Tilmann Altwicker¹, Florian Geering², Daniel Gerber³ and Zhivko Taushanov⁴

1 Introduction

Legal reasoning is fundamental to the function and legitimacy of any court. In the broadest sense, legal reasoning refers to the process of providing justifications for judgments (and other legal acts) (MacCormick 1998). Legal reasoning is studied from a variety of angles: legal theorists examine the methods and the purposes of legal reasoning (Dworkin 1986; Raz 1999, 326–340; Alexander and Sherwin 2008), political scientists and sociologists study how legal reasoning reflects socio-political factors and attitudes (Braman 2009), and linguists examine how language and rhetoric are used in legal reasoning (Tiersma 1999).

The quality and depth of legal reasoning are central elements of the legitimacy of a court's judgment: comprehensive legal reasoning serves to justify a court’s decisions, providing clarity and predictability for stakeholders. When the reasoning behind a decision is transparent, it establishes trust between the court and the parties to a case, but also with the general public. The presence, quality, and persuasiveness of the reasoning is important for the parties’ willingness to comply with a judgment, and therefore for its legitimacy (Tyler 2006, 149, 161–165). Conversely, judgments with weak or opaque reasoning are likely to be met with skepticism, undermining the court’s authority and the perceived obligation to comply (Madsen et al. 2018, 212).

Methods, elements, and length of legal reasoning reflect cultural specificities. Thus, there exist different ‘styles’ of legal reasoning. Some courts have adopted a brief, ‘formalistic’ approach that centers on listing the sources for the judgment (this has been called the ‘French style’ of legal reasoning); other courts (e.g. German and U.S. courts) provide more extensive reasoning for judgments and follow a more ‘substantive’ approach that is more policy-oriented and discusses the various argumentative positions (sometimes at length) (Von Bogdandy and Venzke 2014, 186-187). In spite of these differences, the interpretation of the relevant legal texts is a common starting point. This holds for both common law and civil law systems (Greenawalt 2018, 96). In both types of systems, the quality of legal reasoning depends heavily on sound legal interpretation.

Legal interpretation refers to the procedure of assigning meaning to legal texts and other statements with the intent to establish rights, obligations, and other legal consequences.⁵ Any textual statement requires interpretation: how a reader understands a text is shaped by their preconception of the text (Gadamer 1993, 57). This is particularly true for legal language with its ‘open texture’ (Hart 1997, 128 et seq.), which requires the reader to assign meaning to the words. By resorting to legal interpretation, the court turns arguments or reasons into legal reasons, i.e. arguments that can claim validity in the legal discourse or within the interpretive community (Stappert 2020, 34 et seq.). Legal interpretation serves as a fundamental component of legal reasoning, embodying the necessity to provide justifying reasons for a court’s ruling in any given case (MacCormick 1978, 196 et seq.). Furthermore, it is fundamentally argumentative and thus inherently inter-subjective (Feteris and Kloosterhuis 2009, 314–315; Herdegen 2020, para 1). It builds on and integrates additional information about the text (Cremer 2022, para 2), which is gained through different interpretive techniques (teleological interpretation, originalism etc.) that are used for normative texts.

In the case of the European Convention on Human Rights (ECHR), it is the European Court of Human Rights (ECtHR) that holds the ultimate (though not the only) power to interpret the Convention rights.⁶ Art. 32 of the ECHR extends the jurisdiction of the Court ‘to all matters concerning the interpretation and application of the Convention on the Protocols thereto.’ By virtue of Art. 45 ECHR, ‘[r]easons shall be given for judgments as well as for decisions declaring applications admissible or inadmissible’. The ECtHR has discretion as to how extensive the reasoning in each case should be (Brunozzi 2023). The literature has noted some variation in the legal reasoning by the ECtHR over time: whereas formerly the Court’s legal reasoning was narrowly tailored to the individual case, in the recent past, an increasing number of passages reflect general statements reminiscent of domestic constitutional courts (Grabenwarter and Pabel 2021, § 14 sub para 4).

While the ECtHR is the ultimate arbiter, two aspects make legal interpretation of Convention rights a complex task. First, interpreting human rights instruments is in general difficult due to the highly abstract nature of these rights. Convention rights are often vague (such as ‘respect for personal life’) and employ essentially contested concepts (such as ‘equality’, ‘freedom’). Second, the interpretive community is heterogeneous: there are many different actors that are involved in the interpretation process. Indeed, national courts of member states have their own obligation to interpret and apply the Convention (Sweet and Keller 2008, 24). In addition, there are applicants and their lawyers, domestic authorities and courts (on various levels),⁷ and even third parties such as NGOs, that may all propose diverging interpretations of the Convention, driven by their distinct agendas. In all this, the ECtHR, however, retains a guardianship role. This guardianship role imbues the Court with a unique mandate: to ensure that member states’ practices are in harmony with the Convention and to address any deviations that might arise. Given the diversity of legal traditions, cultures, and historical contexts among member states, the Court’s interpretative function becomes crucial in ensuring consistent application of Convention rights across varied jurisdictions.

This paper seeks to elucidate the practice of Convention rights interpretation at the ECtHR. We are interested in two research questions: first, what patterns in the case-law of the ECtHR emerge from its de facto use of the canonical techniques of legal interpretation? More specifically, we are interested in what techniques of legal interpretation are utilized by the ECtHR in relation to distinct Convention articles. Secondly, we explore the relationship between case outcomes and the employment of interpretation techniques. Essentially, we investigate if arguments grounded in these legal interpretation methods tend to be ‘winning arguments’ with the Court’s finding of a violation.

To this end, we present an analysis of ECtHR judgments between 2006–2011 (N = 8436), examining the techniques of interpretation that are used in each case and the Convention provisions being interpreted. Using logistic regression models, we also explore the relationship between interpretation techniques and the finding of a violation. The article contributes to the existing literature by providing empirical insights into Convention rights interpretation and into legal reasoning more generally. The study’s findings have practical implications for lawyers, as empirical knowledge can inform a more targeted argument construction. Furthermore, understanding how legal arguments are constructed in practice promotes the societal goal of transparency in adjudication. Most importantly, by tracing the patterns in the use of interpretive techniques by the ECtHR, the study sheds light on the rationality of legal argumentation, which is essential for ensuring its institutional legitimacy.⁸

The structure of this article is as follows: after this introduction, section 2 provides a summary of the legal rules pertaining to the interpretation of the ECHR. Section 3 delineates the methodology employed for data collection and analysis. In section 4, we discuss our findings on the use of interpretation techniques, their frequency, and their relationship with case outcomes at the ECtHR. Finally, Part 5 concludes by acknowledging several limitations of the present study and suggesting potential directions for future research.

2 Literature Review

This article focuses on the use by the ECtHR of the four canonical techniques of legal interpretation, namely textual, historical, systematic, and teleological interpretation.⁹ Through the use of these techniques, the ECtHR not only brings clarity and coherence to the Convention’s provisions but also furnishes justifying reasons for its rulings, reinforcing its role as the central authority in interpreting the ECHR. Despite its critical role, empirical insights into legal interpretation at the ECtHR remain scarce. The study that is closest to ours is Christian Djeffal’s work on ‘Static and Evolutive Treaty Interpretation’ which explores the concept of treaty interpretation within the framework of international law, focusing on the effectiveness and adaptability of treaties over time (Djefall 2015). Djeffal analyzed the interpretation practice of the ECtHR and concluded—on the basis of a quantitative analysis of 31 cases—that the technique of teleological interpretation is less often used by the ECtHR than expected, and that other techniques, like textual interpretation and subsequent practice, figure more prominently in the Court’s case-law (Djeffal 2015, 323, 356). His dataset was, however, too small to test for statistically significant relationships, especially with respect to individual Convention rights or groups of rights.

Relatedly, Daniel Peat (2021) explored the ECtHR’s utilization of consensus-based interpretation to manage an extensive array of interpretive choices.¹⁰ He suggested that the consensus doctrine is a response to the phenomenon of choice overload, which can impede interpretation of legal standards such as fairness, necessity, and proportionality. Through the analysis of 461 judgments by the ECtHR, Peat provided insights into the practical aspects of consensus identification and application in the realm of European human rights law, thus contributing to the legitimacy debate concerning the ECtHR and to our understanding of legal interpretation processes.

In an effort to train argument mining models for ‘Mining Legal Arguments in Court Decisions’, Habernal et al. (2023) annotated legal arguments including interpretation techniques in 373 ECtHR judgments. 30 of the 378 documents in the data they have made available¹¹ contain annotations that denote the presence of interpretations. This means that Habernal et al. found evidence of the use of interpretation techniques in 8% of analysed cases,¹² indicating that the ECtHR may not resort to interpretation techniques as often as it would be expected given the centrality of legal interpretation in legal reasoning (see above).

Most existing empirical studies on legal interpretation have focused on courts and tribunals other than the ECtHR. A study that is methodologically aligned to ours is Manley et al.’s ‘Mapping interpretation by the International Criminal Court’ (2023).¹³ They utilized a quantitative approach to systematically map the International Criminal Court’s (ICC) interpretative practices. They based their analysis on data collected from ten ICC case studies (Manley et al. 2023, 783). The authors noted inconsistencies and a potentially unjustifiable application of the Vienna Convention on the Law of Treaties (VCLT), indicating a possible deviation from the requirement of strict construction in Article 22 of the Rome Statute. Another empirical study is ‘The legal reasoning of ICSID tribunals – an empirical analysis’ by Ole Kristian Fauchald (2008). He found clear patterns in the ICSID tribunals’ legal reasoning. These patterns were established using empirical data derived from 72 different cases (Fauchald 2008, 304). The study concluded that there is room for improved alignment with other international tribunals in their approaches to interpretation techniques.

Lastly, several studies examined statutory legal interpretation in the US. In Choi’s ‘An empirical study of statutory interpretation in tax law’ (2020), the author used empirical data to identify patterns and trends in tax law interpretation, with a focus on textualist vs. purposive arguments. In another study, entitled ‘An Empirical Examination of Agency Statutory Interpretation’, Amy Semet (2019) demonstrated empirically that there are identifiable trends in agency statutory interpretation. Finally, in ‘The Phantom Philosophy – An Empirical Investigation of Legal Interpretation’ by Czarnezki and Ford (2006), the authors examined the underlying philosophies guiding the legal interpretation process through empirical investigation. They found that sharp interpretive disagreements among academics are not mirrored in the practice of judging.

3 Rules on ECHR Interpretation

Building on the canonical distinction by Friedrich Carl von Savigny, Robert Alexy (1995, 84–89) distinguishes between four categories of legal arguments: linguistic, genetic, systemic, and substantial arguments (including teleological arguments). This typology was developed in the context of domestic statutory interpretation. The ECHR as the object of interpretation, however, is an international human rights treaty. For international treaty interpretation, Jacobs (1969, 318–320) named three main approaches: textual interpretation according to the language of the treaty, subjective interpretation according to the intentions of the state parties, and teleological interpretation according to the object and purpose of the provisions. This is reflected in the VCLT which enshrines a comprehensive set of rules for treaty interpretation in Articles 31–33 VCLT (Orakhelashvili 2008, 309).¹⁴ Although it should be pointed out that the VCLT rules are not strict instructions for the process of interpretation, but only designate ‘elements to be taken into account’ and their relative weights (Dörr 2012, 521, 522; Manley et al. 2023, 778). In addition, most rules contained in the VCLT are considered to be customary international law (see Dörr and Schmalenbach 2012).

There are different techniques of legal interpretation that the ECtHR can use to justify its decisions.¹⁵ Even before the VCLT entered into force in 1980, the ECtHR established in Golder v. United Kingdom (para 29) that Convention rights interpretation is guided by the principles laid down in the VCLT.¹⁶ Aspects to be considered include the multilingual nature of international treaties and the impact of the sovereignty principle on treaty-interpretation. To ensure the uniform application of the Convention, the ECtHR’s interpretation is also autonomous from domestic interpretations and concepts: terms, whether descriptive or prescriptive, cannot be imported directly from a domestic legal order (Cremer 2022, para 33). In considering the human rights content of the object of interpretation (the Convention), the ECtHR has noticeably widened the interpretative toolbox in its practice and added ‘living instrument’-interpretation, ‘effet utile’ (useful effect)-interpretation, and consensus-based interpretation, to name a few (Dzehtsiarou 2011).¹⁷ These techniques of interpretation are distinctive approaches used by the ECtHR and answer to a practical need encountered when adjudicating international human rights (Dothan 2014).¹⁸ In particular, these distinct interpretive techniques only partially overlap with the canonical techniques of legal interpretation. Our empirical study, however, focuses on the classical techniques of legal interpretation that are both reflected in the VCLT and in the case-law of the ECtHR. It is therefore not detrimental that the ’court-specific‘ or ‘human rights specific‘ techniques of legal interpretation are not mentioned in the VCLT.

Turning to the interpretation techniques in more detail, textual interpretation (i.e. the interpretation of the wording of a provision) is the starting point of any Convention interpretation (Cremer 2022, para 31). In this study, we considered the ECtHR to apply textual interpretation when it analyses the literal meaning of a word, a sentence, or a paragraph to interpret a specific article of the Convention. An example for textual interpretation is the following: ‘[A]s a matter of textual interpretation, the wording of Article 5 § 4, especially its French version, clearly suggests that the court must have the power to order a release if it finds the detention unlawful.’ (Buishvili v. Czech Republic, para 39). Textual interpretation requires that the ‘ordinary meaning’ of a provision is determined by the context of the entire treaty (Article 31 para. 1 VCLT) (Cremer 2022, para 32), which shows the close connection between textual interpretation and systematic interpretation. We define systematic interpretation as the consideration of the context of a Convention provision (Cremer 2022, para 34–36). Systematic interpretation means that a provision is interpreted by taking into account the normative environment (i.e., content or structure) of the article containing the provision or of the ECHR as a whole. This is an example of the use of the systematic interpretation technique by the ECtHR: ‘The two sentences of Article 2 of Protocol No. 1 must be interpreted not only in the light of each other but also, in particular, of Articles 8, 9 and 10 of the Convention.’ (Folgerø and others v. Norway, para 84).

Going beyond the text of the Convention, teleological interpretation is the third interpretation technique mentioned in Article 31 para. 1 VCLT. The Court uses teleological interpretation when it considers the object and purpose of a provision to interpret a specific article of the Convention (Cremer 2022, para 51). An example for teleological interpretation is the following passage: ‘By rendering the administration of justice transparent, publicity contributes to the achievement of the aim of Article 6 § 1, namely a fair trial, the guarantee of which is one of the principles of any democratic society.’ (Dağli v. Turkey, para 28). We consider the Court’s ‘living instrument’ doctrine an instance or variant of teleological interpretation (Cremer 2022, para 51).¹⁹ It should be noted that teleological interpretation does not necessarily lead to a more extensive or even activist application of the Convention, as teleological interpretation can also be restrictive.²⁰

Historical interpretation is often viewed as a rather static method of interpretation (Djeffal 2015). While historical interpretation is the fourth traditional interpretation technique in the domestic sphere, the preparatory works (travaux préparatoires) and the circumstances of the conclusion of the treaty are only considered supplementary means of interpretation in Article 32 VCLT. We regard historical interpretation as present in a case when the Court refers to the origin and preparatory works of a provision to interpret it. This excerpt provides an example: ‘It is noteworthy that the drafting history of Protocol No. 6 reveals a common understanding between the bodies and institutions of the Council of Europe that Article 10, paragraph 1 of the Convention, in its wording as originally drafted, could reasonably be considered as already comprising the ‘freedom to seek information’.’ (Magyar Helsinki Bizottság v. Hungary, para 136).

As stated above, in addition to the four canonical techniques of legal interpretation, the Court also uses additional interpretation techniques, both those based on the VCLT and those developed autonomously in the Court’s jurisprudence. Therefore, in this study, we also considered interpretations based on subsequent agreement (Article 31 § 3 (a) VCLT) and on subsequent practice (Article 31 § 3 (b) VCLT), interpretation based on rules and practice of international law applicable in the relations of the parties (Article 31 § 3 (c) VCLT), autonomous interpretation (when the Court explicitly states that it defines provisions independently of equivalent domestic concepts), interpretation based on comparative legal analysis (when the Court surveys the law and practice concerning specific issues in the member states), interpretation based on the emergence of a European consensus, and interpretation based on dialogue between courts (when the ECtHR refers to the case-law of other [human rights] courts, e.g. to that of the Inter-American Court of Human Rights). It is important to note that there are no clear and accepted definitions regarding the types of interpretation techniques. This made it challenging to classify and categorize interpretative arguments in legal judgments.

It is essential to underscore that no hierarchy exists among the interpretation techniques,²¹ nor is there a prescribed sequence in which the ECtHR must apply them. In practice, the ECtHR uses its discretion to resort to any of these techniques either independently or in combination.

4 Data and Method

Our dataset includes all ECtHR judgments that were decided between 2006 and 2011 (N = 8,436). We identified these judgments in the Court’s HUDOC portal²² and downloaded the English and French judgment texts. For this project, we used 9 variables annotated by law student assistants using a custom interface. Annotation was semi-automated: where possible, we used text mining methods to automatically extract variables from the text, which were then checked and, if necessary, edited by the human annotators. Capturing the interpretation techniques was particularly pertinent for this analysis. In every case, we captured the presence or absence of 11 interpretation techniques.²³ Where an interpretation was present in a sentence, we also recorded the Convention articles that were interpreted by the ECtHR. The application that was custom programmed for this task, searched for text parts that potentially contained an interpretation using a set of pre-defined keywords and phrases. For example, a key phrase for the use of textual interpretation was ‘the wording of Article …’. All text parts that contained key phrases were then shown to the human annotators who checked whether the text part actually contained the use of an interpretation technique and does not, for example, simply reproduce the wording of a Convention article or deal with an article of domestic law. Given our narrow operationalization of legal interpretation, we are able to capture instances in which legal interpretation is used to advance the case-law (rather than, for example, mere self-citations by the Court of past interpretative acts).²⁴ The annotator would also enter the article that was interpreted (for example, ‘3’ if the text part was ‘the wording of Article 3 …’). The holdings were recorded similarly, so that all Convention articles examined in a judgment and the respective outcomes were also extracted. With this approach, we documented for each judgment whether any specific Convention article was examined in the case, whether the article was interpreted or not (and, if so, which interpretation techniques were used), and whether the Court found a violation of the article or not.

To ensure robustness, annotation was carried out on the basis of an extensive codebook, the human annotators were in regular exchange, and ambiguous formulations in judgments were discussed in regular meetings. The application also included equivalent formulations for English and French to generate annotations that were consistent across both languages. To assess the inter-coder reliability, we drew a random sample of 773 cases. These were independently coded by two coders.²⁵ This amounts to 9.2% of all cases in the scope of this study. We then calculated Cohen’s kappa (κ) to evaluate the agreement between coders. The most common interpretation techniques (textual interpretation, teleological interpretation, and systematic interpretation) have kappa values around 0.7, which can be considered ‘substantial’ (Landis and Koch 1977) or ‘good’ (Fleiss 1971) agreement. The other, less common, interpretation techniques show greater variation, ranging from ‘excellent’ agreement e.g. for European consensus interpretation (κ = 0.94) and autonomous interpretation (κ = 0.84), to ‘fair’ agreement, e.g. for dialogue of courts interpretation (κ = 0.4). In summary, our data (especially for the most common interpretation techniques) is robust with good inter-coder agreement. At the same time, this also shows that identifying legal interpretations is a difficult task, even for specialized human annotators.

5 Results and Discussion

5.1 Frequency of Interpretation

Contrary to our initial assumption—given the importance of interpretation for legitimacy—the ECtHR infrequently employs interpretation techniques. Altogether, we find that at least one interpretation technique is used in 2,859 out of the 8,436 cases (34% of cases). This result is in line with those of recent empirical literature.²⁶ In addition, it can partially be explained by our conservative approach to capturing legal interpretation.²⁷ However, when the Court resorts to interpretation, it frequently uses multiple interpretation techniques together: 39% of those cases containing interpretations use two or more interpretation techniques, with up to eight different techniques used in rare cases.

Figure 1. Frequency of Interpretation Techniques Across All 8436 Cases (the additional data used for this figure did not make part of our analyses).

Figure 1 shows how often the different interpretation techniques are used by the ECtHR. Textual interpretation is the most common interpretation technique employed by the Court—a finding that is also in line with the legal literature: it occurs in 18% of all cases.²⁸ Teleological interpretation follows second, being used in 15% of cases. Somewhat surprisingly, systematic interpretation occurs more frequently than we initially assumed based on expert legal knowledge: systematic interpretation was used in 1,104 cases out of 8,436 cases, or 13% of cases. The other interpretive methods are only rarely used for Convention interpretation by the ECtHR: each occurs in 2% of cases or less. This includes historical interpretation (or ‘reference to preparatory works’) as the fourth interpretation technique of the traditional canon, and the Court-specific approaches of ‘autonomous interpretation’ and ‘European consensus’.

Furthermore, as emerges from the literature (Cremer 2022, para 54), the preparatory works of the treaty and the circumstances of its conclusion only play a marginal role in ECHR interpretation. This can be explained by the following reasons: first, it is important to note that historical interpretation and the interpretation based on the circumstance of the treaty’s conclusion are only supplementary means of interpretation, Article 32 VCLT. These supplementary means of interpretation are considered secondary to the primary means outlined in Article 31 of the VCLT, which include the text, context, and purpose of the treaty provisions (Cremer 2022, para 56). Already by virtue of the VCLT, the primary means are given greater weight in the interpretive process because they are seen as more directly reflective of the states’ consent to be bound by the treaty provisions (Dörr 2018, para 2). Second, the preparatory works and circumstances of the Convention’s conclusion can sometimes be difficult to ascertain with precision. As noted on the Council of Europe's website, the preparatory work refers to the ‘various documents that were produced during the drafting of the Convention and its first protocol’, including reports of discussions in the Consultative Assembly, the Committee of Ministers and certain expert committees.²⁹ Analyzing and deriving conclusive interpretations from these sources can be a challenging task.

While the ECtHR has at its disposal a range of Court-specific interpretation techniques—such as ‘evolutive interpretation’, ‘effet utile’, and consensus-based interpretation—it employs them less frequently than we expected. This limited use might be a manifestation of the Court’s cautious approach, keenly aware of the potential criticisms of overstepping or reshaping the Convention beyond member states’ original intent. The nature of the cases presented might not always demand these nuanced techniques, or the Court might be subtly employing them without overtly labelling their use. Furthermore, by emphasizing classical interpretation techniques, the ECtHR may be reinforcing their foundational importance to national courts. This strategic restraint ensures the Court’s flexibility and seeks to maintain its legitimacy amidst the diverse legal cultures and expectations of the member states.

In summary, our findings suggest that the Court rather infrequently engages the canonical techniques of legal interpretation: it occurred only in one third of cases examined here. This result can partially be explained by our conservative way of operationalizing legal interpretation.³⁰

Figure 2. Frequency of Interpretation Technique per Article. 95% confidence intervals for all proportions are drawn around each value in grey.

For the subsequent analysis, we focus on the three most common interpretation techniques: textual, teleological, and systematic interpretation. This is due to the fact that we recorded very few instances where the ECtHR resorted to other interpretation techniques (e.g. European consensus, preparatory works). Figure 2 shows how frequently the different interpretation techniques were used for each article: For example, Article 2 ECHR was interpreted systematically in 55% of cases where it was mentioned, with the grey intervals showing that the 95% confidence interval ranges from 50% to 60%.³¹

Systematic interpretation is particularly important in the context of Articles 2 and 3: Article 2 ECHR (right to life) was interpreted systematically in more than half of cases where it was mentioned. The right to life encompasses a broad range of issues from state-led executions, use of force by law enforcement, medical decisions at the end of life, to issues of public safety.³² This intrinsic complexity suggests a systematic interpretation that takes into account a broad set of legal norms and principles to provide a comprehensive understanding of that Convention right. Article 2 cannot be fully understood or applied in isolation, as it often comes into play with other rights guaranteed by the ECHR, such as the right to an effective remedy (Article 13), the right to liberty and security (Article 5), or the prohibition of torture (Article 3). For instance, in cases involving the use of lethal force by state authorities, both the right to life and the prohibition of inhuman or degrading treatment might be relevant.³³ By considering the wider normative environment, the Court can address the nuances and complexities of the right to life with more depth and clarity.

Textual interpretation appears to be of primary importance for Article 11 ECHR (freedom of assembly and association), and to a slightly lesser degree, for Article 8 (right to private and family life), Article 14 (non-discrimination), and Article 1 of Protocol No. 1 (protection of property). These findings are not easy to explain. Especially Article 8 and Article 14 are characterized by very open and vague legal concepts (e.g. ‘private life’). The ordinary meaning of these concepts can hardly be considered as uncontroversial which makes the frequent use of textual interpretation intriguing. However, by using textual interpretation, the Court may want to signal its close adherence to the text and its ambition to prevent a potential broadening or narrowing of these rights beyond the scope intended by the framers of the ECHR (De Sloovère 1934, 552).

This signal seems especially crucial when grappling with the vague and contentious concepts embedded in for example Articles 8 and 14, minimizing potential accusations of judicial activism or overreach. This static approach to Articles 8 and 14 is counterbalanced in other cases by the equally visible resort to teleological interpretation. Given their textual openness (e.g. ‘non-discrimination’) and vagueness (e.g. ‘private life’), the issues that come under these articles can be complex, involving delicate balances between an individual interest and the interests of the society or the state. It seems that in these cases a textual interpretation is not called for; instead, teleological interpretation allows the ECtHR to look beyond the text and interpret the article in light of its overarching purpose: to protect individuals’ private lives from unjustified interference (in relation to Article 8 ECHR) and to protect the individual from an unjust disadvantage (in relation to Article 14 ECHR). This flexible interpretation can help to ensure that the ECtHR’s judgments respond effectively to changing societal contexts and understandings of privacy. Article 14 of the ECHR, on the other hand, prohibits ‘discrimination on any ground such as sex, race, color, language, religion, political or other opinions, national or social origin, association with a national minority, property, birth, or other status.’ This article also involves a wide range of complex issues that often require a nuanced understanding and interpretation. Applying a teleological interpretation allows the ECtHR to take into account the broader societal implications of discrimination and the purpose of Article 14, which is to ensure fair treatment for all.

However, on the whole, our data does not confirm the sometimes-voiced concern that the ECtHR tends to place central emphasis on the technique of teleological interpretation.³⁴ Systematic or textual interpretation are more frequently used in all Convention articles analysed here except for Article 5 (right to liberty and security). The frequent use of teleological interpretation for Article 5 could be due to the fact that it often involves balancing competing interests and values. Applying a teleological interpretation allows the ECtHR to consider the wider social, political, and legal context, which is crucial when dealing with matters as complex and sensitive as individual liberty and security. It facilitates the Court’s pursuit of the overarching objectives of the ECHR, including respect for human dignity and human freedom, and it allows the ECtHR more interpretive leeway to adapt these objectives to the sensitive law enforcement context.

Finally, Article 6 (fair trial) in particular seems to require little interpretation: none of the four traditional interpretation techniques under consideration here are used to interpret Article 6 in more than 15% of the instances where this provision is examined. We speculate that this intriguing finding may be due to the following reasons: first, the legal rules contained in Article 6 can be considered relatively clear and explicit compared to other articles. They do not incorporate open and vague text to a similar degree such as, e.g., the right to non-discrimination (Art. 14 ECHR).³⁵ The right to a ‘fair and public hearing within a reasonable time by an independent and impartial tribunal established by law’ or the sub-paragraphs that establish a detailed set of rights in criminal proceedings, such as the presumption of innocence, reflect well-established principles in domestic and international law. These principles are typically understood within clearly defined boundaries and parameters, which reduces the necessity for extensive interpretation. Second, the violations of Article 6 often involve very specific and procedural aspects such as undue delays in legal proceedings (Mayer 2022, para 82). These are concrete issues where the interpretation of the right is not necessarily the central question, but rather the assessment of the specific facts and circumstances in relation to the established understanding of the right. Thirdly, while the interpretation techniques might not be explicitly used, it does not necessarily mean that the Court does not interpret the provision. The ECtHR may, for example, apply a common law-inspired approach to Article 6 ECHR and interpret it by deriving principles from the established case-law rather than relying on the techniques of legal interpretation.³⁶ This is underlined by the fact that Article 6 ECHR is the most frequently examined Convention article in our dataset: it appears in 4,514 cases, which is more than half of all 8,436 cases. As such, the case-law regarding Article 6 may be particularly well-developed and robust, and therefore require less interpretation than other articles.³⁷ It is therefore plausible to suggest that any initial ambiguities in its interpretation have been resolved over time reducing the need for legal interpretation based on the canonical techniques.

5.2 Regression Models

In this part of the analysis, we look at winning arguments: Is there a relationship between the use of interpretation techniques and the determination of a violation? Does employing a particular interpretative technique increase the likelihood of identifying a violation? We estimate a series of logistic regression models to answer these questions. To ensure large enough sample sizes and to improve the interpretability of the model, we do not distinguish between Convention articles here, but instead differentiate between the main protected legal interests: life and limb (Articles 2, 3 and 4), personal life (Articles 8, 9 and Article 1 of Protocol no. 1), and justice-related rights (Articles 5, 6, 7 and 13). To control for differences between these groups of Convention articles, we fit separate models for each of the three protected legal interest.³⁸ We fit restricted models that only includes interpretations (Models 1.1, 2.1, 3.1) and full models that also control for the covariates applicant type, respondent state, priority, and Grand Chamber cases (Models 1.2, 2.2, 3.2). Specifically, in each of these models the probability that articles from the protected legal interest were violated is estimated by:

\[\ln\left( \frac{\pi(x)}{1 - \pi(x)} \right) = \beta_{0\ } + \sum_{i = 1}^{I}{\beta_{i\ }{Interpretation}_{i\ }} + \sum_{k = 1}^{K}{\beta_{k\ }{Covariate}_{k\ }}\ \ \ \ \ \]

where \(\pi(x) = P\left( Y = 1\ \middle| x_{1}\ldots x_{p} \right)\) is the probability to observe a violation knowing all the p independent variables. \({Interpretation}_{i\ }\)is the i-th interpretation (I = 4), and \({Covariate}_{k\ }\)is the k-th additional covariate (K = 4). The covariates were only present in the full models.

We disaggregated our case-level data (N = 8,436), so that every time an article (belonging to a protected legal interest) is examined in a case, it became an observation. In every observation, the article (belonging to a protected legal interest) may be interpreted or not, and it may be violated or not. The number of observations varies between the protected legal interests because some articles are examined much more often than others: while there are 7,515 cases where articles from justice-related rights are examined, there are only 1,911 observations for the personal life-interest, and 1,356 observations for the life and limb-interest.³⁹

5.2.1 Outcome Variable: Finding of a Violation

The outcome variable records whether the Court found at least one violation of an article in the concerned protected legal interest. For example, in a case where the Court held that there was a violation of Article 6 ECHR, the ‘justice’-related rights would be coded as violated.

‘Winning’ is the expected outcome for the few applicants that make it past the admissibility stage at the ECtHR and reach the merits stage: Table 1 shows that violations were found in 91% of the occurrences of the protected legal interest of ‘personal life’ (Articles 8, 9, and article 1 of Protocol no. 1), 93% of the occurrences of ‘justice’-related rights (Articles 5, 6, 7, and 13), and 86% of occurrences of the ‘life and limb’-related rights (Articles 2, 3, and 4).

The finding of a violation is therefore the baseline outcome for cases that reach the merits stage. This also shows that the preliminary screening mechanisms of the Court are effective at ensuring that the Court’s resources are focused on cases where there is a substantial likelihood of a violation being found.⁴⁰ With respect to the higher proportion of non-violations of ‘life and limb’-related rights, we suspect that in these severe cases, the filter is more open at the beginning (i.e. the Court is more hesitant to discard these cases easily and more cases that ultimately are non-violations are adjudicated). In other words, our findings imply that even when there is some doubt about the merits of a case, the ECtHR may choose to hear it due to the gravity of the alleged violations. It is also possible that the ECtHR is conscious of the important ramifications for individuals and states alike if such serious allegations go unheard. An alternative explanation is that the judges may be more hesitant to find violations and shame the state in cases alleging the most serious human rights violations (leading to additional findings of non-violation in edge cases).⁴¹

5.2.2 Primary Independent Variables: Interpretation

The main independent variables capture whether at least one article (belonging to a protected legal interest) was interpreted. For example, if the Court employed a textual interpretation for Article 13, the associated legal interest, ‘justice’, would be recorded as interpreted in this manner. The labelling of interpretations was performed by our annotators at the article level, as described above. Textual, teleological, and systematic interpretation were recorded separately, while the remaining interpretation techniques were aggregated as ‘other interpretations’ due to their low numbers.⁴²

5.2.3 Covariates

We included additional independent variables that may predict case outcomes as covariates to address possible confounding. The first of these variables is whether the victim is a natural person or a legal person. This was determined by annotators based on the applicant information in the judgment head. The second covariate is the respondent state, aggregated on three levels: old member states (accession before 1990), new member states (accession since 1990), and Turkey.⁴³ The respondent state was extracted from the judgment title. As a third covariate, we include a Boolean indicator whether the decision was handed down by the Grand Chamber (Yes) or by a Chamber (No), which is indicated in the judgment head.

5.2.4 Results

The regression results in Table 2 show that, perhaps counterintuitively, arguments constructed on the basis of the canonical interpretation techniques are, from the perspective of the applicants, not ‘winning’ arguments: while there are differences between the protected legal interests and interpretation techniques, the use of arguments thus generated generally does not make the finding of a violation more likely. All four analyzed interpretation techniques are associated with a lower probability of finding a violation for at least one of the three protected legal interests. This may be explained as follows: as stated, at the merits stage, a ruling in favor of the applicant tends to be the default outcome. Consequently, if the Court wishes to diverge from this typical resolution, it necessitates harnessing its full argumentative power, including arguments from legal interpretation. Ultimately, this can be viewed as demanded by considerations of legitimacy.

While the primary analytical methods of this study are empirical, the findings yield noteworthy implications about the Court's interpretative choices. Both textual interpretation and the combined category of ‘other interpretation’—which chiefly encompasses autonomous interpretation, comparative interpretation, and interpretation based on the rules and practice of international law—demonstrate a notable negative relationship with the likelihood of identifying violations pertaining to ‘personal life’ and rights associated with justice. In the realm of textual interpretation, it is plausible to postulate that an emphasis on the ‘ordinary meaning’ might restrict the Court’s interpretative breadth. By adhering closely to the text, there is a potential risk of overlooking wider socio-cultural nuances or pressing policy considerations. Consequently, this could make judges less inclined to recognize violations, given the narrower lens of their analysis. Turning our attention to the ‘other interpretation’ category, what stands out is our result regarding the use of transnational interpretative techniques (i.e. those that consider international law and differentiate from domestic understandings of a concept). The data suggests that when judges employ interpretation methods that draw upon international law or diverge from traditional domestic understandings of a concept, there is a diminished tendency to determine violations. This trend calls for further inquiry, especially in understanding how global perspectives in legal reasoning might influence the Court's determinations.

The use of the teleological interpretation technique has a significant negative relationship with the probability of finding a violation of ‘personal life’. Interestingly, when the Court focusses on the purpose behind this protected legal interest, it might establish a higher threshold for determining what constitutes a violation. Only infringements that seriously compromise the core objective behind the protected interest of ‘personal life’ might be recognized as violations.

The influence of systematic interpretation on case outcomes varies depending on the category of the protected legal interest involved. For justice-related rights, systematic interpretations are associated with a lower probability of finding a violation. On the contrary, in cases concerning ‘life and limb’ systematic interpretation shows a significantly positive relationship, making it the lone interpretation technique to exhibit a positive association with the finding of a violation in any category of protected legal interests. In the context of justice-related rights, one possible explanation could be that these rights are often more comprehensively articulated in the ECHR. Therefore, when systematic interpretation is applied—taking into account the broader legal context and interrelatedness of various provisions—it might lead to a narrower understanding of these rights. This could make the Court less likely to find violations as it may interpret the scope of these rights more stringently.

In simpler terms, the richness of legal detail available for justice-related rights could set a higher threshold for identifying violations when viewed through the lens of systematic interpretation. This approach might constrain the Court's latitude, resulting in fewer findings of violation within this category of rights.

The results for the protected legal interest ‘life and limb’ (Articles 2, 3, and 4) are shown in Models 1.1 and 1.2. Model 1.1 only includes the interpretation techniques, whereas Model 1.2 adds the covariates to control for the applicant type, the respondent state group, and whether the case was decided by the Grand Chamber.⁴⁴ The presence of a systematic interpretation significantly increases the chance of finding a violation in both models (OR 2.10 for Model 1.1 and OR 1.82 for the full Model 1.2). In the reduced Model 1.1, the other interpretation techniques appear to have a negative relationship with the finding of a violation, although less important after adding the covariates in the full model (OR 0.28 and OR 0.36 for Models 1.1 and 1.2). It should be noted that in Model 1.2, victim type is not included as a control variable for this protected legal interest, as only natural persons can be victims of violations of the right to life, the prohibition of torture or the prohibition of slavery.

Models 2.1 and 2.2 show the results for the group of justice-related rights (Articles 5, 6, 7, and 13). In both the reduced model (Model 2.1) and the full model (Model 2.2), the presence of textual interpretation and ‘other’ interpretations are associated with a lower probability of finding of a violation (OR 0.49 and 0.51 respectively). Surprisingly, in the full model, systematic interpretation also has a negative relationship with the finding of a violation, which was not statistically significant in the limited model without the control variables (OR 0.69).

Finally, Models 3.1 and 3.2 show the results for the protected legal interest of ‘personal life’ (Articles 8, 9, and article 1 of Protocol no. 1). Here, textual interpretation, teleological interpretation and the ‘other’ interpretation techniques all show significantly negative associations in both models. Their odds ratios are, respectively, 0.66, 0.57 and 0.50 in Model 3.2, indicating also a surprisingly negative association with observing a violation.

All models share a rather limited explanatory power, as expressed in the pseudo R² values. This is not uncommon for logistic regressions on legal data.⁴⁵ In these models, the objective is to understand the determinants of a violation ruling without considering the specific facts of each case. This approach inherently results in a considerable amount of unexplained variance. Nonetheless, the full models consistently explain more of the variance, indicating that the country of origin is also correlated with finding a violation. The separation plots (see Annex Figure 3) illustrate the obtained R² values (Greenhill et al. 2011). Despite the low pseudo R² values, the χ²-tests indicate that the models are significantly better than a model that only includes a constant.

6 Conclusion

While legal interpretation is often likened to an art rather than a science, this perspective should not dissuade us from conducting systematic analyses to discern the underlying patterns and trends in legal argument construction by interpretation (Frank 1947, 1259; Jennings 1967, 544; Sur 1974). The empirical analysis performed in this study is one of the first comprehensive explorations of the use of legal interpretation techniques in the judgments of the ECtHR. It reveals that the use of interpretation techniques plays a role in only a third of the cases under examination here, with multiple interpretation techniques frequently used in conjunction. Among these, textual interpretation appears as the most commonly used technique, followed by teleological and systematic interpretations. Other techniques, such as historical interpretation, autonomous interpretation, and the court-specific approach of the European consensus, are used less frequently. Different articles of the ECHR demonstrate varying tendencies in terms of the preferred interpretation method. For instance, the right to life (Article 2) is often interpreted systematically, while the fair trial provision (Article 6) is less frequently subjected to the four canonical interpretation techniques. These findings suggest that the choice of interpretation technique is influenced by the specific legal question or right being examined in a case.

From the perspective of the member states, it is interesting to observe the ECtHR’s interpretative choices and frequency. Given that the ECHR is an international treaty, the principle of legal certainty in international law and the contention that ‘states cannot be expected to enforce an international obligation they did not initially consent to’ can be construed as setting certain limits to dynamic interpretation (Magyar Helsinki Bizottság v. Hungary, para 150). The Court’s visible preference for textual interpretation, which is the most common method used, echoes the primacy of the text in treaty interpretation and underlines the importance of adhering to the clear wording of the Convention. This adherence might be perceived positively by member states as it signals a level of respect for the original intent and boundaries of the Convention. Moreover, the limited use of Court-specific techniques perhaps reflects the ECtHR's caution against an overly dynamic interpretation. This restrained approach can be interpreted as an acknowledgment of the diverse legal cultures and histories of member states, aiming to ensure a balance between the progressive interpretation of rights and the need to respect state sovereignty and the original intent of treaty framers. The ECtHR's interpretative practices appear to strive for a balance between fidelity to the Convention’s text and the need for its progressive interpretation.

The study also investigated whether the use of certain interpretation techniques could be linked to the case outcome. The logistic regression models used here suggest that the presence of legal interpretation does not significantly increase the probability of an applicant winning the case. Our explanation is the following: The baseline outcome for cases that reach the judges at the ECtHR on the merits stage is the finding of a violation. The conclusion that the use of interpretation techniques does not make the finding of a violation more likely is consistent with this model: Each of the four interpretation techniques examined significantly decreases the probability of identifying a violation for at least one of the three protected legal interests. This observation can be understood in the context that, during the merits phase, a judgment favoring the applicant is usually expected. Thus, for the Court to depart from this expected outcome, it must employ its full argumentative resources, encompassing reasons rooted in legal interpretation. This approach can be seen as a measure to uphold its legitimacy. From the perspective of the applicant and legal practitioners, given that the use of certain interpretation techniques may even decrease the probability of a violation finding, it is important to note that the use of arguments derived from legal interpretation is not necessarily a winning strategy. Further research should analyze the impact of arguments derived from legal precedents on case outcomes. It should also be borne in mind that the association between interpretation techniques and case outcomes can vary considerably depending on the type of legal interest being examined.

Our analysis has certain limitations due to its focus on the presence of legal interpretation techniques without considering the specific factual context of each case. Neither do we consider the submissions of the parties. Nevertheless, the results underline the importance of legal interpretation techniques in the work of the ECtHR and offer a novel perspective on the relationship between interpretation techniques and case outcomes. Another limitation is time, as we only include data from six cross-sections (2006–2011). Certain time trends would only be visible if more data were available for the analysis. Finally, we were unable to measure the importance or practical weight of an argument derived from applying an individual interpretation technique: we treat all interpretation techniques equally, but in practice, a legal argument derived from, e.g., an act of textual interpretation, may be more strongly associated with the case outcome than an act of systematic interpretation.

Further research, integrating more information on the factual context and the influence of different legal cultures at the Court, could shed more light on the complex interplay of interpretation and litigation outcomes at the ECtHR. In addition, interviews with judges, law clerks, and parties could be helpful in corroborating our results. A comprehensive temporal analysis could be conducted to unearth shifts in interpretation methods over the years. This would offer insights into evolving judicial behavior, legal cultures, and case characteristics. Comparative analysis between respondent countries may reveal how different legal traditions or national contexts shape the Court’s interpretative methodologies. Investigating the link between interpretation methods and case complexity may also yield interesting results. Complexity could be determined by factors such as the number of legal issues involved, the number of parties in the case, or the length of the judgment. Examining the alignment between the interpretations advocated for by parties and those adopted by the Court could provide valuable insights into litigant strategy and its effectiveness. Finally, a different research design needs to be adopted to capture the techniques of legal argumentation used by the ECtHR in the majority of cases in which it does not resort to the classical techniques of interpretation.

In light of the empirical evidence presented, this study underscores that while the interpretative techniques are central to the Court’s decisions, understanding their nuanced application requires further exploration into legal cultures, litigant strategies, and the dynamics of judicial reasoning at the ECtHR. In conclusion, the study highlights the complex nature of legal interpretation at the ECtHR. While interpretation techniques play an important role in the Court’s adjudication, their use is not a guaranteed path to a favorable outcome for applicants.

7 Acknowledgments

The authors are grateful for excellent research assistance by Martin Cattaneo, Frédéric Fitzi, Zoé Gianocca, Andrea Vuksic, and Laura Whittlestone. We thank the participants of the 1st Conference European Society for Empirical Legal Studies, Warsaw School of Economics, September 26, 2023, for helpful comments.

8 Funding

Funding by the Swiss National Science Foundation (SNSF) is gratefully acknowledged (Grant no. 170498). The data that support the findings of this study are available upon request from the corresponding author.

9 References

Alexander, Larry, and Emily Sherwin. 2008. "Demystifying Legal Reasoning." Cambridge University Press.

Alexy, Robert. 1995. "Recht, Vernunft, Diskurs – Studien zur Rechtsphilosophie." Suhrkamp.

Alleweldt, Ralf. 2022. "Kapitel 10 - Recht auf Leben." In EMRK/GG Konkordanz-Kommentar (3rd edition), edited by Oliver Dörr, Rainer Grote and Thilo Marauhn. Mohr Siebeck.

Altwicker-Hámori, Szilvia, Tilmann Altwicker and Anne Peters. 2016. "Measuring Violations of Human Rights: An Empirical Analysis of Awards in Respect of Non-Pecuniary Damage Under the European Convention on Human Rights." Heidelberg Journal of International Law 76 : 1–51. http://dx.doi.org/10.2139/ssrn.2631404.

Ammann, Odile. 2020. "Domestic Courts and the Interpretation of International Law: Methods and Reasoning Based on the Swiss Example." Brill, Nijhoff.

Artico v. Italy, app. no. 6694/74, ECHR, 13 May 1980, Series A no. 37.

Bogdandy, Armin von, and Ingo Venzke. 2014. "In Whose Name? A Public Law Theory of International Adjudication (International Courts and Tribunals Series)." Oxford University Press.

Braman, Eileen. 2009. "Law, Politics, and Perception: How Policy Preferences Influence Legal Reasoning." University of Virginia Press.

Brunozzi, Kathrin. 2023. "Art. 45 Begründung der Urteile und Entscheidungen." In Europäische Menschenrechtskonvention, edited by Jens Meyer-Ladewig, Stefan von Raumer, Martin Nettesheim. Helbing Lichtenhahn Verlag.

Buchanan, Allen, and Keohane Robert O. 2006. "The Legitimacy of Global Governance Institutions." Ethics and International Affairs 20 (4) : 405–437. https://doi.org/10.1111/j.1747-7093.2006.00043.x.

Buishvili v. Czech Republic, app. no. 30241/11 ECHR, 25 October 2012.

Choi, Jonathan H. 2020. "An Empirical Study of Statutory Interpretation in Tax Law." New York University Law Review 95 (2) : 363–441. http://dx.doi.org/10.2139/ssrn.3460962.

Costa, Jean-Paul. 2011. “On the Legitimacy of the European Court of Human Rights’ Judgments.” European Constitutional Law Review 7 (2) : 173–182. https://doi.org/10.1017/S1574019611200026.

Cremer, Hans-Joachim. 2022. "Kapitel 4 - Regeln der Konventionsinterpretation." In EMRK/GG Konkordanz-Kommentar, edited by Oliver Dörr, Rainer Grote and Thilo Marauhn. Mohr Siebeck.

Czarnezki, Jason J., and William K. Ford. 2006. "The Phantom Philosophy? An Empirical Investigation of Legal Interpretation." Maryland Law Review 65 : 841-906. http://dx.doi.org/10.2139/ssrn.773865.

Dağli v. Turkey, app. no. 28888/02, ECHR, 27 November 2007.

De Sloovère, Frederick J. 1934. "Textual Interpretation of Statutes." New York University Law Quarterly Review 11 : 538.

Demir and Baykara v. Turkey, app. no. 34503/97, ECHR 2008–V.

Djeffal, Christian. 2015. "Static and Evolutive Treaty Interpretation: A Functional Reconstruction." Oxford University Press.

Dothan, Shai. 2018. “The Three Traditional Approaches to Treaty Interpretation: A Current Application to the European Court of Human Rights.” Fordham International Law Journal 42 (3) : 765–794. http://dx.doi.org/10.2139/ssrn.3241331.

Dothan, Shai. 2014. "In Defence of Expansive Interpretation in the European Court of Human Rights." Cambridge Journal of International and Comparative Law 3 (2) : 508–531. https://doi.org/10.7574/cjicl.03.02.149.

Dörr, Oliver, and Kirsten Schmalenbach. 2012. "Vienna Convention on the Law of Treaties: A Commentary." (1st edition). Springer-Verlag.

Dörr, Oliver. 2012. "Article 31: General Rule of Interpretation." In Vienna Convention on the Law of Treaties: A Commentary (1st edition), edited by Oliver Dörr and Kirsten Schmalenbach, 521–570. Springer-Verlag.

Dörr, Oliver. 2018. "Article 32 – Supplementary means of interpretation." In Vienna Convention on the Law of Treaties: A Commentary (2nd edition), edited by Oliver Dörr and Kirsten Schmalenbach, 617–633. Springer-Verlag.

Dworkin, Ronald. 1986. "Law’s Empire." Harvard University Press.

Dzehtsiarou, Kanstantsin. 2011. “European Consensus and the Evolutive Interpretation of the European Convention on Human Rights.” German Law Journal 12 (10) : 1730–1745. https://doi.org/10.1017/S2071832200017533.

Fauchald, Ole Kristian. 2008. "The Legal Reasoning of ICSID Tribunals – An Empirical Analysis." European Journal of International Law 19 (2) : 301–364. https://doi.org/10.1093/ejil/chn011.

Fayed v. United Kingdom, app. no. 17101/90, ECHR 21 September 1994, Series A no. 294-B.

Feteris, Eveline, and Harm Kloosterhuis. 2009. "The Analysis and Evaluation of Leal Argumentation: Approaches from Legal Theory and Argumentation Theory." Studies in Logic, Grammar and Rhetoric 16 (29) : 307–331.

Fleiss, Joseph. 1971. "Measuring Nominal Scale Agreement Among Many Raters." Psychological Bulletin 76 (5) : 378–382. https://doi.org/10.1037/h0031619.

Folgerø and others v. Norway, app. no. 15472/02, ECHR 2007–III.

Frank, Jerome. 1947. "Words and Music: Some Remarks on Statutory Interpretation". Columbia Law Review 47 (8) : 1259-1279. https://doi.org/10.2307/1118098.

Gadamer, Hans-Georg. 1993. "Vom Zirkel des Verstehens." In Gesammelte Werke Band 2: Hermeneutik II Wahrheit und Methode: Ergänzungen, Register, edited by Hans-Georg Gadamer. Mohr Siebeck.

Glöckner, Andreas and Christoph Engel. 2013. "Can We Trust Intuitive Jurors? Standards of Proof and the Probative Value of Evidence in Coherence-Based Reasoning." Journal of Empirical Legal Studies 10 (2) : 230–252. https://doi.org/10.1111/jels.12009.

Golder v. United Kingdom, app. no. 4451/70, Commission decision of 21 February 1975, Series A no. 18.

Goss, Ryan. 2023. "The Disappearing ‘Minimum Rights’ of Article 6 ECHR: the Unfortunate Legacy of Ibrahim and Beuze." Human Rights Law Review 23 (4) : 1–23. https://doi.org/10.1093/hrlr/ngad024.

Grabenwarter, Christoph and Katharina Pabel. 2021. "Europäische Menschenrechtskonvention." Helbing Lichtenhahn Verlag.

Greenawalt, Kent. 2018. "Realms of Legal Interpretation: Core Elements and Critical Variations." Oxford University Press. https://doi.org/10.1093/oso/9780190882860.001.0001.

Greenhill, Brian, Michael D. Ward and Audrey Sacks. 2011. "The separation plot: A new visual method for evaluating the fit of binary models" American Journal of Political Science 55 (4) : 991–1002. https://doi.org/10.1111/j.1540-5907.2011.00525.x.

Habernal, Ivan, Daniel Faber, Nicola Recchia, Sebastian Bretthauer, Iryna Gurevych, Indra Spiecker gen. Döhmann, and Christoph Burchard. 2023. "Mining Legal Arguments in Court Decisions." forthcoming in Artificial Intelligence and Law. https://doi.org/10.48550/arXiv.2208.06178.

Hart, Herbert Lionel Adolphus. 1997. "The Concept of Law." Oxford University Press.

Herdegen, Matthias. 2020. "Interpretation in International Law." In Max Planck Encyclopedia of Public International Law, edited by Anne Peters and Rüdiger Wolfrum.

Huang, Kuo-Chang. 2009. "Does Discovery Promote Settlement? An Empirical Answer." Journal of Empirical Legal Studies 6 (2) : 241–278. https://doi.org/10.1111/j.1740-1461.2009.01143.x.

Jacobs, Francis G. 1969. "Varieties of Approach to Treaty Interpretation: With Special Reference to the Draft Convention on the Law of Treaties Before the Vienna Diplomatic Conference." International Law and Comparative Law Quarterly 18 (2) : 318–320. https://doi.org/10.1093/iclqaj/18.2.318.

Jennings, Robert Yewdall. 1967. "General Course on Principles of International Law." In Collected Courses of the Hague Academy of International Law Volume 121 : 323–619.

King, Nancy J., and Michael Heise. 2018. "Appeals by the Prosecution." Journal of Empirical Legal Studies 15 (3) : 482–538. https://doi.org/10.1111/jels.12190.

Landis, J. Richard, and Gary G. Koch. 1977. “The Measurement of Observer Agreement for Categorical Data.” Biometrics 33 (1) : 159–174. https://doi.org/10.2307/2529310.

MacCormick, Neil. 1998. "Legal reasoning and interpretation." Routledge Encyclopedia of Philosophy. DOI 10.4324/9780415249126-T010-1.

MacCormick, Neil. 1978. "Legal Reasoning and Legal Theory." Clarendon Press.

Madsen, Mikael Rask, Pola Cebulak, and Micha Wiebusch. 2018. "Backlash against International Courts: Explaining the Forms and Patterns of Resistance to International Courts." International Journal of Law in Context 14 (2) : 197–220. https://doi.org/10.1017/S1744552318000034.

Magyar Helsinki Bizottság v. Hungary, app. no. 18030/11, ECHR, 8 November 2016.

Mamatkulov and Askarov v. Turkey, app nos. 46827/99 and 46951/99, ECHR 2005-I.

Manley, Stewart, Pardis Moslemzadeh Tehrani, and Rajah Rasiah. 2023. “Mapping Interpretation by the International Criminal Court.” Leiden Journal of International Law 36 (3) : 771–806. https://doi.org/10.1017/S0922156523000225.

Mayer, Franz C. 2022. "Art. 6 - Recht auf ein faires Verfahren." In EMRK: Konvention zum Schutze der Menschenrechte und Grundfreiheiten – Kommentar, edited by Ulrich Karpenstein and Franz C. Mayer. Helbing Lichtenhahn Verlag.

Orakhelashvili, Alexander. 2008. "The Interpretation of Acts and Rules in Public International Law." Oxford University Press.

Peat, Daniel. 2021. "The Tyranny of Choice and the Interpretation of Standards: Why the European Court of Human Rights uses Consensus.” New York University of International Law and Politics 53 (2) : 381–432.

Popa, Liliana E. 2018. "Patterns of Treaty Interpretation as Anti-Fragmentation Tools, A Comparative Analysis with a Special Focus on the ECtHR, WTO and ICJ." Springer, Cham. https://doi.org/10.1007/978-3-319-65488-1_5.

Raz, Joseph. 1996. "Ethics in the Public Domain." Clarendon Press.

Semet, Amy. 2019. "An Empirical Examination of Agency Statutory Interpretation." Minnesota Law Review 103 (5) : 2255–2341.

Slapin, Jonathan B. 2022. "SNIS Final Report: A Quantitative Textual Approach of the European Consensus Method of Interpretation in the European Court of Human Rights." available at: https://snis.ch/wp-content/uploads/2020/09/2020-Slapin_Complete-Final-Report_21-09-22.pdf (last visited July 18, 2024).

Stappert, Nora. 2020. "Practice Theory and Change in International Law: Theorizing the Development of Legal Meaning through the Interpretive Practices of International Criminal Courts." International Theory 12 (1) : 33–58. https://doi.org/10.1017/S1752971919000150.

Sur, Serge. 1974. "L’interprétation en Droit International Public." Paris : Librairie générale de droit et de jurisprudence.

Sweet, Alec Stone, and Helen Keller. 2008. "A Europe of Rights: The Impact of the ECHR on National Legal Systems." Oxford University Press.

Tiersma, Peter M. 1999. "Legal Language." University of Chicago Press.

Tyler, Tom R. 2006. "Why do People Obey the Law." Princeton University Press.

Tyrer v. the United Kingdom, app. no. 5856/72, ECHR, 25 April 1978, Series A no. 26.

Velkhiyev and others v. Russia, app. no. 34085/06, ECHR, 8 March 2012.

Zarpli, Omer and Huseyin Zengin. 2022. "Shame, endorse, or remain silent?: State response to human rights violations in other countries." Research & Politics 9 (1). https://doi.org/10.1177/20531680211070344.