Witnessing legal sources of time for better abortion care

Abstract

Moves to time-limited abortion laws have generated novel research questions as pregnant people are timed out of lawful access. Some of these questions were articulated during the successful 2018 campaign to legalise abortion in Ireland and repeal the ‘right to life of the unborn’ in Article 40 3 3 of the Constitution, otherwise known as the Eighth Amendment. Feminist and civil society campaigners asked how would the time in the new criteria for lawful abortion be counted? If the starting point of the 12-week time limit was the last menstrual period (LMP) and a moment of cyclical time, how would the ending point of the time limit accommodate rhythms and overlaps, including those of an abortion timeline? I respond to this everyday witnessing of legal problems by considering the different ways in which legal time runs. Working with different aspects of legal method, such as statutory interpretation, administrative rules, legal determinants of health, and standards of proof for evidence, I source four kinds of legal time: calendared, interruptible, health-determining and provable. I show how these different sources of legal time could be useful for improving legal standards of abortion care under the Health (Regulation of Termination of Pregnancy) Act 2018. Witnessing doctrinal legal methods openly in this way contributes to critical socio-legal methodologies by working the knowledge that laws hold, including laws that have emerged through significant social struggle.

Introduction: The legal problem of time

The adoption of time-limited abortion laws around the world has been characterised as a political compromise between value-based constituencies (Siegel, 2012), rather than as a substantive development in itself. But such developments in Ireland and elsewhere (Baird, 2024; Bergallo et al., 2018; Ely Yamin and Ramon Michel, 2023; Fletcher, 2018, 2024; Millar, 2022) generate new challenges for procedural fairness and operating optimally under the law (Conlon et al., 2022; Donnelly and Murray, 2024; Duffy et al., 2023; Enright, 2022; O'Shea, 2023). If periodic abortion law has facilitated abortion becoming lawful for a limited period of pregnancy, it has also generated new zones of exceptionality for those seeking to become unpregnant (Chakravarty et al., 2023; de Londras, 2020; Grimes et al., 2023; Power et al., 2021). Under the Health (Regulation of Termination of Pregnancy) Act 2018 (HRTPA), ¹ ‘earlier’ abortion, or abortion at less than 12 weeks gestation, has become generally lawful. ‘Later’ abortion has become exceptionally lawful since it is only available on particular grounds such as when there is a risk of serious harm to health. Those seeking access after 12 weeks must prove themselves worthy of the legal exceptions, with significant effects that exacerbate rather than relieve vulnerability (Chakravarty et al., 2023; de Londras, 2020; Grimes et al., 2023). The World Health Organisation has recognised that gestational time limits and such zones of exceptionality cannot be justified (World Health Organisation, 2022: 28). Gestational time limits generate barriers to abortion care, which only serve to make reproductive life worse (de Zordo et al., 2023; Mullally et al., 2020). But while feminism works for a world that makes reproducers free (Duffy et al., 2023; Enright et al., 2020; Gago, 2020; O'Shaughnessy, 2024; Verges, 2020), I show how law, in substance and method, actually provides a wider variety of sources of time than is typically assumed, a collection of sources which could enable diverse interpretations of timely provision. Actually existing abortion law, such as Ireland's new periodic abortion law, is the hard-won product of political struggle (Browne and Calkin, 2020). It contains knowledge that is useful for working out how law intervenes in reproductive labour time (Fletcher, 2024; Kotiswaran, 2023) and how such legal interventions could be developed.

Understanding more about the novel terrain of time-limited abortion also generates useful knowledge at the intersection of law and time (Beynon-Jones and Grabham, 2018; Chowdhury, 2020; Mawani, 2018; McNeilly and Warwick, 2022; Valverde, 2015), knowledge which we could deploy for better and ultimately transformative care. Research on legal temporality has emphasised law's active role in making, rather than passing through, time (Greenhouse, 1996). Much of this work has focused on how legal actors stretch or shape time in interesting ways as they ‘make land liquid’ and erase indigenous histories (Keenan, 2018), or ‘brew legal time’ by mixing human and non-human materials and count the biological time of cell growth by reference to drug manuals (Grabham, 2016). I show how legal sources and methods of reading them also contribute to this concern with stretching and shaping legal temporalities, but prospectively rather than retrospectively. Witnessing legal sources of time is about enabling current and future applications of those sources, and supporting users and makers of time, including reproducers of time, in holding law to account. Time, both decelerated and accelerated, has been such a problem for those seeking care and access to justice in Ireland and beyond (Enright, 2018; Ring, 2018). Life's time has been taken from reproducers through carceral interventions, familial separations and migration flows (O’Donnell et al., 2023). But that time is being reclaimed as reproducers make their own futures and refuse paths that have been ordained by others. Learning more about legal sources of time could generate helpful insights into and beyond abortion law, by speaking to questions about how time is reproduced through everyday care labour (Elias and Rai, 2019; Folbre, 2022).

Counting and measuring time have become key techniques for quantifying unpaid reproductive labour and seeking to have it recognised and valued (Elias and Rai, 2019; Rai, 2024; Waring, 1988), but typically in a context of unpaid domestic labour rather than in the specific context of pregnancy (Folbre, 2022; Kotiswaran, 2021; Kukla, 2008). Thinking through the ways in which law measures reproductive time in the context of abortion law could draw from and develop methods of valuing time spent on reproductive labour. Measuring 12 weeks, or three days or 28 days after birth, or some other time-limited period in abortion law, involves law in making choices about the value of the time spent on the work of growing embryos and foetuses (Browne, 2022; Chadwick, 2018). Watching how law responds to the following kinds of questions could generate helpful knowledge for working and reworking the relationship between law and reproductive time: When does the abortion time limit start running? What kind of clock is appropriate as a means of measurement? How precise should the measurement be when deciding whether the deadline has arrived? Does the purpose and effect of legal measurement matter for the conduct of measuring procedures? These problems of measurement are experienced differently by those actually using legal time limits everyday, including those accessing, providing, supporting and regulating time-limited abortion care. Their questions, and answers, about the effects of time limits may be helpful in reflecting on, developing and reviewing ‘good legal practice’ for applying gestational time limits.

Drawing from science and technology studies, and methods of modelling the making of scientific knowledge, I adapt Haraway and Murphy's approach to ‘witnessing’ knowledge generation (Haraway, 1997; Murphy, 2012) for legal purposes. I apply witnessing to the techniques of conventional doctrinal analysis and develop a method of open legal witnessing (Fletcher, 2021), or ‘law work’ (Enright et al., 2020), in three steps: characterising the factual problem, identifying relevant legal rules, applying relevant rules to the problem in order to generate new and different legal applications. Such a method allows me to approach different legal sources with a view to reading them for knowledge about how time works. Witnessing legal method in this way assumes a non-legal audience, contributes to the commons of knowledge (Fletcher et al., 2017; O’Donovan, 2015; Ross, 2017) and seeks public accountability of legal expertise. Moreover, given that everyone is working on novel legal terrain, and needs to pilot, review and reflect on new legal standards of care, there is merit in being explicit about how different legal sources are relevant for reading and applying time-limited abortion law.

In summary, I identify four sources of legal time that could be helpful in addressing factual problems that have been identified with time-limited abortion law, and I make some suggestions as to how those sources could be applied to the problem. By identifying possible applications of legal sources of time in this way, I hope to provide some legal tools that will need further experimentation, analysis and review in order to assess how well they work (Cloatre and Pickersgill, 2016). The first sourcing of time shows how law uses cyclical as well as clock-driven time, when it measures time by reference to cycles such as menstrual cycles, and necessitates a calendaring of cyclical and clock moments (Fletcher 2024). The second sourcing of time is one which finds that law recognises that timelines overlap with each other and may need to be interrupted or suspended in order to facilitate each other. The third collection of sources comes to us from the public health literature, which recognises that laws and legal provisions themselves can be determinants of how time runs or is experienced (McGuinness and Montgomery, 2020). The fourth collection emerges from legal standards concerning evidence (Fennell, 2020), and the different expectations of evidentiary proof for determining whether a particular legally relevant event has happened or not. These different sources of time are identified for the use of others through a methodology of legal witnessing as open doctrinal analysis.

Witnessing sources of legal knowledge

In feminist science and technology studies, witnessing involves multiple adaptations of conventional scientific knowledge generation, adaptations that repeat aspects of scientific experimentation, modelling and monitoring. While Haraway was interested in revealing the conventions of the scientific expert for challenging the modesty of the impartial witness (Haraway, 1997), Murphy identified how feminists adapted scientific protocols ‘immodestly’ through practices such as vaginal self-examination (Murphy, 2012). In this way, feminist witnessing identifies different possible knowledge uses and outcomes, and encourages engagement in making some future knowledge outcomes more plausible than others. I turn to doctrinal legal analysis to show how conventional legal methods (Egan, 2017; Kilcommins, 2017; Mhuirthile et al., 2016; Schweppe, 2011) may be adapted and repurposed in generating critical knowledge about how law works. This legal witnessing builds on earlier work that showed how legal knowledge may be reproduced through feminist activist interventions with legal consciousness of key concepts such as ‘choice’ (Fletcher, 2020, 2021). It is not ‘my method’, but one that runs through the feminist commons of knowledge (Federici et al., 2021; O’Donovan, 2015; Ross, 2017) and has had many iterations as ‘know your rights’ initiatives and legal explainers, feminist judgments (Enright et al., 2017) and more.

Witnessing doctrinal analysis reproduces law otherwise by adapting legal method for ‘open’ social justice purposes. This open doctrinal method is one that entails (1) demonstrating the basic steps in doctrinal legal methods of knowledge production for non-legal audiences, (2) showing how doctrinal analysis generates useful legal knowledge, which may be deployed for non-doctrinal purposes, e.g. for critical and contextual problem-solving, and (3) generating proposals about how that legal knowledge, including legal standards for behaviour, could be applied in a given factual situation. In short, witnessing law is a public-oriented approach to legal education that adapts basic doctrinal method for the purposes of informing everyday legal practice from a social justice perspective. This method differs from conventional doctrinal analysis because the former is typically focused on deriving the ‘right legal answer’ to the problem based on an analysis of legal rules, whereas critical doctrinal analysis recognises that there are typically a variety of plausible legal answers to a given problem, and some will be more plausible than others because of material power relations (Blalock, 2015; Harrington, 2016).

In repurposing the method of doctrinal problem analysis, I reflect on the problem of time in time-limited abortion law, identify legal sources of time that might be useful for tackling that problem, and propose possible ways of applying those sources to the problem. Conventional doctrinal method responds to fact-based legal problems (such as the appropriate response to a pregnant person whose early lawful abortion with pills has failed) typically in four steps, often called the IRAC method (Whalen-Bridge, 2014). The IRAC method involves identifying the Issues, Rules, Application and Conclusion of a client's fact-based problem. In identifying the legal 'issues' that emerge from a problem, law students are taught how to frame the problem in legal terms (Mhuirthile et al., 2016; Schweppe, 2011) and how to articulate legal questions or issues that emerge from the factual problem a person or community might have.

Conventional legal problem analysis acknowledges but backgrounds the ‘material facts’ out of which the legal issues emerge, and to which the legal rules are applied. This is why one technique of critical legal practice developed by feminist judgments projects has been to flesh out the facts of the case in the drafting of legal interventions (Enright et al., 2017). Understanding more about the actual lived experience of litigants, as they find their lives being poured over by judges and lawyers, usually renders the legal rules less abstract and makes them more accountable to real flesh and blood people. I draw on this technique to foreground the material facts as the first step in problem analysis and I suggest that describing the facts always already involves choice about what facts count, or which facts are material. As a result, characterising the factual problem is the first step in the legal method of problem analysis adopted here, in ways that are accountable and responsive to the needs of law's everyday users.

The second step of conventional problem analysis involves identifying the ‘relevant legal rules’ that address the question. Some of these legal rules will be obvious, some could be more obscure. The rules generated by HRTPA 2018 are an obvious set of rules to apply to a problem someone is having about accessing or providing an abortion. But other relevant legal rules such as constitutional rights (Enright, 2022; Fletcher and McGuinness, 2017), or rules of statutory interpretation, or legal determinants of health, or rules of proof, might be less obvious, particularly to those not trained in legal methodology. Some legal rules will be binding and authoritative as a matter of law, others will be persuasive. Whether a legal rule is binding, e.g. a woman cannot be lawfully stopped from accessing an abortion in the first 12 weeks of pregnancy, depends on the status of its source, i.e. is it based in a binding primary legal source such as a statute or a precedent from case law, and on the fit of the rule with the fact situation. If the fact situation raises slightly different considerations than those envisaged by the statute or the precedent then it might be possible to distinguish the binding legal rule and adapt it to the situation.

Persuasive or non-binding sources of law are regularly used in legal methods, if in irregular ways. A judicial comment in a judgment, an obiter dicta, is a conventional example of a non-binding legal source. Academic research is another example of a non-binding source that might be persuasive in being a research-based articulation of how law should be applied but does not invoke the authoritative status of binding legal sources. Policy and custom and practice are other kinds of non-binding legal rules that also feature in legal problem-solving, sometimes explicitly and sometimes implicitly. Conventional problem analysis is hierarchical in its approach to these different sources of rules and demands that most weight is given to those that are binding in the sense of being authoritative about their application with most legal authority. But there are often gaps in how binding legal rules apply, and non-binding sources regularly feature in official legal work. An open witnessing of legal rules can acknowledge this convention, identify obvious and obscure legal rules that are relevant, and clarify non-binding sources of rules that might also be helpful for problem-solving.

The third and fourth steps of a conventional legal problem analysis are typically ‘Application’ and ‘Conclusion’. An open method of legal witnessing is open in not having to have a conclusion in the narrower sense of presenting a concluding comment with the ‘right legal answer’. Rather an open method ends with the third step of showing how relevant rules may be applied to the material facts and clarifying what these various possible applications might be. Even if a particular legal rule is very ‘settled’ in the sense of being clear and authoritative, a novel fact situation can unsettle it. Applying relevant legal rules to a novel factual situation is conventionally understood as generating a ‘new’ legal rule in a practical everyday law sense, even if the common law method might claim that the rule was always there, waiting to be revealed. In Attorney General v X [1992] IR 1, the key pre-repeal abortion case, the Supreme Court generated the test of ‘a real and substantial risk to the life of the woman’ as the legal rule that set a standard for determining when a pregnant woman would be entitled to end her pregnancy under the Eighth Amendment (de Londras, 2015; Fletcher and McGuinness, 2017). This was a ‘new’ or newly articulated legal rule, which derived from the application of the right to life of the pregnant woman under the Eighth Amendment to the novel fact situation of having a young suicidal pregnant woman seeking abortion before the court. X had a right to abortion in Ireland because the pregnancy threatened her life as a matter of fact. Therefore, applications of rules to facts are powerful legal techniques, which may involve repetition, but may also innovation, and certainly involve assessment and accountability to the facts of the situation. The difference between conventional and critical doctrinal method is that the latter recognises that there will be a plural range of applications and that what makes some more plausible than others is not typically a legal matter but socially determined. Recognising this means that we can intervene by testing out different applications, reviewing them, and assessing how well they work for feminist purposes and promoting access to justice on the ground. Through active legal witnessing in this way, we can develop practices that strengthen legal protection for providers and users as we all work to make existing abortion law the best it can be for abortion-seekers.

Sourcing cyclical, clock and calendar time

One of the key problems identified by the O'Shea review of the operation of Ireland’s new abortion law (Conlon et al., 2022; Duffy et al., 2023; O'Shea, 2023) is that of timing out of lawful abortion care (O’Shea 2023: 89). Many people have had and will have trouble with fitting the factual experience of pregnancy into the legal timeline for abortion care (McGuinness, 2024; O'Shaughnessy et al., 2021, 2023). Questions were raised during the campaign to legalise abortion about how the 12-week time limit in (what is now) section 12 ² on ‘early abortion’, and the 3-day waiting period, are to be measured, questions which make the complexity and ambiguity of legal time evident. ³ Nonetheless, to date, this question has been treated as having a self-evident answer by the policy actors who have been tasked with answering it. Moreover, the answer given by the Chief Medical Officer (CMO) ⁴ has involved counting 12 weeks as 12 weeks + 0 days, and a shorter period of lawful abortion time, than the 12 weeks + 6 days which is how 12 weeks is typically counted as a matter of clinical practice. Here I argue that the CMO is implicitly adopting a clock approach to time by going for an interpretation of the time limit which is more mechanical and predictable in choosing a precise +0 days cut off point, rather than one which is more embodied and variable allowing for +6 days period of transition from 12 to 13 weeks. ⁵ But such a clock approach to the question of how to count legal time is not as self-evident as it might seem. There is more than one way in which material time runs as a matter of law. Various everyday witnesses of the legal reform process knew this when they asked questions about how legal time would be counted.

Early in the preparation for implementing the new Act, civil society organisations and clinicians raised questions about the operation of the 12-week time limit and the 3-day waiting period. Clinicians pointed out that in clinical practice typically 12 weeks means 12 weeks plus 6 days, or not yet 13 weeks. Feminist and social justice groups pointed out that counting gestational time from the date of a woman's last menstrual period (LMP) meant that the starting point of pregnancy had changed as a matter of law from conception to LMP. This legal change means that the 12 weeks counts unconscious gestational time in two different ways, retrospectively from conception when pregnancy is unknowable because not established, and retrospectively from a missed period or equivalent sign when pregnancy becomes knowable. As a result of pregnancy legally starting with LMP, a woman or pregnant person could be pregnant as a matter of law before they have the heterosex or insemination needed to make them pregnant. For two weeks of the 12 weeks (or any other qualifying period of gestational time) they are not actually pregnant when pregnant time is running contemporaneously because they have not yet conceived. They become pregnant retrospectively as law looks backward from the identification of a pregnancy to the first day of their last menstrual period as a starting point. And for four to six weeks they may not know that they are pregnant because they have not missed their menstrual period or acquired some other sign of pregnancy. In other instances, a lack or change of knowledge about the pregnancy can also contribute to much of the pregnancy time not being abortion-conscious time. If the pregnant person does not know that they are pregnant for at least part of the 12 weeks this creates problems for the fair application of the key time-limited rule permitting access to abortion.

Clinicians, women and civil society groups also raised questions about the meaning and significance of three days in the mandatory 3-day waiting period. Having to wait three days to have an abortion when up against the deadline, e.g. at 11 weeks and 4 days, could take someone over the deadline and render them ineligible for the abortion for which they had been approved. And applying the three days from a clinician's point of view could be difficult if the three days began on a Thursday and had to accommodate a weekend off and so in effect turned into 5 days, again making it more difficult to meet the deadline. Being able to count 12 weeks as 12 weeks + 6 days could have significant practical effect in helping to minimise the harsh effects of the 3-day waiting period. The possibility that the 3 days could count as 3 working days with the first appointment happening during the hours of the first working day and the second happening during the hours of the third working day, was aired during the run-up to reform. Another variation that raised the possibility of scheduling the 3 days beginning from the pregnant person's first point of contact with their abortion provider was also discussed. ⁶

In December 2018, Peter Boylan, the doctor given responsibility for advising on the setting up of the abortion service, wrote to the Chief Medical Officer to ask when 12 weeks ended. In replying to ‘the issue that has arisen in relation to the definition of 12 weeks’, the Chief Medical Officer's answer focused on the verb ‘exceeded’ in the relevant statutory provisions and gave a view in light of the meaning and significance of ‘exceeded’ rather than considering the issue of how to count 12 weeks more broadly. The letter stated: ‘For the avoidance of doubt, I wish to clarify that the legislation states “… that the pregnancy concerned has not exceeded 12 weeks”. Therefore 12 weeks is 12 weeks + 0 days’. The CMO adopted a mechanistic approach to counting time as if three days of time flowed evenly over 72 hours, rather than as if a period of time may stop and start, accelerate and decelerate. This answer does not address how the legal period of time itself is to be measured, or what legal standards might be relevant to such measurement. Neither did he address the significance of such a finding for achieving one of the objectives of the Act, i.e. removing barriers to lawful abortion care and facilitating women's negative statutory rights to abortion under the Act. Given these gaps in the legal analysis underpinning the CMO’s answer, it is possible to revisit the issue and consider how legal time could be measured in the round.

The Chief Medical Officer refers to the statute as the key relevant binding legal source, and identifies the language of ‘must not exceed 12 weeks’ as providing an answer to the question. But there are other aspects of section 12 and the statute at large that may be relevant and helpful in answering this legal question of how to understand 12 weeks (or 3 days or 28 days after birth), and there are rules of statutory interpretation which he, or his legal advisors, could have used more extensively. Statutory rules of interpretation allow those determining a statute's meaning to pay attention to the purpose of the statute and its provisions, as well as to the ordinary and specialist meaning of its words when legal actors are tasked with deciding what a statutory provision means. Here I am going to confine myself to focusing on the literal rules of interpretation and the usual techniques/methods that are used in identifying the ‘ordinary meaning’ of a statute.

As Ní Mhuirthuile et al. explain (Mhuirthile et al., 2016), the ordinary meaning of law is distinct from specialist meaning, where the latter refers to a meaning that has been given to a particular phrase by a specific legal source or practice for a specialist legal purpose. Theft might have an ordinary meaning in everyday language, but it also has a specialist meaning that it has been given by statute and common law. If the terms do not have a specialist legal meaning, however, ordinary legal meaning applies. It is common to read one statutory provision in light of others when coming to a view about what the meaning of a particular provision is, especially where there are a number of possible interpretations. Section 12 explicitly uses the cyclical time of bodily rhythms when it is counting out legal time because it states that pregnancy begins on the first day of the woman's last menstrual period. Those operating and interpreting the provisions of the statute, including how its time limits run, could legitimately point out that the statute itself draws on cyclical time, and therefore provides cyclical time as a source of legal interpretation. While the Act refers to 12 weeks as a time limit and uses abstract clock time in that sense, it also uses cyclical time and the first day of the last menstrual period to count the beginning of the 12 weeks, and it brings calendar time into being by anticipating the scheduling of other moments (Fletcher, 2024), such as certification and waiting periods, along the way.

Legal time limits do not in themselves create an obligation to count a particular period of time, such as 12 weeks, in an exclusive clock-centric way. Rather the statutory source of the time limit in this instance explicitly recognises that cyclical time fixes a starting point for the time limit to run. The 12-week time limit is a calendared period of time that draws on abstract and concrete moments to plot a calendar that begins with menstrual cyclical time, ends with the clock, and is punctuated with moments of professional certification and waiting. Given that the statutory time limit itself draws on these different kinds of material time, there is no reason why the 12-week end point could not be measured in a way that understands time as rhythmic and scheduled in its accommodation of an attachment to predictable progress and unpredictable interruption. There is nothing self-evident or legally necessary about using a mechanical clock approach, rather than a lively cyclical approach, to measure the 12-week time limit.

Sourcing timeline overlaps and interruptions

The O'Shea review recognises the problem of people timing out of the 12 week deadline in the specific case where they have had a failed medical abortion (O'Shea, 2023: 90). Such a possibility of abortion failure – predicted in 1-2% of cases of abortion with pills – makes the question of how the 12-week time should run, practically significant. Indeed the Health Service's consent form for early medical abortion makes the stakes explicit. It asks abortion users to accept that it is illegal for a doctor to provide an abortion unless the other grounds apply ‘if my abortion fails and I have an ongoing pregnancy that goes beyond 12 weeks of pregnancy’. ¹⁰ There are clearly some significant problems with the status quo which have been recognised by providers, users and O'Shea herself. Through a predictable effect of the drugs, a woman who has acted on her wish to end the pregnancy has not achieved her goal and is now carrying a pregnancy that may have been harmed by the drugs. I cannot address the issue fully here, but it is worth noting that denying a repeat abortion in such circumstances may be a breach of her constitutional and human rights since her rights to bodily integrity, privacy and dignity are engaged and failing to provide her with lawful abortion potentially breaches her procedural and substantive rights under Article 8 ECHR (Enright, 2022). It is also possible that such abortion care is being delivered at sub-optimal professional standards. But professionals are being legally prevented from providing conscientious care (Donnelly and Murray, 2024) by this interpretation of the 12 week time line as continuous uninterrupted time.

Rules of procedural fairness mean that there will be occasions when it is fair to stop the clock, not least because there has been some intervening event that requires it. In a paid labour context, if an employee is contractually obliged to clock in for 35 hrs a week, they may be allowed to have that time arranged ‘flexibly’ or in a pattern of attendance that suits their care obligations and not just those of their employer. One legally obliged timeline may be interrupted or paused by another. Scholars of time have noticed how legal time may run between events rather than as an uninterrupted flow (Greenhouse, 1996; Beynon-Jones and Grabham 2018). If legal time runs between events then that means that it can stop and start, accelerate or decelerate, intensify or relax. This sense of time as moving between events and things is one that informs administrative procedures which allow the interruption of a clock that is timing one particular process when another process intervenes or overlaps. If other legal processes allow interruption of the clock as it counts down a legally defined period, then in principle the legal process of applying the 12-week deadline could also accommodate the practice of suspending the timeline or interrupting the clock. When different life events have different timelines that may clash, this can be accommodated as a matter of law. The repetition of a medical abortion in cases of predicted failure is legally possible and does not have to be excluded as the consent form assumes.

Part of the problem with the consent form's interpretation seems to be that the legal timeline is being measured here by reference to a foetal development timeline, rather than by reference to the timeline initiated by the abortion as an act intervening into and interrupting the pregnancy. The description of ‘having an ongoing pregnancy’ is determined by the fact that the abortion although initiated has not succeeded in expelling the contents of the womb and so the pregnancy is ongoing because the products of conception are still present in the womb. This is a foetocentric way of counting whether there is an ongoing pregnancy or not since it determines the ongoingness by foetal presence. However, from the pregnant person's point of view the pregnancy is not ongoing because they have already taken steps to interrupt that pregnancy and they are now in the process of having an abortion, an abortion that has been started, has failed, and that needs to be repeated in order to be completed. From the pregnant person's point of view the gestational timeline has been interrupted, and arguably suspended, by the abortion's timeline as an intervening life event. If the abortion was initiated within the lawful period of 12 weeks, then the 12-week timeline could be paused or suspended while the abortion's timeline is enabled to come to completion.

Could such a foetocentric perspective on measuring the timeline of a failed abortion be justified by reference to the statute or other legal sources? If we look at the statutory provisions allowing abortion as a whole, we can see that section 11 is the only one that adopts a foetocentric approach in grounding the access criteria in issues of foetal anomaly. Section 12, the section at issue here, centres the pregnant person as the relevant legal subject. Moreover, the general purpose of the referendum and the legislation was to change the constitutional framework from a foetus-centred perspective to one centred on the pregnant person. And in M v Minister for Justice and Equality [2018] IESC 14 the Supreme Court held that there was no implicit constitutional protection of unborn life in a way that gives rise to state protective obligations. An ‘unborn child’ did not have any constitutional rights beyond the right to life as then enshrined in Article 40.3.3 of the Constitution, and since repealed. Adopting a woman/person centred perspective for the purposes of calculating the 12-week timeline in cases of failed abortion means recognising that the act of initiating the abortion interrupts or suspends the gestational timeline, while the abortion's timeline has to be allowed run its course and achieve its objective. If the abortion was lawfully initiated within the 12-week time period but failed, then repeating the procedure is necessary to achieve the legal objective of terminating the pregnancy, and any such repetition ought to be regarded as happening within the suspended 12-week timeline.

The position that a failed abortion initiated within 12 weeks cannot be repeated, in the absence of exceptional grounds being met, adopts an illegitimately foetocentric perspective for calculating the timeline. It assumes that gestational time continues as before because the foetus is still present in the womb, rather than regarding gestational time as being interrupted or suspended once a lawful abortion has been initiated for the purpose of making the woman unpregnant. But such an implicitly foetocentric version of how to apply the 12-week deadline in cases of failed abortion with pills is not the only possible interpretation. Rather fair procedures in the administration of legal timelines typically recognise that there may be occasion to interrupt the timeline, either as a result of other rights which the person has, such as the right to request flexible working arrangements for care activities, or as a result of accommodating exceptional circumstances which make rigid adherence to the timeline unfair. On this understanding of interruptible material time, the 12 week gestational timeline could be legally suspended by the intervening act of the abortion and the abortion once lawfully initiated with the pregnant woman could be allowed to finish.

Sourcing health-determining time

The O'Shea review identifies that there is significant confusion about the scope of section 9 ⁸ and that fewer people are accessing abortion on the ‘risk to health’ pathway than might be entitled. The risk to health ground is likely being under-used (O'Shea, 2023: 53–58). The legal standard for section 9 requires that doctors must identify that the pregnancy poses a risk of serious harm to the health of the pregnant person, or a risk to her life. The problem at the moment is that the threshold for this intervention is very high in practice and less people are accessing abortion on this ground than could be the case (Grimes et al., 2023). Lack of clinical guidance is contributing to a situation where some doctors seem to be implementing this ground as if it requires a risk of death and as if it mirrored the pre-repeal order (see further O’Shea 2023: 56), rather than enabling a broader range of threats to health and life to be considered.

Continuing pregnancy in a range of different circumstances may pose a risk of serious harm to health and meet the terms of access under section 9. Such circumstances include two situations that were much discussed during the referendum campaign, pregnancies in circumstances where the pregnancy resulted from sexual violence, or pregnancies where there is concern over a significant foetal anomaly. While pregnant people in these circumstances will be able to access abortion if they meet the terms of section 12, they will only be able to access abortion after 12 weeks if they meet the other legal grounds. There is significant evidence that people in such circumstances are not being found to meet the other legal grounds, and are having their rights infringed in the process. But access could be lawfully enabled on grounds of a risk of serious harm to health through more careful fleshing out of the different kinds of experiences that generate such a risk (Baird, 2024; Bellamark and Paterson, 2023; Enright, 2022; World Health Organisation, 2022: 26–27). Moreover missing the 12-week deadline, due to factors such as trauma-related delay in recognising pregnancy and the need for abortion, could also be recognised as an aggravating factor that requires a policy intervention in the name of harm reduction, rights protection and/or justice promotion.

It is well recognised now that legal rules and processes may contribute to ill health and may require evidence-based assessment and human rights evaluation. In the realm of public health this phenomenon is increasingly framed as a question of ‘legal determinants of health’ (Gostin et al., 2019; McGuinness and Montgomery, 2020; Thomson, 2022). Although Gostin et al. emphasise the positive harm-reducing role that law can play in determining health, McGuinness and Montgomery point out the negative impact that law may have on health, particularly in the context of stigmatised care such as abortion. The move towards abortion's decriminalisation recognises that criminalisation has contributed to ill health and vulnerability, since criminalisation of abortion contributes to worse health outcomes for women and pregnant people. Decriminalisation is a mode of public redress since it recognises the harms caused by criminalisation and removes them. A risk to health ground could operate as a more positive legal determinant of health, in the way imagined by Gostin et al., if it is operationalised in a way which reduces vulnerability to ill health, including situational vulnerabilities generated by poverty or gendered violence.

The current under-use of section 9 is an example of how restrictive interpretation of the risk to health may negatively determine health by requiring that pregnant people become more ill before qualifying for an abortion. There is certainly scope for a wider range of abortion care under section 9. Mitigation of the risk to health imposes an obligation on medical professionals to consider vulnerable circumstances, such as poverty, homelessness, and experiences of racism and gendered violence, that are capable of generating risks of serious harm to health. Medical professionals could be supported in using their legal power under the Act to convert the evidence of a risk to health arising from vulnerable circumstances into a recognised ground of abortion. Secondly, there is potential for a fast-track protocol given that missing the 12-week deadline may be an aggravating factor that could exacerbate ill health. Delay is a determinant of health and those who are beyond 12 weeks of pregnancy are likely to have any risk to their health exacerbated by the passing of time. The 12-week deadline is itself a legal determinant of health.

Sourcing proofs of timeliness

O'Shea found that women and couples were not accessing abortion under the section 11 ground even though they may legally qualify, because professionals felt that they had to be ‘certain’ in predicting future death of the foetus or child after birth (O'Shea, 2023: 9). Section 11 imposes a limit of 28 days after birth as the horizon that counts for deciding whether abortion will be lawful or not on grounds that there is a condition where the foetus is likely to die. ⁹ The certifying practitioner is required to assess the future likelihood of foetal or neonatal death on the basis of diagnostic information available to her in the here and now about the condition of the pregnancy. ¹⁰ Clinicians respond by identifying a range of conditions that are likely to have these consequences if diagnosed (Power et al., 2021). While we know that this pathway is the second most frequently used pathway and that there clearly is a need (Grimes et al., 2023; Jackson et al., 2023; Power et al., 2021), we also know that this pathway is failing those who are in significant distress about the development of their pregnancies in a direction that they find unsupportable. ¹¹ In an environment where doctors operating this legal test are heavily scrutinised, and fearful of falling outside of the law, they are holding themselves to a very high standard of proof. The question is whether this level of near ‘certainty’ is legally required when coming to a view about the presence of a condition likely to lead to death of the foetus within 28 days of birth, as stipulated by section 11 of the 2018 Act. What are legal sources of proof and predictability on the facts that could be applied to the question of deciding whether such a condition has been found or not?

When doctors are making assessments under section 11, they are providing lawful clinical care in good faith in light of the available clinical evidence. Therefore, their conduct is legally regulated by civil law standards, as it is the civil law of tort that holds their care to account. If civil law standards apply, then the civil standard of proof (on the balance of probabilities) applies to any evidential requirements (Dennis, 2020; Fennell, 2020; Roberts and Zuckerman, 2010) rather than the criminal standard of proof (beyond a reasonable doubt). This means that the doctor is not required to be ‘certain’ that foetal death is likely when operating according to civil legal standards. Rather the doctor has to come to a view that foetal death is more likely than not. This is a likelihood ¹² that is typically quantified as in more than 50% of cases but could also be proved by reference to qualitative evidence about the kind of condition.

The Irish courts have established a civil standard of proof that is higher than the ordinary one, but this applies under the Mental Health Act 2001 and relies on statutory language requiring ‘serious likelihood’ rather than ordinary likelihood. In MR v Byrne [2011] 3 IR 211 the High Court considered the question of what standard of proof applied where the statute provided that a person could be detained under section 3 of the Mental Health Act 2001 where there is a ‘serious likelihood’ of the person causing serious harm to himself or others because of illness, disability or dementia. O’Neill J in the High Court ruled that the phrase ‘serious likelihood’ meant that a higher standard of proof than the ordinary standard of proof in civil actions applied. This ‘high level of probability’ standard ‘went beyond the normal standard of proof in civil actions of more likely to be true’, but it falls short of the standard of proof that is required in a criminal prosecution namely beyond a reasonable doubt’ (Moriarty, 2015: 35). Since the 2018 Act refers to the condition being ‘likely’ and not to ‘serious likelihood’, it is arguable that the ordinary civil standard of proof, and not the higher than ordinary standard, applies.

The criminal standard of proof ‘beyond a reasonable doubt’ applies when someone is being prosecuted and/or tried for a criminal offence. This is the standard closest to the virtual certainty that doctors are requiring of themselves when assessing the evidence under section 11. And if doctors were to be prosecuted under section 20 for committing an offence outside the terms of the Act, the criminal standard of proof is the one that would apply. But it is not the standard of proof that should apply when doctors are engaged in ordinary lawful care under the Act. Nonetheless the threat of criminalisation is contributing to a chilled atmosphere where doctors are afraid to provide care to which pregnant people are lawfully entitled. This atmosphere has contributed to a blurring of the boundaries around the different possible standards of proof that can apply in different legal moments of healthcare. If likelihood (that a future foetal death would happen) was to be proved by reference to the kind of evidence-based approach to event time that is reflected in the civil standard of proof ‘on the balance of probabilities’, that would require showing on the facts that foetal death was more likely than not, or more likely in 50+% cases (Pardo, 2019; Symon, 2007). Different standards of legal proof mean that there are different sources of provability for calculating retrospectively whether the running of legal time has contributed to a legal event, or whether it will prospectively contribute to a future legal event. Doctors do not have to hold themselves to a legal threshold of certainty or near certainty when it comes to authorising access to a future abortion for conditions likely to lead to death of the foetus.

Conclusion: Towards future applications of legal sources of time

The adoption of time-limited abortion law brings novel legal criteria to the regulation of abortion care, criteria which raise questions about how sources of time could and should be applied. Social actors raised questions about how time runs as a matter of law during the campaign to legalise abortion in Ireland. They have begun a process of identifying key reference points for answering such questions. Here I have drawn on and developed a methodology of legal witnessing for the purposes of contextualising those reference points and identifying legal sources that could be useful for answering such questions. I do not provide legal answers as such here, because critical socio-legal knowledge production recognises that ‘legal answers’ are not provided by law alone and are inevitably the product of the power relations in which they are enmeshed. In such circumstances, working out what the legal answers should be is a process that needs to be open and conducted with others who use the law. Rather I have focused on demonstrating a method of open doctrinal analysis and showing how it could generate knowledge about sources of legal time which could be useful as part of this broader process.

Open doctrinal analysis responds to a legal problem by first characterising the ‘material facts’ on which the problem is based, second by identifying relevant legal rules – binding, persuasive, analogous – for problem-solving, and third by clarifying different possible and plausible applications of the rules to the facts. Using this method, I turn to the O'Shea review and other research into the operation of Irish abortion law to identify particular factual problems that indicate abortion care is not as good as it could be under the law. The problems concern cases of timing out of the calendared pathway, including in cases where medical abortion fails, under-use of the risk to health pathway to abortion, and practitioner unease at the standards of near certainty required of them in predicting future foetal death. Second, I identify relevant legal rules for addressing these problems including the statute itself, and rules of statutory interpretation, procedural fairness, health determination and standards of proof. Third I show how reading the problems in light of the knowledge contained in these rules, generates different possible applications of different sources of legal time. These legal sources know time as a material process that is calendared (with cyclical and clock moments), interruptible (for an overlapping timeline), health-determining and provable according to different legal standards.

If legal time is calendared and accommodating of cyclical time, then the end point of the time limit could fade rather than end abruptly and could operate as 12 weeks plus 6 days rather than 12 weeks plus 0 days. If legal time is rich with multiple overlapping timelines, then one timeline such as the gestational timeline can be suspended, and allow for repeat of a failed medical abortion, when another timeline such as the abortion timeline has begun but not completed within the legal 12 week period. If the passing of legal time can make health better or worse, then delay in access may become evidence of a risk of serious harm to health and trigger this exceptional ground for access as available after 12 weeks of pregnancy. And if the likelihood of a future event, such as foetal death, happening can have different standards of proof, including the civil ‘on the balance of probabilities’ standard, then doctors applying the fatal foetal anomaly ground do not need standards of virtual certainly for their prognosis. They need to establish that it is more likely than not that the foetus will die. These applications are obviously not full answers for the problem of restricted abortion and the injustice of making reproducers unfree. But they do provide sources of knowledge for holding law to account. The next step is to enable and support the development of better abortion practice in testing the operation of these different possible interpretations.