Strengthening the protection of the right to remain silent at the investigative stage: What role for the EU legislator?

This article reflects on the possible contribution of the European Union towards safeguarding the right to silence at the investigative stage of criminal proceedings in EU Member States. The analysis is not limited to the Directive 2016/343/EU and other procedural rights’ Directives. Rather, it focuses on the role of the EU as a legal and political player, pursuing the goal of enhanced protection of procedural rights in criminal proceedings. The article first examines compliance of the legal provisions of the four examined jurisdictions with the Directive. It then identifies the relevant areas, not addressed or insufficiently addressed in the existing EU instruments, which appear problematic as far as the effectuation of the right to silence is concerned. The article argues that a more detailed binding EU regulation is not an appropriate solution to address the existing problems. Instead, it suggests that the EU legislator should consider other, more indirect means of action.


Introduction
The opening article to this Special Issue argued that the provisions of Directive 2016/343/EU ('the Directive') relevant to the right to silence are unlikely to lead to significant changes in the laws, regulations and judicial practices of Member States ('MS'). As noted in that introduction, 1 of the three represented jurisdictions -Belgium, Italy and the Netherlandsthat were under obligation to transpose the Directive, 2 none, at time of writing, have done so via any perceptible change in their national law. 3 A number of other EU MS consider their laws to be compliant with the Directive in respect of the right to silence. 4 In the first part of this contribution, we analyse the laws and regulations of the four MS covered in this Special Issue 5 in light of the Directive provisions. Although Ireland has opted out of the Directive, it is included in this and subsequent analyses alongside the three remaining jurisdictions to illustrate issues that may potentially exist in this area in domestic systems. Ireland is also a useful example of a jurisdiction that despite having opted-out of the Directive, nevertheless largely (and coincidentally) complies with its provisions. Such examples certainly engender questions as to the usefulness of the relevant provisions of the Directive, and the extent to which it has (or has not) facilitated greater protection of the right to silence.
We conclude that it is impossible to point at manifest non-compliance of the laws and regulations of the examined MS with the respective directive provisions, especially given the general nature of the latter, and the possibilities for exceptions mentioned in the recitals. This does not mean, however, that these jurisdictions afford optimal protection of the right to silence at the investigative stage, nor that no problematic issues arise in law or in practice in these jurisdictions. In the second part, we consider whether adoption of a legislative instrument with more detailed provisions, such as a new or an amended directive, presents a suitable and practicable solution for the issues that we have identified. We argue that a more detailed binding EU regulation is not an appropriate solution. Instead, we suggest that the EU legislator 6 should consider other, more 'indirect' means of action.
The comparative findings Compliance with the directive As discussed in the opening article, 7 most of the Directive provisions pertaining to the right to silence are not particularly clear, requiring further interpretation. As ever, the dilemma is whether the Court of Justice of the European Union ('the CJEU') will follow the interpretation given by the European Court of Human Rights ('the ECtHR'), or whether it will provide an autonomous interpretation, possibly opting for an expansionist reading of the respective rights (as compared to the ECtHR standard). 8 So far the CJEU has preferred the first approach. 9 The existingalthough still very limitedcase law interpreting the Directive also suggests that the CJEU would follow the ECtHR interpretations in the absence of further guidance in the Directive. 10 It suggests, further, that the CJEU would be mindful of the principle of 'minimum degree of harmonisation' implied in art 1 and recital 10 of the Directive, stating its purpose is to lay down common minimum rules with the view to facilitating mutual recognition and mutual trust. 11 It would therefore strive to ensure that its interpretations remain confined to the literal scope of the respective provisions, and afford a margin of appreciation to the MS. On that basis, we consider that, in the absence of respective CJEU pronouncements, the reading of the Directive provisions relevant to the right to silence would closely follow their textual meaning (interpreted in conjunction with recitals) and the interpretations given by ECtHR. Furthermore, due to the 'minimum degree of harmonisation' principle, doubts concerning whether the Directive implies a more exacting or far-fetching standard are likely to be resolved in the negative.

Article 2: Scope of application
Article 2 of the Directive states that it applies to suspects in criminal proceedings. None of the four countries, besides Italy, reported issues of potential restriction of the right to silence due to the classification of certain procedures as non-criminal by domestic law. 12 In respect of Italy, the CJEU has held, in a recent preliminary reference that 'natural persons who are subject to an administrative 7. Pivaty  . 12. This does not mean that such issues do not exist in these jurisdictions, but rather that they were not reported (and if they exist, they were probably not viewed as central in the given systems as far as the operation of the right to silence is concerned).
investigation for insider dealing have the right to silence when their answers might establish their liability for an offence that is punishable by administrative sanctions of a criminal nature, or their criminal liability'. 13 With the exception of Italy, 14 the use of information obtained under compulsion in administrative proceedings in subsequent criminal proceedings was likewise not discussed in the country-specific articles.
Article 7(1): General provisions on the right to silence Article 7(1) read in conjunction with recital 26 states that MS should ensure suspects' right to remain silent, and that this right and the right not to incriminate oneself should apply to (all) questioning relating to a suspected crime. The right has a firm footing in legislation or case law in the four jurisdictions. For illustration, in Belgium, the right to silence and the privilege against selfincrimination can be found, at least in part, in art 47bis of the Belgian Code of Criminal Procedure (Belgian CCP). Article 47bis states that no onesuspect or non-suspectcan be obliged to make a statement against his or herself, be obliged to plead guilty, or be compelled or forced to (physically) cooperate with the authorities. The latter, in particular, protects suspects from being forced to cooperate with other investigative acts 'when these have a self-incriminating character'. 15 Note that, not dissimilarly to Italy, while the privilege against self-incrimination applies to all, the explicit right to silence is only granted to suspects. 16 In Belgium, the Netherlands and Italy, the right to silence is viewed as a fundamental expression of the nemo tenetur principle (privilege against self-incrimination), according to which no one can be compelled to take responsibility for a criminal action. While all three jurisdictions do not extend the right to the giving of personal identity details 17 , in the Netherlands, unlike in Belgium and Italy, nemo tenetur is translated as the right to silence in of itself. Accordingly, while it was initially declared nearly absolute, the case law has since allowed inroads into this right. Thus it is not limited to incriminating evidence, but it also includes all information of a factual nature. 18 In Ireland, somewhat unusually for an EU MS, the right to silence is protected as a corollary of freedom of expression. 19 Accordingly, it may be restricted in order to serve the exigencies of public order and morality. The Irish legislature has created several statutory provisions on inferences from pre-trial silence as discussed in the Article 7(5): Adverse Inferences From Silence section. 20 In contrast, an accused at trial cannot be compelled to testify and no inferences may be drawn from their failure to give evidence at trial in their own defence. Of the four jurisdictions considered, Italy seems to afford the broadest justification for the right to silence. Here, the right is considered a means by which the right of passive self-defence may be exercised; accordingly, it is linked to both the suspect's moral freedom and self-determination. 21 In all jurisdictions, the right to silence applies to formal questioning of suspects. Its application to investigative actions which may include elements of (informal) questioning is somewhat of a grey area in Italy, Belgium and the Netherlands. In Belgium, for example, and similarly to the Netherlands, although the right to silence applies to police interrogations, the concept of interrogation is not expressly defined in national law. Belgian case law defines the interview rather narrowly, excluding, for instance, exchanges between police officers and a future suspect in the context of a house search. 22 Voluntary statements, likewise, do not fall under the concept of interrogation. 23 Interestingly, Belgium also allows the (voluntary) use of polygraphs, either to give 'direction' to the investigation, or if requested by the deference to demonstrate a defendant's innocence. In the Netherlands, the status of statements given by suspects during undercover operations is uncertain, although recent case law indicates that in some circumstances such statements may fall under the ambit of the right to silence. 24 It is, however, unclear whether and to what extent the uncertainty surrounding the statements given outside of a formal interrogation indicates non-compliance with the Directive. As noted above, all examined jurisdictions provide for the legal right to silence during police questioning. Whether or not statements from (future) suspects taken outside of formal questioning, without informing them of their rights, is an infringement of the right to silence depends on the particular case circumstances (namely, whether the situation resembled that of a formal questioning). It therefore seems that any question of possible non-compliance with the right to silence in such circumstances must be resolved not at the level of national legislation, but by the national courts.
Article 6: Burden of proof Article 6 of the Directive focuses on the burden of proof stating that 'MS shall ensure that the burden of proof for establishing the guilt of suspects and accused persons is on the prosecution'. All examined jurisdictions contain legislative provisions or case law stating that a defendant has no burden to prove or disprove the alleged facts against them; in other words, the onus of proving the defendant's guilt is on the prosecution.
The issue reported in the Dutch and Italian articles, namely that according to established case law, invocation of the right to silence may bar the possibility of obtaining compensation for unjustified pre-trial detention 'because the former suspect had made no attempt to prove their innocence…' 25 would fall outside the ambit of art 6. That is because the CJEU has clarified that art 6 applies only to 21. Marchesi and Panzavolta (n 5) ? Concerning moral freedom, the Italian Code of Criminal Procedure (Italian CCP) forbids using methods or techniques that could influence 'the freedom of self-determination of the person' and alter the ability to remember and assess facts. The prohibition covers both those that directly aim to obtain an answer (e.g. torture, narcoanalysis and hypnosis) and those that aim to assess whether the statements of the declarant are true (e.g. polygraphs the judicial decision of guilt (i.e. final verdicts), and not to other procedural acts (the decision on the compensation for unjustified pre-trial detention would presumably fall under this category). 26

Article 7(3): Material other than oral statements
Article 7(3) states that 'the exercise of the right not to incriminate oneself shall not prevent the competent authorities from gathering evidence which may be lawfully obtained through the use of legal powers of compulsion and which has an existence independent of the will of the suspects or accused persons'. Thus, the Directive simply repeats the ECtHR language in Saunders 27 which, as argued in the opening article, 28 remains confusing. The respective recitals of the Directive likewise refer back to the ECtHR case law.
In the Netherlands, the Supreme Court has not interpreted the right to silence in such a way as to allow the suspect to refuse to cooperate with a blood test or a breath analyser test. The Court held that the right cannot be invoked in the context of material evidence that is independent on the will of the suspect (meaning, existing externally of the suspect's mind). The complexities of interpreting both Saunders and art 7(3) are apparent with the distinction made by Dutch case law between unlocking a (smart)phone by using the passcode and doing so by using a biometric technique such as the fingerprint scanner or facial recognition software. While the former is protected by the right to silence, the latter is exempted through the court's interpretation that such biometrics exist independent of the will of the suspect. 29 Accordingly, the suspect may be compelled to provide such data information to the investigative authorities.
In Italy, suspects do not have a right to withhold evidence in their possession, except for information covered by state secret and professional privilege. 30 The suspect cannot be compelled to produce documents or other real evidence, but the authorities always have the right to acquire it, even coercively if needed. 31 In Belgium, art 88quarter of the Belgian CCP allows the investigative judge to compel anyone with 'special knowledge' of computer systems to provide information about the computer system under investigation, including on how to gain information to the system, or the data therein. 32 While the courts had previously prevented such compulsion from being requested from a suspect, in order to protect their privilege against self-incrimination, this is no longer so: recent jurisprudence of both the Court of Cassation and the Constitutional Court has suggested that so-called 'decryption orders', issued pursuant to art 88quater, do not infringe the privilege against self-incrimination where such orders clearly relate to evidence or material that 'exists independently of the will of the suspect'. 33 In Ireland, although the law is unclear on compelling the production of documents, in certain circumstances (those concerning financial, theft or fraud-based investigations) a person may be required to do produce documents, with a failure to do so amounting to a criminal offence. Irish courts have held that forensic samples may be obtained from a suspect prior to their accessing legal 26 advice, which is viewed by some as confirming that the right not to incriminate oneself does not extend to items of real evidence. 34 The legal interpretations adopted in the four jurisdictions under analysis rather reflect a narrow reading of the respective ECtHR case law, according to which material other than oral statements does not fall within the ambit of the privilege against self-incrimination. As explained in the opening article, 35 it is unclear to what extent the latter applies, under the ECtHR regime, to information or material obtained from the suspect, for the ECtHR has extended the privilege to other types of evidence such as documents. 36 However, on its face, and as also explained in the opening article, this reading may be considered compliant with the Directive (as one possible interpretation, given the latter's repetition of the language in Saunders).
Article 7(4): Taking into account cooperative behaviour of the suspect Article 7(4) of the Directive states that it is up to MS to 'allow their judicial authorities to take into account, when sentencing, cooperative behaviour of suspects and accused persons'. It is of note that this article does not explicitly mention the possibility of taking into account uncooperative behaviour for sentencing purposes.
In the jurisdictions under analysis, we encountered examples of both types of reasoning. In the Netherlands, according to the Supreme Court, the suspect's failure to assist the investigation, including refusal to provide relevant information, may be taken into account when determining the appropriate sentence. 37 In Italy, likewise, according to some case law, the exercise of the right to silence can be considered as a negative element when assessing the behaviour of the defendant(s) for sentencing purposes. In contrast, in Belgium, judges are normally forbidden from taking into account the way in which the defendant has organized his defence when sentencing, with the Court of Cassation clear that the right to silence shall have no role in the sentencing decision. 38 Article 216/1 of the Belgian CCP, however, does permit the public prosecutor, in certain circumstances, to promise a reduction, mitigation or the like to any person who makes, inter alia, 'substantial statements' concerning either their own participation or that of third parties. 39 Such a promise may lead to an incentive for a suspect to speak (thus perhaps providing an impetus to break their silence); such a scenario, however, would seem to be outside the remit of art 7(4).
In Ireland, cooperation with the gardaí during a period of detention in custody may be taken into account in mitigation at the sentencing stage. 40 Early guilty pleas are, perhaps unsurprisingly, particularly encouraged through a reduction in sentencing. In a similar vein, the Italian legal system has established a number of incentives, rewards and benefits to encourage the person to offer their contribution of knowledge. An example is the offering of benefits to suspects who confessjudges tend to value confession as a mitigating factor. Rewarding cooperation and punishing non-cooperation seems to encourage waiver of the right to silence in very a similar fashion and they may be difficult to distinguish from one another in practice. It is thus unclear whether the Directive also extends to taking into account uncooperative behaviour. In any event, whilst mitigation schemes seem to have a statutory nature, the practices taking into account non-cooperative behaviour emerge as part of judicial reasoning in sentencing decisions in concrete cases. Whilst national legislatures might attempt to correct judicial practices, they cannot always pre-empt the discretionary decisions taken by a judge in practice.

Article 7(5): Adverse inferences from silence
The Directive prohibits the use of adverse inferences from silence as 'evidence (of guilt)', but it is unclear if it would tolerate, according to recital 28, their use as an element of 'corroboration of evidence'. Suspects' silence, or the failure to provide a satisfactory (innocent) explanation is indeed being used for evidentiary purposes in all four jurisdictions, although in a different fashion. Ireland is the only jurisdiction where the possibility to draw adverse inferences is provided for by statute.
Here, ss 18-19 of the Criminal Justice Act 1984 allow for 'the drawing of inferences from the accused's failure or refusal in the pre-trial period of investigation to account for the presence of any object, substance or mark on their person, clothing or footwear, or in their possession, or in the place where they were arrested, or on their presence at a particular place'. Section 19A provides that inferences may be drawn at trial from a suspect's failure in the pre-trial period to mention any fact, when being questioned, charged or informed that they might be charged with a particular offence, that they later rely on in their defence at trial, being a fact which in the circumstances existing at the time 'clearly called for an explanation'. A number of safeguards, aimed at reflecting both constitutional and wider ECtHR jurisprudence on the right to silence and privilege against selfincrimination are provided for suspects under those statutory inference-drawing provisions. 42 Notably, the Supreme Court of Ireland has held that, first, restrictions on the right to silence as occasioned by such provisions are proportionate to the aim pursued; 43 secondly, the inferences must be evidential in nature only, and they may not be the sole basis for the conviction of the accused; and thirdly, the court is obliged to act in accordance with the principles of constitutional justice, including having regard to an accused's right to a fair trial. 44 On initial analysis, Ireland's legislation on inferences may be viewed -as compliant with art 7(5), which, as argued in the opening article, 45 can be read to reflect the ECtHR case law guarantees. 46 In Italy, Belgium and the Netherlands, the evidentiary consequences of silence or the failure to provide an account emerge from the case law. In the Netherlands, the court may take a suspect's silence into account when the suspect does not give a reasonable explanation with regard to the circumstances which, taken separately or in conjunction with other evidence, are incriminating for the suspect. The circumstances which call for a reasonable explanation from the suspect are a prima 46. Note, however, that the ECtHR has linked its acceptance of inference provisions with the right of (timely) access to legal advice. Daly (n 5). While detained suspects in Ireland are entitled to access legal advice, and have been allowed to have solicitors attend at interview, there is as yet no statutory entitlement to the latter aspect of the right to legal assistance in Ireland. See supra, the Right to Legal Assistance section.
facie case, where considerable evidence is already present against the suspect and where the suspect's silence can thus be used for corroboration of other evidence. 47 In Italy, judges have distinguished between two different uses of silence. While the use of a suspect's silence during pre-trial questioning 48 as evidence to prove their guilt is clearly prohibited as it would violate the onus probandi rules, 49 the use of silence for other evidential purposesto different extentsseems to be allowed by the case law. The Court of Cassation, for example, has considered the suspect's silence at the pre-trial stage as an 'argomento di prova', that is to say, a means to interpret other evidence. 50 In such cases, some safeguards are also put in place. Moreover, Italian courts have consistently underlined that the rules concerning the burden of proof cannot be violated. 51 Nonetheless, it is clear that some adverse inferences may be tolerated, particularly where a suspect refuses to provide exculpatory explanations or remains silent throughout the proceedings on potentially inculpatory circumstances. 52 And finally, in Belgium, although the law does not explicitly prohibit the drawing of adverse inferences from a suspect's silence, such a prohibition can be found in a consistent strand of case law. 53 This states, particularly, that the judge can neither rely on the silence of the suspecteither directly or indirectlynor draw any conclusions as to their guilt or use their silence to corroborate other evidence. 54 Nonetheless, the courts may 'attach probative value' to the silence of the accused, insofar as such silence does not challenge the incriminating evidence. 55 Accordingly, while the judge cannot base their decision on the fact a suspect remained silent, they can use such silence to draw arguments against the defendant, or to 'further confirm' a fact which remains undisputed. 56 The compliance of the four jurisdictions with art 7(5) of the directive would in part depend on how art 7(5) will be interpreted by CJEU: as prohibiting any kinds of adverse inferences from silence and thus invalidating recital 28 (which in our view is unlikely), or as tolerating some kinds of inferences. In any case, however, any non-compliance is also likely to surface on the level of individual cases, where an argument may be raised concerning the de facto weight of silence attached by the judges in the given circumstances.

Article 10(2): Remedies
Article 10(2) of the Directive states that 'without prejudice to national rules and systems on the admissibility of evidence, MS shall ensure that, in the assessment of statements made by suspects or accused persons or of evidence obtained in breach of the right to remain silent or the right not to incriminate oneself, the rights of the defence and the fairness of the proceedings are respected'.
As noted in the opening article, this formulation is particularly vague, a point demonstrated by the equally ambiguous remedies available in each jurisdiction.
A trend was detected in all jurisdictions towards a more restrictive application of the exclusionary rule, also echoed in the ECtHR case law (particularly, with regard to pre-trial restrictions of procedural rights). 57 Nonetheless, the examined jurisdictions differ in their approaches to the exclusion of involuntary statements. For illustration, in Ireland, involuntary statements, namely those obtained by threats, inducement or 'oppression', are excluded from evidence. In the Netherlands, statements obtained under unacceptable pressure during (police) interrogation are not per se excluded, but other remedies are used such as reduction of sentence (which arguably constitutes a weak deterrent mechanism against involuntary confessions). In contrast, the failure to inform a suspect about the right to silence would usually lead to exclusion of evidence. 58 Whether or not a remedy would be afforded with regard to other possible violations of the right to silence depends on the circumstances of the given case. In the Netherlands, in order for either exclusion of evidence or a stay of proceedings to be upheld, any procedural violation must be found to amount to a violation of 'a considerable extent' (neither has been much used to remedy breaches of the right to silence.) 59 Similarly, in Belgium, although the courts have clarified the criteria 60 based on which a violation of the right to a fair trial should result in the exclusion of evidence, it is difficult to predict how courts will apply these criteria to the right to silence, as this depends on the factual circumstances of the case.
Thus, also with regard to the application of art 10(2) it appears that, even if national laws are considered in conformity with the Directive, the practices of national courts when interpreting and applying the respective provisions to the individual cases would significantly affect MS compliance.

Other issues identified in the country studies
The right to silence and cautions The duty to provide information about procedural rights, including the right to silence, forms part not of the a caution on the right to silence to be given, the scope, content and the manner of delivery of the caution differ. In the Netherlands, for instance, the content of the right to silence caution is not defined in law or regulations, 62 in contrast to, for example, Belgium and Italy where the duty to inform a suspect about the different elements of the right to silence is envisaged (and thus safeguarded) by the law. 63 In all jurisdictions, the obligation to inform about the right to remain silent applies only to suspects (in the Netherlands, from the moment when the criminal suspicion is sufficiently formalized), and not, for example, to witnesses (although generally there is a requirement to stop an interrogation to provide a caution if a witness turns into a suspect in the course of an interrogation). There are differences with regard to the moment at which a caution must be given: in the Netherlands, the obligation to caution applies to suspect interrogations, but it does not need to be repeated at every interrogation. In Ireland, according to the Judges' Rules, a caution must be provided at interrogation and at the time of the arrest, whilst in Belgium the requirement to give a (written) caution at the moment of arrest is missing. All jurisdictions require the provision of written information on rights, including the right to silence.
Italy, Belgium and the Netherlands do not require suspects to be informed about the consequences of remaining silent, but in Ireland there is a specific requirement to inform suspects about the potential consequences of a failure or refusal to provide information in response to specific questions with regard to interviews where provisions on adverse inferences from silence may be invoked (see the Article 7(5): Adverse Inferences From Silence section). The debate regarding whether a caution should be more detailed, so that a suspect is made aware of the potential negative consequences of remaining silent, is illustrated well by the example of Italy where the scholarship argues, on the one hand, that a detailed caution ensures that suspects are fully aware of the implications of exercising the given right, and on the other, that giving awareness of all the negative consequences may be akin to compelling (or at least, encouraging) the person to speak. 64 Although the examined jurisdictions seem compliant with the respective Directive 2012/13/EU provisions, the manner in which a caution is administered has significant impact on the exercise of the right to silence in practice.

Disclosure of case-related information
Early disclosure of case-related information affects the effective exercise of the right to silence because suspects (and their lawyers) can only make an informed decision about whether to make use of this right when they are given sufficient prior disclosure. 65 This is particularly so given that the strength of the evidence against the suspect has proven to be the most important factor influencing the choice of procedural strategy, including whether or not to remain silent. 66 Whilst the right to disclosure of case-related information is provided by art 7(4) of Directive 2012/13/EU, this provision is generally not interpreted as granting the right to the suspect and their 62. Article 29(2) of the Dutch CCP states that 'before the interrogation the suspect should be informed that he is not obliged to give an answer'. In Ireland, the obligation to provide the (right to silence) caution does not have a statutory nature, but is envisaged in the Judges' Rules. lawyer to be informed about the evidence in advance of the interrogation (although there is a duty to provide information about the criminal charge). 67 Similarly, the extent of the authorities' obligation to disclose case-related information at the investigative stage in the four examined jurisdictions is unclear. In the Netherlands, for instance, the suspect is allowed to request access to inspect the case file from the public prosecutor, from his first interrogation after arrest (subject to the prerogative to withdraw access to certain materials in the interests of investigation). 68 In practice, however, access is usually given with some delay and certainly not before the first interrogation.
In Italy, there is no obligation to provide access to the (full) case file in advance of suspect interrogations conducted by the prosecutor. The only exception is when the suspect requests that they be questioned. Either way, before the beginning of any questioning the authorities are obliged to inform the suspect of the facts alleged against them and to disclose the incriminating elements collected against them (and the sources, if this is not detrimental to the investigation). The issue of disclosure in the Garda (police) station is also notable in Ireland. While Ireland has opted into the Directive on the right to information, it is unclear what real impact this has on procedures in garda custody. It has not been transposed into Irish law but is directly effective. Nonetheless, there are no statutory guidelines in place in relation to disclosure in the Garda station. The Garda Code states that 'an Garda Síochána is not obliged to disclose any information that could prejudice an investigation'. 69

Right to legal assistance
The effective exercise of the right to silence, particularly at the investigative stage, is also closely connected to the right to legal assistance. A lawyer is best placed to inform the suspect about the consequences of the exercise of the right to silence, give advice on whether or not to use the right, and support the suspect in exercising it, including during police interrogations. 70 From the four examined jurisdictions, all have statutory provisions on the right to legal assistance at the investigative stage. Ireland, which has not opted into Directive 2013/48/EU, has recognized the right to legal advice as constitutional in nature and has provided for it, in limited form, in legislation. Despite the absence of a recognized legislative or constitutional right specifically to have one's lawyer present at garda interview, Ireland has introduced a system providing for just that. 71 In the absence of a statutory footing for the right of access to a lawyer at the investigative stage, however, the provision of legal assistance at police interviews is haphazard (there is, for instance, no Garda station should be to protect and advance the legal rights of their client'. 73 This does not make any reference to the effective participation of the lawyer embedded in art 3(3) Directive 2013/48/ EU, though is arguably broad enough to allow for the same, depending on interpretation.
In contrast, Belgium, Italy and the Netherlands (being party to Directive, 2013/48/EU) contain detailed legislative provisions on the right to legal assistance at the investigative stage, including during suspect interviews, and on the appointment of (legal aid) lawyers to provide such assistance. However, in both Belgium and the Netherlands, restrictions are imposed on the lawyer's interventions during police interviews, which arguably create tensions with art 3(3) of Directive 2013/ 48/EU. 74 In the Netherlands, for instance, the lawyer may not address questions and comments to the interrogating officer except for at the end of the interview; they may not answer questions on behalf of the suspect; they may not request multiple time-outs for additional private consultation; and they may not intervene for other reasons than to signal that undue pressure is being exercised or that the suspect does not understand a question or that they are unfit for questioning. 75

Rules pertaining to the conduct of interrogations
Although the rules on the conduct of suspect interrogations often do not directly refer to the right to silence, the vagueness of these rules, as illustrated by the country analyses, may seriously undermine the effectiveness of this right.
In this regard, it is remarkable that in the Netherlands and Belgium, no statutory definition of 'interrogation' exists, and no clear guidance in given on how the interrogation should be conducted. Consequently, there is little guidance on what methods of interrogation are permitted (with the exception of methods that have been reviewed by courts) and what constitutes unlawful pressure or coercion. None of the examined jurisdictions, with the exception of Ireland, seems to impose upper limits on the duration of interrogations. 76 Whilst in Ireland and the Netherlands police are generally permitted to continue posing questions to suspects after they indicate their wish to use the right to silence, in Belgium the legal position on this matter is ambiguous (the Court of Cassation seemingly endorsing the view that questioning may continue). 77 The rules and practices of audiovisual recording of interrogations differ in the examined jurisdictions. For instance, whilst in Ireland recording of interviews is prescribed 78 and interviews in relation to serious offences are almost invariably audio-visually recorded, in Belgium there is no formal obligation to audio-visually record interrogations (although this remains a possibility).

Intermediary conclusions
We observe, first, that the rationales of the right to silence are not always made explicit in the countries under examination, and that different jurisdictions have translated various aspects of the rationales differently into their laws and judicial practice. Secondly, we note that although some of the issues with regard to the protection of the right to silence overlap between jurisdictions, each jurisdiction also has its own unique areas of concern. Thirdly, we observe that the 'problems' related to the legal protection of the right to silence as described in the country reports 79 are very diverse.
Considering the comparative findings described above, the four examined jurisdictions are mostly compliant with the Directive's provisions related to the right to silence. However, as noted, other problems exist which are not mentioned in the Directive and which do not evidently fall within its scope, or within the scope of the other procedural rights' directives.
Addressing the problems: the role of the EU The next question to consider is whether any outstanding issues related to the protection of the right to silence could or should be addressed via a more detailed binding EU-wide legislative instrument such as by an amended Directive 2016/343, a new directive, or by other means. Although directives are the most known and commonly used legislative means to advance EU policies, other types of instruments grouped under the label of 'soft law' are increasingly used for this purpose. 80 'Soft law' instruments, such as recommendations, guidelines or resolutions issued by the Commission or Council continue to play an important role in the Justice and Home Affairs ('JHA') area even after the third pillar has been communitarised by the Treaties of Amsterdam and Lisbon. 81 A number of soft law instruments on procedural rights in criminal proceedings exist to date, including the Commission recommendations on the right to legal aid for suspects or accused persons in criminal proceedings 82 and on procedural safeguards for vulnerable persons suspected or accused in criminal proceedings. 83 In the following section, we argue that a more detailed directive is not an appropriate means of action. We consider three types of arguments: those related to the EU legislative process, the nature of the right to silence, and practical effectiveness of EU legal action in this area.

Concerns related to EU legislative process
First, it is highly unlikely that a proposal for more detailed and stronger provisions around the right to silence would pass through the EU legislative negotiation process. 84 As demonstrated in the opening article, 85 during the original negotiations for the Directive, many details were deleted and several qualifiers were added to limit the scope of the respective rightswhich in the case of this Directive resulted in particularly vague and confusing legal provisions. The lack of consensus about the right to silence 86 was palpable during the negotiations, and this is unlikely to change in a new negotiations round. More generally, the limitations related to the legal basis for EU action on procedural rights in criminal proceedings are likely to pose an obstacle towards further enhancement of EU legislative standards on the right to silence. As observed by Ouwerkerk, art 82(2) of the Treaty on the Functioning of the European Union ('TFEU') is too confined to develop expansive 87 and comprehensive legislation on procedural safeguards in criminal proceedings. 88 Article 82(2) states that binding legislation (directives) can be adopted in the field of JHA to establish common minimum rules 'to the extent necessary to facilitate mutual recognition of judgements and judicial decisions and police and judicial cooperation in criminal matters having a cross-border dimension'. The mention in this article of 'common minimum rules' and the functionalist definition of EU legislative competence imposes obvious constraints as to the scope of legislation to be adopted thereunder. 89 The reference to 'common minimum rules', for instance, might suggest that the EU is not entitled to adopt higher standards than those established by the ECtHR (where such exist). Indeed, the Commission was careful to stress that the goal of proposed legislative action on procedural rights is to 'give life to' or 'concretise' the ECtHR standards, rather than to expand them 90 (although at times, directives seem to have the effect of enhancing ECtHR standards). 91 Furthermore, references to the functionalist competence of the EU and to the concept of 'minimum rules' provide ample ground to undermine efforts to push for (more comprehensive or detailed) EU legislation in this area. Indeed, the final adopted texts of directives are generally less detailed, and provide for more limited rights, as compared to the initial drafts presented by the Commission.
The special nature of the right to silence As compared with other procedural rights of suspects, the right to silence has a particularly theoretical, almost declaratory, nature. Like the presumption of innocence itself, the right to silence (brought under the umbrella of the presumption of innocence by the Directive) is closer in nature to a legal principle than a formal procedural right. 92 Unlike, for instance, the rights to information, interpretation or legal assistance, the right to silence cannot be expressed as a set of concrete 87. As regards the scope and the degree of protection of the respective rights. . 92. The right to silence is often described as a legal principle (e.g. as a fundamental or essential principle of a legal system) and it is closely linked and is often considered synonymous with the nemo tenetur se ipsum accusare ('no man is bound to accuse himself') principle.
procedural steps, such as when, by whom and how exactly this right should be effectuated. Only the 'caution' part of the right to silence appears amenable to such regulation; this is already included in the scope of the right to information. 93 Akin to a legal principle, the right to silence is deeply entrenched in the fabric of individual criminal justice systems. As evidenced by the country-specific articles, an assessment of whether this right is sufficiently respected requires a review of the national legal system as a whole, which goes far beyond the provisions on the right to silence as such. The operation of the right to silence in domestic systems is related to the rules of evidence and its assessment (including the distribution of roles between professional judges and the jury if present in the legal system), burden of proof, regulations on interrogations and (the use of) suspect statements, rules pertaining to disclosure and the case file, regulations on pre-trial detention, and review of detention. In other words, the practical operation of the right to silence is highly dependent on the particular procedural and institutional context. Needless to say, the national procedural contexts in which the right to silence operates, differ greatly. The EU would clearly lack the legal basis to impose one particular procedural context on all MS. It is also counter-productive, as it is not clear whether one or another procedural systemfor instance, adversarial or inquisitorial or mixed, jury trial, trial by a judgeis more suited to optimally protect procedural rights. 94 Related to the above, the meaning of the right to silence sparks a lot of controversy. Beyond the general statement that 'no one should be compelled to incriminate oneself', its scope and meaning are incredibly difficult to define. The lack of consensus on the very rationales of the right to silence in academic literature and in European law (namely, ECtHR case law) is a vivid illustration of these difficulties. 95 On a more practical level, the lack of clarity on the rationales leads to debates concerning the scope of application of the right to silence. Questions of relevance arise, to name only a few, with regard to certain information or material (e.g. written statements and digital data) or investigative actions (such as posing questions during an undercover operation), rules and conditions pertaining to suspect interrogations, or the types of negative consequences which can follow from the exercise of silence (for instance, effects on pre-trial decisions, or the use of silence for 'evidence corroboration').
As described earlier, some problematic aspects related to the protection of the right to silence are common to (some of) the four jurisdictions under analysis, while others are quite diverse. Where the general issues identified are common, their manifestation in each jurisdiction is quite specific and closely related to the procedural context. For example, all four country-specific articles consider the possibility that suspects' silence (or failure to provide an exculpatory statement) might have a role in the assessment of evidence and/or sentencing decisions as potentially problematic. However, the ways in which silence affects evidence assessments or sentencing decisions are very different. Whilst in Ireland inferences are subject to clear safeguards introduced by law, in the other three jurisdictions inferences evolved from court case law (facilitated by the intime conviction principle) 96 and are therefore not (always) accompanied with explicit safeguards. The conditions under which a court draws inferences from silence, likewise, differ significantly: whilst in Italy and Belgium silence at trial 97 can be taken into account by courts as 'corroboration' of other evidence, 98 the Netherlands seems to allow for a broader scope of inferences both from pre-trial and trial silence, 99 whereas in Ireland no inference is to be drawn from a failure to give evidence at trial, though pre-trial silence can have consequences. In this context, we should also bear in mind that our study included only four EU jurisdictions, and it is very likely that other issues would emerge if we considered other jurisdictions as well. Nonetheless, it is questionable whether all these different issues or areas may be (adequately) addressed in one EU-wide binding legal instrument.
The theoretical and contextual nature of the right to silence leads to considerable difficulties in translating it into a set of universally applicable procedural requirements. Whilst certain aspects of the right to silence, as described further in the section What Further Action Can the EU Take?, might be amenable to more detailed EU-wide regulation, 100 many other aspects related to the protection of the right to silence cannot be regulated uniformly and in sufficient detail at the EU level, if at all.

Effectiveness of (further) EU legislation
The country-specific articles make it clear that the issues around the protection of the right to silence in the four studied jurisdictions have a varied origin. Some problems exist partly as a result of inadequate legislative frameworks (for instance, the lack of legislative definition of 'interrogation' in Belgium), but most issues arise in the judicial practice and in the practice of other actors (for example, police interrogation practices in the Netherlands). The respective EU policy documents likewise suggest that problems mainly arise in the practice of criminal justice actors, who allegedly do not pay due respect to the right to silence in their day-to-day business. 101 If this assumption is correct, then the suggestion that EU-wide legislative pronouncements would help eliminate these practices is, at best, tenuous. Despite the growing interest in multilevel implementation of EU policies, research on the effects of EU legislation on the practices of legal operators is still scarce. 102 The existing studies of practical enforcement of EU legal instruments in the area of criminal justice in general and procedural rights in particular rather point at the lack or inconsistency of such enforcement. 103 In fact, if a gap between the 'law in the books' and 'the law in action' in this area does exist, adoption of binding EUlevel legislation might even have a counter-effect on the improvement of the respective practices. MS might assume they comply with the legal standards on paper, and so they do not need to undertake any further action. An instrument containing a list of formal procedural requirements might be interpreted by 'street-level bureaucrats' in the criminal justice system (police, lawyers, prosecutors or judges) 104 as requiring them to undertake a formal tick-box exercise, rather than protecting the substance of the respective procedural rights. 105 Cultural resistance to the Directive's provisions by 'street-level' actors, documented, for instance, with regard to resistance by police (and lawyers) to the active role of lawyers during interrogations endorsed by Directive 2013/48/EU 106 may further hinder their practical enforcement. 107 The lack of appropriate remedies for breaches of procedural rights is often considered a factor which causes lack of compliance in practice. Whilst providing for stronger remedies (for instance, exclusion of evidence obtained in breach of the right to silence and all other resulting evidence) might improve practical compliance, given the trend in ECtHR case law discussed in the Article 10(2): Remedies section it is questionable whether the EU can ensure, at least in the short term, that MS introduce such remedial regimes.
What further action can the EU take?
Our conclusion that a binding EU-wide instrument containing more detailed standards around the right to silence in criminal proceedings is not a suitable means by which to address the problems identified in the country-specific articles does not mean that there is no role for the EU to play in this area. Rather, we argue that EU action should focus on encouraging MS to monitor the operation of their criminal justice systems with a view to ensuring optimal protection of suspects' right to remain silent.
First, it is worth noting that the political goal of EU action on fair trial rights in criminal proceedings is broader than the legal basis of the respective legislative action to be found in art 82(2) of the TFEU. 108 Politically, EU action on fair trial rights in criminal proceedings rests on broader considerations than facilitating smooth cooperation in the field of JHA. Rather, it pursues the objectives of legitimizing EU action in this field (by remedying imbalance in the EU action on cooperation in criminal matters), 109 as well as more broadly enhancing procedural fairness and protection of individual rights across all EU Member States, which is deemed necessary to promote 104. The term 'street-level bureaucrats' covers those state officials who have direct contact with the members of the public.
Their action is characterized by a certain level of discretion in enforcing the rules, laws and policies which they are assigned to implement. 105. See a common EU-wide area of 'justice'. Enhanced procedural fairness and protection of individual rights as a policy goal should, in our view, be interpreted as striving to guarantee the rights that are 'practical and effective' both in law and in practice across all EU MS. This policy goal can be partly achieved via directives, but also via other means such as 'soft law' instruments, as well as various forms of political and practical action. A non-binding instrument such as a Commission recommendation could in our view usefully contribute to the policy objectives. 110 Whilst it is often assumed that 'soft law' instruments are less 'effective' as compared to binding laws due to their non-mandatory nature, the factors determining the effectiveness of EU legal instruments in practice are more complex. 111 A recommendation is not subject to the objections related to the (lack of) legal basis, whilst (still) allowing the EU to introduce reporting and other follow-up obligations for MS. 112 Another advantage lies in its flexible nature: it allows, for instance, the introduction of various regulatory options (such as different types of remedies, or different types of safeguards against 'undue compulsion' during interrogation); the describing of different factors relevant to the assessment of whether the right to silence has been respected; the provision of examples of good practices; as well as the designing of 'action plans' for MS related, for example, to monitoring the practice of effectuating the right to silence (among other procedural rights), or to training procedural actors. 113 Additionally, a recommendation could provide further clarification of certain aspects of the right to silence. As transpires from the analysis in the opening article, certain aspects relevant to the right to remain silent might benefit from further EU-wide interpretation, even if not legally binding. 114 These include, for example: safeguards against waiver, including against the risk of undue pressure to waive the right exercised by sentence mitigation schemes; defining 'undue compulsion' in the context of police interrogations; formulating explicit safeguards with respect to drawing adverse inferences from silence or attaching other negative consequences to silence; 115 defining requirements to (effective) remedies against breaches of the right; and clarifying the legal status of information or material obtained from the suspect other than an oral statement. 116 In addition to concrete provisions on the right to remain silent, a non-binding EU instrument could clarify the rationales behind the right and its relationship with the privilege against self-incrimination (which as explained in the introductory article 117 remains highly controversial), which might ultimately be used as inspiration by the ECtHR, the CJEU and by domestic courts.
Furthermore, a non-binding EU instrument could address the interdependency aspects, not covered in the Directive or other procedural rights' directives, of the right to silence with other procedural rights. The country articles demonstrate that the scope and operation of the right to silence at the investigative stage is closely connected with the (scope and operation of) other procedural rights, most importantly, of the right to information (the right to silence 'caution' and 110. See access to case-related information) and of the right to legal assistance. Whilst these rights are addressed in other procedural rights' directives and in the ECtHR case law, they have been interpreted differently by MS, and these interpretations might have important consequences for the operation of the right to silence in these states. One example is early disclosure of case-related information to suspects and their lawyers: Directive 2012/13 does not clarify the earliest moment when suspects and their lawyers should receive such information, and law and practices concerning disclosure at the pre-trial stages of criminal proceedings vary considerably among MS. 118 However, it may be argued that disclosure of (certain) case-related information even before the first interrogation of the suspect is desirable to effectively protect the suspect's right to silence. 119 Yet, the main focus of the respective EU action should in our view be directed at engaging in a dialogue with MS concerning the optimal ways of effectuating the right to silence in their legal systems, and of overcoming any potential obstacles. Due to the inherent limitations on any binding legislative instrument on procedural rights in the JHA area, and because of the particularly complex and subtle nature of the right to silence, the Directive contains only general provisions, which provide little if any guidance to MS concerning which situations should be avoided in order to effectively protect the right to silence, as well as how to effectively address such situations when they do arise.
Many of the issues of concern related to the right to silence described in this Special Issue exist on the level of (judicial) practice, rather than the respective legislative frameworks. This relates, for instance, to the lack of clarity surrounding the statements given outside of a formal interrogation (the Article 7(1): General Provisions on the Right to Silence section), the impact of 'non-cooperative' behaviour of the suspect on sentencing and other judicial decisions (the Article 7(4): Taking Into Account Cooperative Behaviour of the Suspect section), or the effectiveness and adequacy of procedural remedies for specific violations of the right to silence (the Article 10(2): Remedies section). Consequently, these issues can only be adequately resolved at the national level, but not via EU-wide regulation. A request for preliminary reference to CJEU, another common means of EU law implementation, would also not be a suitable way of action, or in the least of limited effectiveness. 120 The purpose of the preliminary reference procedure is to request authoritative interpretation of EU law to be applied by national courts. However, the difficulty in the above-mentioned situations lies not so much in interpreting the Directive provisions, but in their practical application to the given circumstances.
For instance, an EU-wide regulation or a CJEU ruling could propose some kind of definition of 'undue compulsion' in the context of police interrogations, but it would be insufficient to determine whether a certain approach to suspect interrogations employed in the given country includes elements of 'undue compulsion'. As shown by the example of the compliance review of the undercover 'Mr Big' investigation method with the provisions on the right to silence described in the Dutch article, 121 often such review cannot be undertaken in abstracto, but only in the context of concrete cases. Likewise, whilst an EU instrument could describe the safeguards against improper use of adverse inferences from silence, the question of whether national courts appropriately use such inferences can only be assessed in the given national context, and often in the context of particular cases. and (e) design a plan of action for MS related to monitoring the practice of effectuating the right to silence and training the respective procedural actors.
Moreover, we would argue that EU institutions, and notably the Commission, should continue a dialogue with MS concerning the challenges they might be facing in implementing the Directive in law and in practice. As argued in the opening article, the (perceived) lack of compliance of MS with the Directive 126 could result, at least in part, from the lack of clarity concerning how compliance should look. As this article and the country studies in this Special Issue demonstrate, due to the highly contextual nature of the right to remain silent, determining what constitutes 'compliance' with this right in law and in practice is not straightforward. Furthermore, providing one universal set of guidelines for compliance with the right to remain silent to all MS is unlikely to be sufficient. Rather, it seems that the situation of each MS needs to be assessed individually. Finally, as we argued in this article, compliance of MS with the Directive provisions on the right to remain silent is frequently determined by the day-to-day practices of judges (namely, by how judges interpret and apply the right to silence in individual cases), prosecutors and other relevant actors. Accordingly, this calls for monitoring the implementation of the respective right from a multilevel perspective.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Directorate-General for Justice European Commission and grant number is 802102.