polysemy of its content and—what is important to stress—form. Such an argument stands in sharp contrast to an argumentation that law represents a somewhat natural polysemic narrative (see e.g. [30])—which should be contested as inaccurate at least whenever it relates to the description of a regular behaviour of a regulator or a regular assumption of law users.

3 Multiligual Nature of European Union Law and the Question of its Unity

European Union law is published in 24 official languages and applied in 27 legal order of EU member states.Footnote 1 Texts of law published in the Official Journal of the European Union in all of its official languages are considered to render perfectly equivalent versions of respective EU pieces of law.Footnote 2 This implies that that there are not controlling and controlled versions of that law in these official languages. Moreover, the Court of Justice of European Union’s jurisprudence emphasizes the unity of meaning of these different official language versions of respective acts of EU law.Footnote 3Footnote 4 It is of some general significance that EU law represents a legal system autonomous from respective national systems of EU member states.4 This autonomy is pronounced also in EU law semantics—as its wording is meant to convey unique EU meaning unless it itself allows that this wording should refer to or mean anything which is defined in national laws of EU member states.Footnote 5

From the language of law perspective, EU CJ’s concept of unity can be interpreted to have three aspects:

  1. (a)

    semantic meaning unity aspect—meaning that provisions of law expressed in respective EU official languages should have perfectly equivalent logical value and meaning;

  2. (b)

    relative processual uniformity aspect—meaning that EU law should be applied in the same manner—at least to the extent prescribed in EU law itself, beyond that limit, relaying on national institutional and legal arrangements meant to render EU law as effective as it is required by it;

  3. (c)

    relative pragmatic uniformity aspect—meaning that EU law should have bear the same result in all EU member states,Footnote 6 regardless of any processual diversity allowed under the enshrined also in the Treaties (respectively, on European Union and on functioning of the European Union) principle of institutional autonomy of EU member states, i.e. their freedom to use whatever suitable institutional/organisation arrangements they have adopted (under their national laws) to properly (i.e. according to the EU standards) enforce EU law.

Considering all these characteristic material, processual, and contextual aspects of unity (in the pragmatic realm implying uniformity), it is obvious that the EU legal system is arranged in the manner aimed at achieving and sustaining it. Most importantly, the EU legislator does the utmost to achieve semantic (lexical) equivalence of EU law. How this semantic meaning unity is achieved and how it is supported or augmented by the CJ EU goes beyond the subject matter of this article unless the sematic realm of EU law utterance blurs with the syntax so much that the two are difficult to distinguish—as it is especially the case of very relevant for law sentences in which modal verbs are used.

One should notice that the most important, bottom-line equivalency the EU legislator aspires to achieve is definitely pragmatic. Pragmatic aspect of unity is decisive for making EU law an effectively and efficiently applied and enforced in the legal order where it is to enjoy priority of its application before any piece of national law (also in EU member states’ courts and national administrative bodies). As it has already been argued, such an aspect of unity is achieved by using semantic means; it is also achieved by its very special means and methods of teleological interpretation,Footnote 7 where “spirit and objective” of the interpreted piece of EU legislation is to be discovered and become an essential premise for legal reasoning in a given situation.Footnote 8 The question is whether also syntactic or even style/register-related considerations (i.e. formal considerations) matter in achieving the EU unity/equivalency.

The very first argument which should be recalled in response to this question draws from formal features of EU law. These features include, so called, synoptic layout in which each and every a piece of binding law produced in 24 different official language versions is promulgated. This means that equivalent pages of the same EU legal act are reproduced in the Official Journal of the European Union in such a way that they look the same regardless of their EU official language. Thus, the same provisions reproduced in different EU official languages can be easily identified (or located) not only by their referential numbering (especially by articles) but also by their mere place in the legal act promulgated in the EU Official Journal. That synoptic layout may have some (albeit definitely not fully intentional) bearing on the formulation of respective legal acts as they tend to be of the similar length. Baaij [3] considers it an important trace that this painfully formal resemblance of reproductions of EU law in different official languages is achieved intentionally. It, however, safer to interpret the relatively similar lengths of EU provisions reproduced in different languages as a product of general resemblance of all Indoeuropean languages. As a result, there is no firm evidence that synoptic reproduction of texts of EU legal acts have any direct bearing on the syntax of utterances which have the form of provisions of these legal acts. In the language of the concept of corpus, this means that syntax is not very likely to be subject to any meaningful change/adjustment meant to achieve synoptic consistency of EU texts of law. It is more reasonable to argue that, in the context of each and every EU official language, cataphaticity motives prevail thus leading to some simplification and standardisation of both semantics and syntax.

The syntax failure to reflect synoptic mode of reproduction of EU legal acts can be explained by the fact that syntax is a very language structure-sensitive matter. As it has already been mentioned, in the language of the law, syntax often blurs with semantic arrangements (see also Baaij [3]), as in many languages sentences expressing normative expectations make recourse to modal verbs. This solution, essentially because of semantic reasons, cannot produce absolute equivalency, at least where it is conceived in purely linguistic terms. In order to explain this point we should remind that modal verbs have a very important meaning-creating function in the so called normative lexemes. According to Finlay, such lexemes are utterances which “refer to probabilistic relations in which things stand to particular „ends” or potential states of affairs that vary from content to content” [14]. The presence of normative lexemes can be considered a purely formal element of utterances (i.e. the element which emerges at the syntax level of analysis). Regardless, any text containing such lexemes can immediately be recognised as the one which intends to regulate behaviour, and therefore, could be a text of law. Only a more contextual study of such a text (allowing for e.g. a verification whether the text contains binding rules), conceived to be a social act, can ultimately reconfirm such an assertion.

Widely accepted differentiation of prima facie normative lexemes (or—more specifically—modal verbs which grant these lexemes their character) distinguishes (e.g. Palmer [33]; Goosens [15]:

  1. (a)

    dynamic modal verbs, which—in the context of relevant lexemes—express ability or willingness to perform an action because of a supporting objectives circumstances located internally to the person considered;

  2. (b)

    deontic modal verbs, which—in the context of relevant lexemes—express acceptability of an action or an obligation to perform it—both with regard to its content and effect(s) where the source of this acceptance or obligation is located in the environment external to the person considered;

  3. (c)

    epistemic modal verbs, which—in the context of relevant lexemes—express, essentially not materially substantiated, opinion of the person considered that a certain phenomenon can likely occur in a given circumstances;

  4. (d)

    evidential, which—in the context of relevant lexemes—express materially substantiated opinion that a phenomenon occurs or is likely to occur.

Thus, a very important criterion of division of various modalities is the source of their binding power, which in linguistics is referred to as modal source (in German Modalfaktor). According to the definition of modal source offered by Bech (the researcher which came up with this concept), it is a “factor (…) which makes the content of the modal field necessary or possible” (German „…den faktor (…) der den inhalt des modalfeldes notwendig (…) macht bzw. ermöglicht”, [4, 5].

Semantic differentiation of normative lexemes, especially those most relevant for texts of law, transpires also in their specific logical values. Even an exemplary comparative analysis of these values of normative lexemes in Danish, English and German (all being genetically close to one another Germanic languages) indicates striking non-equivalence of many of them [8, 9], as shown in Table 1.

Table 1 Comparison of logical values of basic modalities (normative lexemes) in Danish, English and German
Full size table

Many normative lexemes in Danish, English and German have overlapping ranges of meaning but cannot be considered mutual equivalents when it comes to their logical value. Many do not have any counterparty in one or two languages considered. Yet, as it has already been raised, the usage of certain modalities in different languages can—at the formal, syntax level of text analysis—indicate that the text expresses legal norms (and thus, determine its genre and register). In other words, it is the formal (syntax) analysis which is conductive to more common statements about the meaning and social function of the text considered than any more substantive analysis of respective language versions of this text. Nevertheless, their presence cannot be decisive about the legal content of the text (i.e. its meaning). Moreover, as normative lexemes expressed in different languages (such as Danish, English and German in the example considered) cannot always be mutual counterparts. In consequence, unity of EU law cannot rely only (if at all) on them.

The said extensive conclusion has quite important caveats. Namely, similar to the German linguistic practice of expressing illocutions of normative character can be encountered in Romanic languages (where e.g. English shall, Danish skal or Swedish skall find their non-perfect equivalent in e.g. Spanish form deberá or Portuguese dovem). However, when it comes to Slavic languages, no such analogy exists, altogether. In such languages as Polish or Croatian, or in many cases French (the Romanic language), such illocutions are expressed in legal provisions not by using modalities (normative lexems) but simply by recursing to declarative sentences in present tense. To illustrate this point, let us consider the following example of just one provision of EU law, namely Art. 37(1) of Regulation of the European Parliament and the Council Nr 1215/2012 of 12 December 2012 on jurisdiction and the recognition of judgments in civil and commercial matters (recast)Footnote 9:

The comparison of the first part of the first sentence of Art. 37(1) of regulation Nr 1215/2012 indicates that even the sequence of its constitutive lexemes is different in different language versions. The syntax structure used is the same as in national legislation. The differentiation transpiring on the grounds of EU law is obviously determined by the differences in the generic syntax structure of the respective official languages considered as well as a specialised syntax structure of used to express texts of law in these languages. The comparison in Table 2 presents this differentiation. In this presentation, the lexemes of the same syntactic function (and therefore the same logical/semantic value) are marked in the same colour. Thus semantic differences are exposed by different positions of the lexemes marked with the same colour.

Table 2 Syntax of the provision of Art. 37(1) of Regulation of the European Parliament and of the Council Nr 1215/2012: Comparison of selected Slavic, Germanic and Romanic languages
Full size table

The analysis presented in Table 3 substantiates that the reproduction of the considered provision of Art. 37(1)—itself being an attributive sentence with extended complement—in respective EU official languages render the same logical value. Yet, obviously, these sentences differ with regard to both semantics and, especially, syntax. The most striking (and, indeed, untreatable) syntax/semantics difference which can immediately be detected among those different reproductions of Art. 37(1) arises from the lack of articles in Slavic languages (with just one exception of Bulgarian). For this reason, in the text quoted, indefinite English article „a”, Danish and Swedlish „en”, Spanish „la”, „un”, „una”, Portuguese „as”, „num”, „uma”, and French “la”, “un”, “une” render a bit more information than the Polish and Croation reproductions of this provision.

Table 3 Sequence of the first part of the provision of Art. 37(1) of Regulation Nr 1215/2012 in selected Slavic, Germanic, and Romanic languages
Full size table

More fundamental and purely syntax differences occur with respect to the construction of sentences uttering art. 37(1) in respective EU official languages. They occur only in a narrow, yet significant area where the verb, the adjective, and the attributive operate, i.e. with regard to the element 3–5 in Table 3 with respect to each and every language considered (in the English version with respect to the syntagm „…wishes to invoke in a Member State…”). In that English version, the sequence is 3–4–5:

„(3) „…wishes…”, (4) „…to invoke…”, (5) „…in a Member State…”.

In other Germanic languages considered (i.e. in Danish and Swedish), English sequence 3–4–5 instead becomes the equivalent of the English 5–3–4; for the Slavic languages versions, the English sequence is 3–4–5 in Polish, and 3–5–4 in Croatian; the Romanic versions considered follow the English sequence pattern (3–4–5). This conclusions cannot be considered absolute for the Scandinavian languages and, especially the Slavic languages as their syntaxes allow some types of inversion—already visible when Croatian and Polish texts are compared (where Polish 3–4–5 version is not an inversion) For, respectively, Polish and Croatian, a possible inversion (with respect to the original wording used in the legal act) would be:v

Polish „..chce w Państwie Członkowskim powołać się” or Croatian „…se želi pozvati u državi članici” (i.e. which, for Polish, would represent exact equivalent of the English sequence 3–5–4 but for Croatian restoring the natural, not inverted, 3–4–5),

or.

Polish “…w Państwie Członkowskim powołać się chce” or Croatian “…u državi članici pozvati se želi ” (which represents equivalent of English 5–3–4, but in both languages sounds archaically),

or even unacceptably from the point of view of the legal tradition and/or rules of reproduction of legal provisions:

Polish “…powołać się w Państwie Członkowskim chce…” or Croatian „…pozvati u državi članici se želi” (3–5—4).

Yet another syntactic possibilty to produce a proper wording of this phrase in Slavic languages arises from the fact that the Slavic equivalent of English verb “invoke” is the reflexive word “powołać się” in Polish “želiti se” which offers even more possibilities for inversion, such as:

Polish “…w Państwie Członkowskim się chce powołać…” or „…się w Państwie Członkowskim powołać chce…” and in Croatian „…u državi članici se želi pozvati…” or „se u državi članici želi pozvati…” .

— where these syntactic forms are grammatically correct, yet do not comply with the legislative standards. Thus, if applied, they would produce a misleading register and, as a result, possible misconceptions of readers as to the genre of the text—especially where no addition, contextual information would be available to them.

The broader possibility of syntax inversion in Slavic languages which is possible without breaking not only syntax rules but (in many cases), but also genre/register (specific for the law) ones reflects their higher level of compositionality understood to be the language ability to express the same meaning using diversified syntax structure.

In Romanic and Scandinavian Germanic languages such wide possibilities to modify the sequence without impairing its general or legal syntax quality is narrower. In Danish, there is just one (seemingly) possible way of modification, i.e. to have it, instead of the present form „…i en medlemstat ønsker at påberåbe sig…” (i.e. as the equivalent of English 5–3–4) as „…ønsker at påberåbe sig i en medlemstat…” (i.e. as the equivalent of English 3–4–5). Yet, even that modification is impossible because it is destined to produce anacoluthon sentence when put together with the entire remainder text of the provision. To put it shortly, no modification could make a viable sentence here—regardless of whether produced in legal register or a general one.

It is of utmost importance to note that some of the syntactic sentence arrangements are coerced in the sense that they reflect necessary rules of syntax of a given language. Some, however, are more reflective of the legal genre in a given language. A good example of this phenomenon is the Polish version of the analysed sentence” (“…przedstawia (a) odpis orzeczenia spełniający warunki niezbędne do stwierdzenia jego autentyczności”, i.e. literally „…presentsFootnote 10 (a) a copy of the judgment satisfying the conditions necessary to establish its authenticity”) whereas a non-legal syntax could have allowed that sentence to be arranged in (non-legal) Polish by inversion, in a rather strange in the English syntax way: “przedstawia (a) spełniający warunki niezbędne do stwierdzenia jego autentyczności odpis orzeczenia” (i.e. literally “…presents (a) satisfying the conditions necessary to establish its authenticity a copy of judgment” or even impossible in Danish “fremlægge (a) opfyldende [alternatively der opfylder] de nødvendige beingelser for at fastslå dens ægthed en kopi af retsafgørelsen" (with an analogic anacoluthon in Swedish). The inversion is also not apt to produce a proper sentence in French.

The only syntactic modification of this sentence in any EU official language could be to give it a passive voice format. Yet, such format is not accessible in either legal text considered because it would go astray of syntax-related legislative traditions. The tradition is constantly reinforced by anaphorisation in the creation of law—the legislative practice which is based on the continuous referral of legislators to the already existing law. Such an anaphorisation is most important with regard to the semantic aspect of law. Yet, it is the most prevalent with regard to law-specific syntactic structures where one can observe the most radical orthodoxy in their repetitive application.

All this is indicative that respective EU official language versions have fundamentally differentiated syntax—not only at a surface, but also when deeper structures of it are concerned. Moreover, the comparison indicates that the syntax rules applicable to legislation (including EU legislation) are stricter than these applicable to general genre language even with regard to the languages where there is relatively much more scope for syntactic manoeuvre. This limitation seems to reflect an intention to preserve illocutionary character of utterances, i.e. to make the language of law somewhat different from the general language. As a result, on the grounds of each EU official language use, specific genre of EU law is produced in the same manner as it is produced with regard to national laws.

The analysis presented above obviously reflects different levels of syntax compositionality (i.e. ability to produce proper sentences of logical equivalence by modifying the location of the lexemes constituting these sentences, see e.g. Szabó [37]; Kracht [25]) available in different EU official languages considered, which is a well-known phenomenon. Yet, it also exposes a new facet of compositionality—namely that the language of law puts constraints on the general form of compositionality of a given language. Thus, in languages of relatively higher compositionality (such as e.g. Polish or Croatian), inversions otherwise acceptable from the point of view of general syntax rules, may not be acceptable if they are meant to produce utterances of law.

The coerced, limited compositionality structures available to produce EU legal texts serve well the general purpose of achieving their equivalence. This desired effect is achieved because it reduces the risk that different degree of openness of texture (in German referred to as Porosität der Begriffe, i.e. the ability to denote whatever (see: [2], [41]) of different languages on the equivalence of logical outcomes of sentences they produce.

4 Conclusions

In the EU system of law expressed in 24 official languages, there is no single syntax system which would prevail or apply as a controlling any other language systems. Thus, each language version of EU law relies on respective language structure.

Syntax structures used in respective language versions of EU law are just narrower form of syntax structures of a general form of these language versions. They use patterns which grant legal text the relevant register (the mode of which is determined by legal traditions of respective EU member states) and which are limited to that register. Notwithstanding, syntactic structures of EU law produced in different official languages do not represent a single EU-wide pattern. This failure of multilingual EU legal system results from diversified mode of producing illocutionary value of legal provisions in respective EU official languages. The diversification arises from the language-legislative practice where, on the grounds of each EU official language, strongly idiomatic modal verbs (modal operators) and even special modal structures not containing modal verbs are used to express legal norms.

Syntax used in respective language versions of EU law are least apt to be uniformised in any way. Indeed, this conclusion shades an important light on the pragmatic, semantic and syntactic structures of EU law perceived as a body of law expressed in 24 natural languages. Namely:

  1. (a)

    in the pragmatic aspect, the EU law is most uniformised, because of the requirements of unity of application of EU law;

  2. (b)

    in the semantic aspect, the EU law is:

    • when it comes to the fundamental mode of reference—diversified because of 24 languages used;

    • when it comes to the mode of that reference interpretation—uniform because of the uniformity and autonomy of EU meaning;

    • in the syntactic aspect, the EU law is—generally speaking—not uniform; some, rather rudimentary, uniformity is somewhat likely to emerge with the intent to produce synoptic mode of legal texts (i.e. the same outlay of published texts); yet, there is no firm evidence that it is a definite motivation for reproduction of respective provisions in 24 EU official languages.

Relative uniformization of syntax of EU legal acts within the groupings of Germanic, Romanic and Slavic languages considered in this study appears to be the only implausible text uniformity format explained by historical affiliations of languages within each group. Deeper-structure syntax transpiring through the register of respective legal acts is also used to signal that respective texts indeed contain the law, yet there is no evidence of the EU-specific nature of that aspect of syntax. In other words, legal acts produced in respective EU official languages are offered in the register specific for acts of law to allow adequate reconstruction of apocatastasis situation conceived by the legislator.

Yet, at the syntactic level, this register is not EU-law specific. Instead, it is using the same structures which are used to produce national legislation. Thus, the scope of manoeuvre available to EU legislators in making texts of law are different in different EU official languages reflects differences between these languages when it comes to compositionality (higher in some EU languages and lower in some other). It also differs with regard to their law-specific compositionality, which can be construed as flexibility of language-specific syntax structures in expressing law without losing their genre/register producing ability.

Considering this, it is quite possible that the EU legislator takes this in element into account avoiding highly idiomatic syntactic or semantic structures, i.e. the ones which cannot be directly, and with no much “compositionality fuss”, reproduced from one language to another and vice-versa. This penchant is only reenforced by self-referential nature (or anaphorisation) of development of every legal system, i.e. the phenomenon where new pieces of legislation respect the existing body of law and even attempt to mimic its syntax structures as well as semantic arrangements—all to achieve its systemic consistency and coherence.

This recursive approach (i.e. the approach based on form and content anaphorisation and on using the existing body of law as a reference material for any new piece of law) might be one of the most important drivers of producing EU texts of law in national languages which already have a nucleus of EU-specificity or, in other words, a common EU register. This arguments, however, requires some further, more detailed study. All in all, the analysis presented in this study indicates that, there is no EU-specific syntax which would, as such, contribute to unity of EU law. Instead, unity is achieved in the area of semantics and pragmatics.