Court Forms as Part of Online Courts: Elicitation and Communication in the Early Stages of Legal Proceedings

The article explores court forms as an interactive genre essential for legal-lay communication in civil and family proceedings: court forms elicit key information from predominantly lay users for the purposes of court administration and the judiciary. The information presented in court forms defines the agenda and communicative focus of the subsequent hearings and settlement negotiations, and in some instances even the path the proceedings would take. It is thus important to consider court forms in terms of their comprehensibility as well as functionality for eliciting legally coherent narratives and facilitating efficient engagement of lay participants with the proceedings. The case study presented in the article draws on the combination of corpus linguistics and discourse analysis to explore two versions of the court form most frequently used by self-represented parties in England and Wales, ‘Form C100: Apply for a court order to make arrangements for a child or resolve a dispute about their upbringing’. The comparison of the paper form and its redesigned online version identifies the improvements made as part of the digitisation process but also indicates communicative challenges which can prevent the lay person from presenting the relevant information. The discussion provides an opportunity to reflect on the role of language in online courts.


Introduction
The existing research on language and the law has contributed to our understanding of a wide range of lexical and syntactic complexities in written legal texts [10,11,28,38,42,48,50,67,70]. The linguistic complexity of legal discourse has been linked to two key processes: the inherently conventional nature of legal language (e.g. reflected in complex lexical constructions and archaic forms) and the requirement for legal texts to mirror the complex reality (e.g. reflected in complex syntactic constructions, such as conditional sentences with multiple if-clauses or main clauses). Both aspects create barriers for the lay audience when it comes to active participation in legal discourse [26,74].
Communicative challenges for the layperson have so far been studied in relation to written legal discourse, such as contracts and legislation [60,65,66,70] or other legal documents [22,41,68,71]. A related strand of research is on legal discourse which is delivered in the spoken mode but often draws on the written mode, such as police caution [2,40,54,58] or jury instructions [19,21,35,72]. The emphasis in the above areas tends to be on the comprehension of legal texts by the layperson (1) as a passive addressee (e.g. the improved comprehension allows consumers or contract participants to make informed decisions) or (2) as a limited participant (e.g. the comprehension of Miranda rights allows a detainee to make decisions on what to say but not control the course of their case).
This article considers lay people who are not only the recipients of legal texts but also active contributors to the communication process during legal proceedings, i.e. people who represent themselves in legal proceedings-litigants in person (LIPs) or self-represented litigants (SRLs). The high numbers of SRLs across different jurisdictions and countries have made semi-represented and fully unrepresented hearings a standard rather than an exception in civil and family proceedings [e.g. 57,78]. Their journey towards accessing justice services starts with the act of completing a court form (for claimants/applicants) or responding to a court form (for respondents). From the linguistic point, the pre-requisite for accessing justice as a self-represented party, is thus comprehension of court forms and accompanying documents.
The article thus deals with court forms in their function of a communication tool between legal and lay audiences (i.e. the judiciary/lawyers and self-represented litigants or lawyers' clients). 1 For the judiciary, the main role of court forms is to elicit key information, which consequently impacts the court's decisions on the case trajectory [55]. From the court users' point of view, court forms present an opportunity to frame satellite narratives and the first reiteration of the master narrative [28: 155] within the relevant legal and professional discursive conventions [8]. The challenges SRLs experience in these initial pre-court stages 1 3 Court Forms as Part of Online Courts: Elicitation and… go beyond comprehension and extend to discursive challenges as they need to apply the generic guidance to their individual circumstances and contextualise their case within the boundaries of the relevant legal concepts [73]. The study presented here thus addresses two gaps in the current research: the insufficient focus on the lay court users and the limitations of focusing solely on the comprehension process when it comes to legal-lay communication. These aspects gain on significance in relation to the growing number of digitised court applications: online forms are designed to be user friendly for lay court users as the baseline users without the pre-requisite legal or procedural knowledge.
The article focuses on two versions (downloadable pdf version and an online version) of the most frequently used court form in family courts in England and Wales, i.e. "Form C100: Apply for a court order to make arrangements for a child or resolve a dispute about their upbringing". The rationale for focusing on this particular court form, apart from its frequent use, is three-fold: the ratio of SRLs is highest in private family law cases, for which the chosen court form is designed [25]; cases related to child arrangement orders tend to require at least several hearings and proceedings sometimes start anew as children's needs change with age; the private and public child-related proceedings are currently being digitised and the court form C100 was the first one to be introduced as part of the online application process [15]. Contrasting the paper version of the court form to its online counterpart allows us to reflect on the user experience with the two media, explore elicitation techniques, and provide an insight into strategies for supporting the users' conceptualisation and construction of satellite and master narratives [28: 155]. The wider objective of the article is thus to reflect on (1) the challenges lay participants experience when engaging with legal language in court forms; (2) the improvements and limitations the digital paths to justice currently offer to court users; (3) the importance of embedding procedural guidance and offering genre-specific support to enhance the court user's discursive competence.

Court Forms in Civil and Family Proceedings
In the context of civil and family proceedings in England and Wales, there have been several developments in the last decade which had an impact on how court forms are compiled and administered by HMCTS (His Majesty's Courts and Tribunals Service) as well as how and by whom court forms are being completed. The first change occurred as a result of legal aid cuts [44], which came into force in 2013 and had a direct impact on the increase in the numbers of LIPs/SRLs as many people found themselves with few options but to represent themselves in court without a solicitor or barrister. In family courts, for instance, the number of hearings where at least one of the parties is not represented has risen from app. 55% in 2012 to 81% in the period between April-June 2022 [25]. Court forms are thus often being completed by lay people instead of experienced lawyers. Although there are charities, law clinics and pro bono services (e.g. the charity 1 3 Support through Court) which support lay people with filling in court forms and provide preliminary advice, their services are limited in the geographical reach, number of clients they can support and type of support they can offer [47]. Predictably, without tailored support lay people struggle to identify legally relevant information [80] and frame the narratives in a coherent way following the discursive and genre-specific conventions [74]. The narratives presented (in response to open questions) may contain irrelevant details, emotional accounts or unsupported claims [16,76], which makes it difficult for the judiciary to discern legally coherent arguments.
The second development which impacts on how court forms are being administered, completed and used in court proceedings is currently taking place as a result of the HMCTS reform programme, which "aims to bring new technology and modern ways of working to the way justice is administered" [39]. The reform was launched in 2016 in order to enhance the accessibility and efficiency of the justice system. As part of the programme, HMCTS recognises that user experience is one of the decisive factors for accessing justice; the institution thus aims to design userfriendly court communication, including court forms. A careful consideration of the LIPs/SRL's perspective is crucial for the introduction of online routes to justice, which is one of the most forward-looking changes being implemented as part of the HMCTS programme; digital paths to justice incorporate opportunities to start proceedings online and attend video hearings where all the participants (including the judge) appear in a virtual courtroom instead of a physical building [38]. In order to enable court users to initiate proceedings online, HMCTS is gradually turning physical court forms into online forms which aim to elicit the key information from the parties when starting the proceedings. For instance, it is now possible to start and complete uncontested divorce proceedings online; this digital service was introduced in 2018 and has seen 58% of digital applications uptake with only up to 1% of applications being returned due to user error in comparison to 40% when using the old paper format [38: 16]. The online private family services started with the digitisation of the court form C100 in September 2020 and the service has so far reported 92% user satisfaction, though the user satisfaction measures the online user experience (webinar presentation at 4th HMCTS Annual Public User Event, November 2020) and cannot thus be linked to more substantial access to justice issues, such as the success rate of the applications or to what extent users felt they knew what they needed to do, where to find the relevant guidance or how to phrase their narratives in response to open questions.
The vision is that gradually more court users would be starting proceedings online and accessing justice via digital paths, but in the meantime both paper and online court forms are being used concurrently. There are several concerns raised by legal professionals about the management of online paths [55], with three of them being directly linked to language use: (1) the comparability of the user experience between the printable and the online versions, (2) digital exclusion of the most vulnerable in the society from the more time-efficient and up-to-date method of accessing justice and (3) the experience and effective engagement of self-represented litigants with court proceedings. Although HMCTS offers some assistance with digital 1 3 Court Forms as Part of Online Courts: Elicitation and… aspects of starting the proceedings online [36], many comprehension-related challenges and guidance limitations remain to be problematic in both versions [55].
The current developments within the private family law digital services present an opportunity for this study to provide a timely contribution to the reform programme. Despite the seemingly narrow focus of the article on two versions of one court form, the outcome of the case study is relevant not only for the design of the C100 court form, but also other court forms or legal discourse types across different jurisdictions in the UK or internationally as many countries experience an influx of self-represented litigants [57] alongside the shift towards digitisation of legal proceedings [3].

Court Forms as Part of Legal Discourse
The requirements posed on legal texts are often contradictory: written legal discourse needs to accurately cover the complexities of the metalinguistic reality it relates to and at the same time ensure clarity and unambiguity of expression, which leads to the tension between the need for under-specification and over-specification. While under-specification limits opportunities to express legislative intentions and complicates the interpretation process, over-specification reduces readability and accessibility [12]. Another tension is between the exclusive status of the legal profession, preserved through the use of arcane linguistic features which make legal texts less intelligible to the 'outsiders' [28: 162-169] and the urgent need to improve the comprehensibility of legal documents in order to enhance the engagement of the wider public with the justice system. Yet another level of complexity is introduced by the fact that legal texts are often created for multiple purposes and diverse audiences (e.g. a police interview is elicited for the purposes of investigation as well as potential court proceedings). The same texts are thus re-interpreted and reframed as they progress through the justice system (e.g. police interviews are recorded, transcribed and extracts read out loud in court as supportive evidence for witness/ defendant examination [34]). The study discussed here contextualises the impact of these tensions on the communicative function of court forms in the context of DIY Law and digitisation of legal proceedings. By adding to the body of literature on comprehensibility of legal discourse, the article argues that it is important to explore legal texts beyond readability and comprehensibility and reflect on the implications of audience design and elicitation strategies to enhance more effective participation of the lay parties with legal proceedings.
Legislative discourse and discourse types drawing on legislation (court forms) display a wide range of lexical and syntactic complexities: long sentences, nominalisations with embedded phrases/clauses, complex prepositional phrases, multinominal expressions, syntactic discontinuities [9: ch. 4, 11: 106, 13]. Since 1970s, the Plain Language movement and the affiliated organisations across the world (e.g. Clarity) have campaigned to reform legal language and improve intelligibility of institutional and bureaucratic discourse [1]. Perhaps one of the most successful and consistent implementations of plain language principles across the wide spectrum of official and legal settings has taken place in Sweden, where the government endorses the quality of public information provision and each government authority has a language expert appointed to work on official texts and research audience needs [33]. The Swedish model of language planning incorporates micro and macro textual elements and includes: corpus planning for compiling the guidance and a database of texts; status planning for endorsing language as a democratic tool; and speech and text planning for ensuring a high quality of public facing information [33]: 25].
But plain language principles are limited when it comes to establishing transparency in such linguistically, cognitively and procedurally complex settings as legal contexts. Assy (2011) argues that full comprehension of legal discourse by the wider audience is not a feasible aim because law reflects complex reality and thus cannot be simplified. For the very reason, Bhatia [9] suggests two versions for legislative documents, an easified version and a simplified version. While the easified version would merely enhance cognitive processing (by reducing information load, indicating legislative intention and illustrating legislative issues) and address the wider specialist audience, the simplified version would take an additional step and include more specific explanation for the benefit of the wider public. This dual approach can be useful for legislative documents, but because court forms are completed, or at least signed, by a wide pool of court users, enhancing their comprehensibility is a key aspect of access to justice.
Another complication with plain language principles lies in the fact that their focus tends to be on micro linguistic features related to lexical, grammatical and syntactic choices, and only some of the discursive characteristics are included in the guidelines (e.g. organisational structure). A more pronounced emphasis on coherence, linguistic framing, and different approaches to defining legal terminology is often not explored in detail. The discourse level approach to simplifying legal language thus requires a more in-depth exploration and testing [28]: 165]. Yet, within the field of language and the law, there have not been many experimental studies on the comprehensibility of written legal discourse. Yeung and Leung's [81]: 90] experimental study showed that language specific simplification strategies can have a positive impact on the readers' prospects to understand guidance documents in Chinese. The authors mainly incorporated lexico-grammatical simplifications and only limited discursive simplifications (e.g. the inclusion of a diagram) in the userfriendly version and the changes increased the average score of respondents' correct answers from 48 to 79%.
Within applied psychology there is a stronger tradition of experimental research into the comprehensibility of legal texts, including court forms, but the inclusion of discursive strategies is not always clearly defined or tested. The conclusions which can be drawn from the existing research in applied psychology is that following plain language principles in the design of court forms improves their comprehensibility [53]. In addition to the micro-linguistic features (such as the removal of archaic and obsolete expressions, reducing lexical and syntactic complexity, shortening sentences and using personal pronouns), what also makes a positive impact on comprehensibility are the following macro-linguistic aspects: ensuring that discourselevel features reflect spoken genres, clarifying the intention of the text and creating a coherent representation of the propositional content [49]. Simplifying the language, engaging the audience through the appropriate mode of interaction and visualising 1 3 Court Forms as Part of Online Courts: Elicitation and… or illustrating key concepts can improve the comprehension by up to 30% [46]. It is easier to deal with the propositional content which is expressed explicitly though because the combination of increased readability versions and the explanation of legal terms help the layperson apply newly acquired concepts to new situations and explain their reasoning 2 .
What is difficult to achieve is to create a clear representation of all aspects inherent in the propositional content because specialized legal meaning is often expressed implicitly without accompanying keywords signalling the key messages; it is the implicit meaning that often renders the text unintelligible for the lay participant [6]. The professional expertise in a specialised area cannot be supplemented by following plain language principles or even additional explanation of terminology [20]. Assy [5] maintains that understanding legal language and engaging with the legal system involves "the ability to identify the pertinent legal rules, principles, and doctrines, to recognize the relevant facts and classify them into the pertinent legal categories, and to engage in a particular type of interpretation and reasoning" (378), which cannot be expected from the layperson reading a plain version text. Masson and Waldorn [49] show that the improved comprehension of more readable texts does not necessarily translate into improved performance when participants in experimental settings complete activities related to narrative development, such as answering questions and paraphrasing. The discrepancy between the legal and lay knowledge schematas, alongside common misconceptions about law, often prevent lay people from fully comprehending legal texts and engaging with legal proceedings [49,54]. Common misconceptions seem to persevere even after the introduction of the relevant information [49].
What could help dispel common misconceptions about law is public legal education [28]: 199, 49] as it creates an interactive and reiterative process of a feedback loop (known in pedagogy as a cycle between completing a task, receiving feedback and repeating the task), thus strengthening the acquisition of new knowledge and skills [4]. It has been shown that repeat SRLs tend to perform better in court as they gain court experience and improve their understanding of court procedures [78]. To a certain degree, this is even reflected in SRLs' improved linguistic performance, such as the use of more effective questioning strategies and challenging third turns during cross-examination [75], and the development of an inter-language, or an inter-genre [77], as a coping strategy. An important part of public legal education and enhanced understanding of court procedures is the availability of clear procedural guidance throughout the court proceedings, especially at the very initial stages which involve completing court forms. The role of court forms is thus to facilitate legal-lay communication and information exchange, paying special attention to the provision of guidance and supporting discursive competence of court users.

Methodology and Data
The focus of the comparative analysis is on the downloadable pdf/paper version and the online version of the C100 court form used for applications "under Sect. 8 of the Children Act 1989 for a child arrangements, prohibited steps, specific issue order or to vary or discharge or ask permission to make a Sect. 8 order". 3 While the printable form includes the questions and guidance applicable to all the scenarios and multiple intertextual links to additional documents, the online version is a smart form which uses prior responses to lead to the relevant sections and guidance links.
Both versions elicit information about the nature of the application and provide guidance on how to fill in the form. The paper version includes the guidance in three ways: intertextual links to booklets, mainly CB1 ("Making an application-Children and the Family Courts") and CB7 ("Guide for separated parents: children and the family courts"); guidance embedded in between individual sections of the court form; and a one-and-a-half page guidance note at the end of the court form. This results in the duplication of some information as well as multiple sources being used without a pre-specified thematic distinction (CB1 and CB7 cover similar topics but different aspects of self-representation). In contrast, the online court form was reorganised to streamline much of the information from the leaflets CB1 and CB7 as well as include additional court forms C1A ("Allegations of harm and domestic violence") and C8 ("Apply to keep your contact details confidential from other parties in family proceedings"). The digitisation of the court form has therefore involved the

3
Court Forms as Part of Online Courts: Elicitation and… redesign of the application process as well as the reduction of the number of court forms to be completed (if there are additional concerns) and the number of guidance documents to be followed. Despite the differences between the two versions, both types of applications are offered as equal routes to justice; it is thus important to compare the two versions from the linguistic point of view and explore potential comprehension or discursive challenges to court users. The data analysis draws on corpus linguistics methods and discourse analysis. While the corpus linguistics enables a detailed reflection on lexical and grammatical characteristics of the texts [cf [52], the critical discourse analysis component facilitates the exploration of the texts as a tool for eliciting specific information    and narratives from court users in the family justice context. The all-encompassing nature of critical discourse analysis is well-suited for the exploration of texts in legal settings [63]: 437] as it provides a framework for investigating the discursive practices and their societal impact in institutional and ideological contexts [62]: 5]. This allows the study to contextulise lexico-grammatical characteristics within the wider societal and justice aspects of court forms: the lack of comprehensibility as an instrument for exclusion of non-professional users; the explanations provided as a tool for pre-determining the options court users have, elicitation strategies as a method for framing the relevance of court users' narratives.
In order to prepare both versions for the corpus-based analysis, it was necessary to (1) extract the non-textual elements from the printable court form C100 and the supplemental forms C1A and C8 and (2) collate the text from all the sections incorporated into individual paths of the online version. 4 Table 1 summarizes the key information about the size of the sub-corpora.
The online version is longer and includes approximately 20% more words due to the multiple paths embedded within the form. In practice, the court user may be faced with the shorter text in the online form as only the relevant paths would be shown to them. The number of types is almost identical in both versions due to the similar content. Interestingly, the sentences are on average longer in the online version, but this is mainly due to the attempt to provide more explanation and guidance (see below). The following sections present the quantitative and qualitative analyses of the two versions of the court form, exploring the lexical and syntactic complexity, elicitation processes and guidance embedded within the two media.

Reflections on Readability
The starting point of the analysis is a reflection on readability due to the focus of the paper on the use of court forms by self-represented litigants and the frequent use of readability measures by court administration and management for estimating to what degree court communication meets plain language criteria (e.g. the Flesch-Kincaid readability measure is routinely used by HMCTS-personal communication with a member of HMCTS staff, September 2019). Despite the many limitations of standard readability tests, they provide an indication of syntactic and lexical complexity of the texts and have therefore been used in research on comprehensibility in legal contexts [e.g. [40]. For instance, the Flesh-Kincard measure [27], one of the most established measures, relies on the average number of words per sentence to evaluate the sentence-level complexity and the average number of syllables per word to assess the word-level complexity. Other measures also rely on average sentence length, but the lexical complexity is measured slightly differently: ratio of words longer than two syllables for the Gunning Fog Index [32]; average number of letters per word for the Automated Readability Index [82]; number of words with at least six letters for the Laesbarhedsindex score [56]; number of words with at least 1 3 Court Forms as Part of Online Courts: Elicitation and… three syllables for SMOG [56]. Such measures are often insufficient as they do not consider discourse level characteristics or such grammatical characteristics as word classes, clauses or phrases [26]. Furthermore, such formal characteristics as the length of sentences or words can be difficult to link to their comprehensibility [40]: sometimes it is commonly used words with a distinct legal meaning or short words (e.g. conjunctions) which can create comprehensibility challenges for unrepresented litigants [80]. Similarly, longer sentences with clausal elaboration can provide more useful information and thus aid comprehension more efficiently than shorter sentences [14,40]. The results of the standard readability tests can nevertheless be helpful in comparing the readability of the two texts.
Even on the basic level of word and sentence length, both texts are extremely complex with the estimated postgraduate reading grade (Table 2). Yet, only a minority of unrepresented litigants have postgraduate qualifications; Lee and Tkacukova [43] report 12% of those surveyed in the Birmingham Civil Justice Centre had an undergraduate degree with almost two thirds (63%) of unrepresented litigants having either no formal qualification (25%) or a pre-A level qualification. In addition, the online version is seemingly more difficult than the paper version, but the difference in the scores between the online and paper versions is given by the longer sentences in the online form (Table 1) and the reliance of standard readability measures on sentence length; it is thus the qualitative analysis that is more reliable for comparative purposes (see the qualitative analysis below).
To evaluate the lexical complexity, the forms were explored using the New General Service List, which contains approximately 2800 most frequently used words from a sub-section of the Cambridge English Corpus (Browne 2014). The results showed that the online form contains a slightly higher proportion of commonly used words (93.4% of words would be considered high frequency words) than the paper version (90.4%). Although this does indicate that the predominant majority of the vocabulary in the forms would be considered as crucial even for L2 learners (for whom the list was originally designed), this does not account for the fact that lay people are often not aware of distinct meanings of legal homonyms [80] and that it is inferred meanings that make legal texts less comprehensible to the lay audience [6]. The corpus-based lexico-grammatical analysis and qualitative investigation of concordance lines in the following sections provide a more contextualised exploration of the linguistic complexity within the court forms.

Lexical Aspects of the Corpus-Based CDA Analysis
The corpus-based analysis draws on single word and multi-word keywords, frequency lists, collocations (Word Sketch function) and CQL searches to explore the differences between the two versions in detail. The single-word keywords (Table 3) and multi-word keywords (Table 4), extracted by comparing the two sub-corpora against each other, provide more insight into predominantly lexical but also some grammatical characteristics of the two texts.
The keywords from the paper form show characteristics of clunky grammatical constructions (e.g. plural forms child(ren), order(s) or conditional clauses (e.g. 'contact address (if known)'); unwieldly instructions for filling in the form (e.g. which box to tick); requirements set for supporting paperwork ('you must ensure that the family mediator completes and signs Sect. 4a'). The concordance lines for the keywords in the online version illustrate several types of changes introduced in that version. Firstly, these include strategies for engaging with the court user more directly through the pronoun 'we' to personalise the institutional responsibility of the court staff ('Why do we need this information and what will we do with it?') and manage the users' expectations by guiding them through the process ('We'll need to ask you some further questions about your concerns.'). Secondly, the online form also shows more leniency in terms of requirements: 'Add an approximate date if you're unsure'; 'If you do not know all of the details, you can still complete your application but it will take longer to process'; 'Briefly describe what happened and who was involved, if you feel able to'; 'This information will be treated sensitively and you'll have the opportunity to give further details later in the court proceedings if you wish'. Thirdly, the online form provides more explanation of procedures: 'Lawyer negotiation is suitable for people who prefer not to meet because their relationship is still difficult, or because there's a lack of trust'. More importantly, such explanations attempt to categorise specific types of situations, such as types of abuse ('Have you ever been financially abused?' with other adverbs related to abuse, such as sexually, physically, psychologically, occurring lower in the list). The prominence of the explanatory function of the online version is further evident in the list of multi-word keywords. Table 4 illustrates the significance of legal terminology in the paper version ('parental responsibility', 'non-court dispute resolution', 'parenting plan'). The concordance lines further show that legal terms are often used in complex The high number of intertextual references in the form requires the reader to have detailed procedural and legal knowledge. The previously long-standing tradition of legal aid in the family context means that in the past court users and the judiciary could rely on the expertise of legal professionals. The current average court user would, however, often not be able to obtain legal aid or afford to pay for legal representation [45]: 17]. This poses a financial and, inadvertently, a communication barrier for access to justice as the form requires an awareness of what the terms mean, or at least skills to find the relevant sections in the Children Act 1989; yet, full comprehension and the correct application of the legal concepts to the court user's circumstances is exactly what many litigants struggle with [73].
The list of keywords in the online version includes lexical items which are more closely linked to the circumstances of the case and focused around two semantic fields: court procedures (e.g. urgent hearing, lawyer negotiation) and types of abuse. While the paper form mainly refers to generic 'domestic violence/abuse' without providing much explanation, the online form differentiates between different types of abuse ('financial abuse', 'psychological abuse', 'physical abuse', 'emotional abuse', 'withholding money') and contextualises these with specific examples of behaviours which would fall under the individual categories of abuse (see the discussion below). The redesign of the online version of the court form thus contextualises the text around child arrangements and child welfare and, as a result, creates more opportunities to elicit wellbeing concerns.
Another positive change from the point of view of comprehensibility in the online version is the substitution of some terminology with everyday vocabulary (e.g. the shift away from 'prospective applicant' or 'other party' to 'other parent'; from 'the same or substantially the same dispute' to 'the same or a very similar dispute'). Similarly, there is a noticeable difference in how the same concepts are framed (e.g. the move away from 'non-court dispute resolution' to 'collaborative law' or to 'negotiation' and 'reaching an agreement' as shown in Table 3).The lexical simplification links to the procedural simplification embedded within the digital paths to justice and the move towards mediation and negotiation [38].
Interestingly, the high occurrence of 'lawyer' in the online form (Table 3) shows that legal professionals are often mentioned in relation to the proceedings. The word occurs, for instance, in explanations of advantages and disadvantages of lawyer-led negotiations in relation to practical aspects (e.g. a quicker and less stressful but more expensive process) and communication aspects (e.g. no requirement to deal directly with the other parent, but being potentially less in control of the process). Although this shows an attempt to provide procedural advice and even incorporate communication-related aspects, such explanations are structured within the cognitive frame of legal representation as a starting point (see example on lawyer negotiation in relation to Table 3), rather than self-representation. Despite the high number of unrepresented litigants using private family courts to resolve their family disputes, the lack of information on self-representation in the redesigned online application illustrates that SRLs have an ambiguous institutional standing. This also means that SRLs are not provided with crucial procedural and practical information relevant to their role, e.g. how and why it is important to conduct negotiations with a lawyer representing the other party [74]. It is the court users' institutionally ambiguous standing, alongside the unawareness of the key aspects of the judicial decision-making process applicable to their case [73,80], that make SRLs feel suspicious about dealing with the other party's lawyer or nervous about how to proceed [43,78]. The mere presence of sufficient information on self-representation (and not only legal representation) could normalise and institutionalise the role of SRLs, who otherwise feel left out from the process [43], and support them to engage with the process more effectively [73]. Therefore, in addition to the positive developments already in place, such as lexical simplifications and semantic unifications of the online court form, HMCTS should consider the impact cognitive framing can have on court users and include more information on self-representation.

Grammatical Aspects of Corpus-Based CDA Analysis
In line with the findings in the previous section, this part shows that some grammatically complex constructions were also made more coherent in the online version. Lexical and grammatical changes can be illustrated by the prominence of the pronoun 'we' (see Table 3) and twice higher occurrence of pronouns 'you' and 'your' in the online version (413 occurrences, 4.39% of the sub-corpus) than in the paper version (189 occurrences, 2.2% of the sub-corpus). A slightly different simplification, the syntactic simplification, can be illustrated through the differences in the use of conjunctions (if, and, or, either, unless). Interestingly, despite the similar frequency (app. 5% for each sub-corpus), their use is different. While in the paper version conjunctions often occur in complex sentences with intertextual references to legal documents (see Example 1 in the section above), in the online version they occur in structurally simpler sentences related to court procedures rather than legally relevant exemptions and conditions ('You can ask for a hearing to take place sooner and a court will consider if there's good reason for the urgency'; 'As there are or have been safety concerns about the children, you do not have to attend a MIAM').
The lexico-grammatical simplifications can also be seen in nominalisations which include the word 'order' as these are often part of complex sentences (see Example 1). The word 'order' is among the top ten most frequent nouns in both versions: in the paper version, it is the second most frequent noun with the frequency of 70 (0.84%) and comes after 'child'; in the online version, it is the ninth most frequent noun with the frequency of 44 (0.47%), illustrating the lower occurrence of complex terms and nominalisations (it comes after 'child', 'court', 'application', 'abuse', 'violence', 'example', 'MAIM, 'detail'). There are two lexico-grammatical constructions in which the word occurs in the paper version: with pre-modifiers related to the Children Act 1989 ('Sect. 8 order'); pre-modifying nouns ('child arrangements order', 'consent order', 'contact order', 'special guardianship order', 'non-molestation order', 'other injunctive order'); and with a post-modifying -ing or -ed clause ('an order related to enforcement of a contact order', 'an order terminating parental responsibility', 'a court order binding a prospective party'). In the online form, the word predominantly occurs in the latter form, but in a more descriptive function explaining the context ('a court has made an order relating to you', 'any orders related to child abduction'). The reduced frequency and the explanatory function of nominalisations containing 'order' provides further evidence that the online version is lexically, grammatically and syntactically simpler.
Nominalisations and passive constructions tend to be stable features of legal and official documents [13]. The exploration of nominalisations and passive voice across the two forms leads to a more detailed analysis of the structural complexity (there are two sets of numbers for the online version: the number for the sub-corpus overall and the number which does not include features in the repeated text included in the multiple paths) ( Table 5).
The frequency of verb phrases in passive voice and nominalisations is lower in the online version; the context in which these features occur and their lexical composition also differ for both sub-corpora. In the paper form, the main verbs in passive constructions are lexically more complex ('at risk of being abducted', 'if consideration is sought within 48 h') and often occur in complex constructions (Example 2 below). In the online version, passive voice occurs in the headings for explanatory sections ('What evidence is accepted'), instructions ('Briefly describe what happened and who was involved, if you feel able to. This information will be treated sensitively') and definitions ('[The court can] decide who they [children] live with and when. This is known as a Child Arrangements Order'). The main verbs tend to be less advanced in the online form ('evidence that a prospective party has been arrested', 'this needs to be approved by a judge'), but some remain the same as in the paper version ('your leave was granted under paragraph 289B of the Immigration Rules', also see Example 2 below).
The online form also includes fewer nominalisations (only the occurrence of the N + -ing pattern is the same in both versions). Nominalisations in both formats include references to legal documents and orders (child arrangement order, child protection concerns, emergency protection order, Family Law Act, family court proceedings, forced marriage protection order, multi-agency risk assessment, violence protection notice) and references to support systems (violence support services), with the paper form incorporating more frequent references to procedural steps (supplemental information form). The below example illustrates a structurally complex extract (with passive voice and nominalisation) used in both court forms:

Example 2 (paper and online versions)
: an expert report produced as evidence in proceedings in the United Kingdom for the benefit of a court or tribunal confirming that a person with whom a prospective party is or was in a family relationship, was assessed as being, or at risk of being, a victim of domestic violence by that prospective party.
The extract comes from a longer section which lists types of evidence that can be submitted if the party wants to be exempt from the otherwise compulsory mediation. The complex noun phrase includes two post-modifying non-finite clauses after the head noun 'report' (the past participle clause 'produced as evidence in proceedings in the United Kingdom for the benefit of a court or tribunal' and present participle clause 'confirming that a person with whom a prospective party is or was in a family relationship, was assessed as being, or at risk of being, a victim of domestic violence by that prospective party'). Furthermore, there is another noun phrase embedded in the first one ('person' is post-modified by a relative clause 'with whom a prospective party is or was in a family relationship' and at the same time acts as a subject for the verb in passive voice 'was assessed' and complement 'as being, or at risk of being, a victim of domestic violence by that prospective party'). This creates a structure where a finite clause is embedded in a relative clause and the whole construction is then embedded into one of the post-modifiers for the head noun of the main noun phrase ('report').
Despite positive improvements and lexical and grammatical simplifications introduced in the online court form, there still remain specific challenges related to the overall comprehensibility of the substantive content of court forms, which creates an unnecessary barrier for an effective participation of unrepresented litigants in legal proceedings. The following section contextualises the changes introduced in the online form through a qualitative analysis of one section from both versions.

Qualitative Analysis
Drawing on the corpus-based quantitative overview above, this section focuses on the discourse analysis of the extracts related to risks of harm to children and parties (the paper version requires a supplemental court form C1A for this purpose). The identification of any form of domestic violence is essential for child arrangements proceedings because the paramount principle they are guided by is the well-being of children. The chosen extract serves the dual purpose: providing guidance on relevant concepts and eliciting responses in the form of short narratives (both versions are presented in Appendices 1 and 2). It is thus important to explore to what extent the two texts provide sufficient guidance, manage to define relevant concepts and support the court user with the development of the micro-narratives [cf [17]. As both versions present different attitudinal, cognitive and communicative challenges, they are explored separately below.

3
Court Forms as Part of Online Courts: Elicitation and…

The Paper Court Form
In line with the previously established grammatical and lexical complexity of the paper form, this sub-section explores further issues with definitions, explanations, coherence problems, difficulties with intertextuality, and limitations of elicitation strategies. The first question (see Appendix 1) alone demonstrates lexical, grammatical and syntactic complexity, which aims to cover all possibilities and includes lexical and grammatical features indicating generalisability and at the same time accuracy of description. This is accompanied by challenges related to textual coherence and procedural ambiguity (e.g. it is not clear if anyone who has had contact with children should be included even if they are not party to the case) and challenges related to intercultural and/or discursive competence (e.g. the verb 'alleging' implies suspicion and raises questions about the trustworthiness and credibility of the applicant, which may discourage reporting or identifying potential risks). Furthermore, the terms are listed without any attempt to explain their legal meaning (e.g. 'child abduction'). The definitional boundaries for these terms can be different in the legal context from their everyday use or, alternatively, the meaning could be misconstrued under the influence of the socio-economic background of the users (e.g., perceptions of domestic violence can differ among communities [e.g. [7]. Further complications can be caused by the potential semantic overlaps between some categories (e.g. 'any form of domestic abuse' and 'child abuse' or 'other safety or welfare concerns'). The principle of child welfare and welfare concerns is notoriously known as a concept which is often misunderstood due to cultural differences [23] or inexperience with legal proceedings by self-represented litigants [73].
The definitional shortcomings are further evident in the 'C1A Supplemental Information Form', to which users are directed in case they recognise alleged risks. It is in this document that the applicants would find definitions of terms relevant to domestic violence, but there are two limitations to the explanation offered here: (1) unclear relevance of the risks of harm (the first sentence mentions children and the second one uses a very generic reference to adults and 'family members', whereas in practice it needs to be about risks to children and the applicant); (2) the discrepancy between the terms defined as part of the relevant legislation, the Children Act 1989, and those the court user is required to identify in the subsequent table in Sect. 2 (the definitions for types of abuse are thus missing). This can result in the lack of clarity as to what instances of abuse to include or how to differentiate between different types of abuse (e.g. emotional and psychological). Given that the definitions and the types of abuse appear on the separate form, those completing the C100 form may not realise that, for instance, emotional or financial abuse could also be part of welfare concerns and may thus not proceed to the supplement form C1A.
The elicitation of specific information as well as open narratives is also flawed with problems. The elicitation of specific types of orders (see the question 'Have you had or do you currently have any of the following orders and are they current?' in Sects. 2 and 5) is not comprehensible to lay court users, who are more likely to be aware of the contents of the court order (i.e. the rules and limitations it sets) rather than the type of court order they have/had. In addition, some court orders may overlap: restraining orders are part of criminal proceedings and can effectively replace 1 3 non-molestation and occupation orders, which are part of family proceedings, but the differences are not explained in the form.
The elicitation of open narratives does not provide sufficient space to expand on the nature of behaviour (Sect. 2). Unlike experienced legal professionals, many selfrepresented litigants would typically struggle to clearly identify relevant issues and express them in a succinct way [16,76]. Ryan [56] reports on the appeal case in which a judge criticised the lack of space on the court form used for the transfer of sentenced prisoners in Ireland and determined that the design of the form was inadequate as it did not encourage the inclusion of sufficient information. In domestic violence cases, initial narratives help the judiciary determine the first intervention steps (e.g. whether the children would have contact with the alleged abuser and whether this contact would be supervised or unsupervised or in what capacity social workers would be involved) and in some cases even determine the urgency of the hearing. These decisions are often based solely on the information provided in the forms and can set the tone for the rest of the proceedings (e.g. pace with which contact with the non-resident parent would be expanding) and, more importantly, impact gatekeeping decisions (e.g. whether the case is of a more serious nature and should be heard by a district judge [e.g. [18]. Another issue with narrativisation appears in the Sect. 5 where the court user is expected to identify a court order they are applying for. The definitions of orders provide generic categories and it may thus be difficult for applicants to identify how they can proceed in advance of the hearings, i.e. before they know how the case may be developing or, what is even more important, before the direction hearing takes place (the direction hearing is usually the second hearing, or a later hearing, when the judge explains what would be important for the judicial decision-making process and how general principles would be applied in the circumstances of the case). Making the decision in the pre-court stages limits the micro and macro narrativisation options the court users could have at a later stage during the proceedings.
Overall, the paper form makes an attempt to explain some legal concepts, but there is no consistent strategy which would align the guidance with the elicitation process. And the lack of space for a narrative, alongside the lack of support for providing more clarity on judicial decision-making, means that unfortunately unrepresented litigants' applications may be misjudged and, anecdotally, this is possibly the reason many child arrangements cases with domestic violence issues proceed before magistrates, who are lay judges and are meant to deal with less serious cases, instead of being heard before district judges [18]. In addition, some SRLs may not even attempt to identify less apparent cases of domestic violence if they misunderstand the initial question in C100 and do not continue to the separate supplemental form C1A.

The Online Court Form
As shown in the corpus-based sections, the online form incorporates elements of procedural guidance and explanation as to why the courts need specific information ('what will we do with it' in Screen 1, Appendix 2) and what court procedures are important (e.g. information about Cafcass in Screen 1). Yet, there are definitional problems, coherence challenges and some issues with the elicitation of narratives throughout the form.
For instance, the first question about substance abuse (Screen 2) requires a short description of the nature of abuse (the length would depend on the applicant as there is no character limit set), but there is no narrative support provided as we saw in the paper version (e.g. information on when it started, the nature or frequency of abuse is not elicited) and the types of risks are not clearly defined. In fact, the definition of harm and ill treatment (Screen 1) is the same as in the paper version and thus equally complex, and possibly even less contextualised as there is no reference to the Children Act 1989. Furthermore, there are cohesion issues: Screen 2 breaks the cohesive continuity between Screens 1 and 3-4, the focus of which is on identification of concerns; the order of Screens 3 and 4 (and possibly even the fact that there are two of them) first elicit safety concerns for children and then move on to the risks for the children and the applicant. Although children experiencing domestic violence between parents/carers is part of safety concern for children (which justifies the order of content presented in the screens), the text fails to clearly establish that the following screens will be eliciting information about children and the applicant separately.
The rest of the screens elicit the information in a much more coherent manner, adopting the staged approach to first eliciting information about the children (Screens 4-16) and then the applicant (Screens [17][18][19][20][21][22][23][24][25][26][27][28][29]. The other positive aspect is that different types of abuse (in the Screens 6-17 and 19-30) are defined by exemplifying behaviours and categorising them. The use of exemplification to support the addresses' comprehension is well-documented in different communicative domains [83] and presents a crucial strategy for supporting self-represented litigants' comprehension of legal concepts. Nonetheless, there remain certain aspects which are not clearly defined: for instance, a common issue is that non-resident parents may stop paying child maintenance if the resident parent does create obstacles for contact with children [e.g. [64], yet it is not clear if withholding child maintenance is part of the financial abuse of the applicant (Screen 23) and there is no explanation or examples of financial abuse on children (Screen 10).
Similarly, it is worth noting that many of the explanations rely on prototypical examples, which certainly helps with the conceptualisation of the meaning through the identification of sets of features [69]. But including examples which would be considered outside of the semantic framing and legal conceptualisation of different types of abuse would be equally useful for avoiding misinterpretation and common misconceptions [49,54]. For instance, if someone regularly gives a child fast food, it does not constitute a safeguarding concern, unless the child has allergies or intolerances associated with that type of food (some parents may try to construe this as a safeguarding concern, but from the point of view of courts, this is a well-being concern only).
In terms of identifying and defining types of abuse in clearer terms, Screen 1 includes external links (which are then repeated in screens 3, 5, 18): the link to NSPCC for more information on abuse of children and the link to Refuge for more information on abuse of partners/applicants. This raises several issues: external links 1 3 are maintained externally and frame abuse as more general social problems; they do not provide information on the legal aspects relevant to abuse (e.g. what legal measures can be taken, what orders can be applied for, how evidence would be tested or even what type of evidence courts recognise); they focus on aspects which are potentially biased (the charity Refuge focuses on supporting women, which means that men following the link may be discouraged from making domestic violence allegations in the application) or not directly relevant (NSPCC covers all types of abuse, including in school settings, so applicants would need to go through a lot of text to identify relevant information). Although external links are undoubtedly useful for signposting to additional support, the court needs to provide the information clearly contextualised for family legal proceedings to ensure the information justice is served [cf 61].
Further shortcomings in relation to definitions and elicitation of narratives are related to, firstly, unnecessarily institutionalised discourse and, secondly, insufficient understanding of the situation many domestic violence victims typically find themselves in. For instance, in relation to the first issue, the distinction between the emotional and psychological types of abuse (Screens 12, 14, 25 and 27) is too technical and not necessary as there is no precedence as to which type of abuse would be more serious or whether the two should be treated differently (in fact, both NSPCC and Refuge only mention emotional abuse); eliciting a more general category is sufficient for the initial assessment and gatekeeping. In relation to the second issue, two of the four specific questions in Screen 7 (and subsequent ones identical to Screen 7) are about the reporting process. The nature of domestic violence is that it is often unseen and unreported [29] or it is reported mostly after the victim has left the perpetrator [24]. In either case, these questions may create a cognitive barrier and discourage applicants from including the details of specific instances. The mention of a GP as an example, may also be confusing: when domestic violence is reported, it is mostly reported to the police or agencies which can help with rehousing rather than GPs [24]; for some types of abuse (e.g. financial abuse) it may not be straightforward to find the most appropriate agency which could interfere without a court order. Despite these shortcomings, the positive development in the online form lies in the fact that there is no limit on the number of characters. Though, on the other hand, SRLs tend to provide extensive overly emotional narratives without a clear focus on the legally relevant aspects, which makes it time-consuming for judges to extract relevant aspects [16,76]. Therefore, any indication as to the amount of detail and the type of information required could help SRLs present the narratives in the most efficient way which would support the court in making preliminary gate-keeping decisions.
Overall, the online form provides certain advantages: more procedural information, clearer definitions of relevant concepts, and a step-by-step approach to eliciting different types of abuse. Yet, the form still presents potential challenges with the conceptualisation of some types of abuse, textual and structural cohesion, and insufficient support for eliciting an appropriate narrative. Given the progress already made as part of the digitisation process, some changes required would be fairly minor (e.g. organising Screens 1-4 more coherently), whereas other changes are linked to a reform of institutional ethos, such as the provision of in-built procedural guidance. The importance of procedural advice has been highlighted by Trinder et al. [78]: 83] who show that previous experience and understanding of court procedures has a positive impact on the capacity to self-represent. The preliminary findings of the Civil Money Claims Online beta version, for instance, also show that the majority of assistance required by the users is related to procedural information and advice [55]: 91]. Yet, there seems to be institutional resistance from HMCTS when it comes to embedding procedural advice into online proceedings, mainly because this could potentially be perceived as legal advice [55]. Embedding advice and information (rather than outsourcing it or relying on references to external links) would create a more sustainable support within court applications. The challenge is to balance (1) the need to elicit the narratives which can be difficult to share or which people may not feel confident enough to share due to misconceptions with (2) the need for ensuring the narrative elicitation discourages false allegations [51,79].

Reflections on Digital Paths to Justice and Conclusion
Court forms play a key role in determining the path the legal claim or grievance takes (e.g. whether the hearing is conducted in person or via a video link 5 or whether mediation is appropriate for the circumstances of the case), the pace of the proceedings, the management of evidence gathering and the planning of support measures (e.g. safeguarding measures, interpreter). The communicative purpose of court forms is thus two-fold: they elicit the relevant information from court users and at the same time need to support them in framing their claims in a way that makes it easy for the judiciary to extract the legal aspects of the case. Comprehensibility is therefore a key pre-requisite for court forms to function as a legal-lay communication tool. The other crucial components are clear elicitation strategy; structural and content support with the construction of relevant narratives in response to open questions; and coherent alignment between questions, procedural guidance and information provision.
The digitisation offers new opportunities of working with court forms and engaging with court users. One of the clear advantages is that smart forms make it easier to see the relevant information and answer respective questions, but this requires careful planning and alignment of the provision of information and procedural guidance with elicitation strategies. The article illustrates that the digitisation of the C100 form has introduced considerable improvements in relation to linguistic style and content: lexical and syntactic simplifications, more explanation of administrative steps, clarification of procedural or case management steps, and exemplification of some of the key legal concepts. These improvements illustrate that despite common scepticism as to the real impact of linguistic simplifications on comprehensibility of complex legal concepts [cf [5], it is possible to enhance court users' understanding of relevant principles and their active engagement with the development of narrativisation to support their case.
The article also shows that the current digitisation improvements are only partial. The main barrier for court users at the early stages of the proceedings remains to be linked to communicative challenges: some passages are carried over from the versions with grammatical and lexical complexities, persistent definitional issues, lack of support for court users with building their narratives. Furthermore, the disregard for the informational needs of lay court users means that the complex narrative development occurs in the discursive space where lay litigants are positioned as outsiders. Even just the fact that some definitions are outsourced via external links which do not provide contextualisation of the concepts for legal settings can create confusion for court users and, unfortunately, demonstrate how the functionality of digitised spaces (e.g. ease of establishing intertextual links) can be mismanaged. The digitisation of court forms or other aspects of legal proceedings needs to be accompanied by changes in the institutional ethos, recognising the role language plays in procedural justice and prioritising transparency and access to justice for lay court users.
As courts move to incorporate digitised elements and embed online applications and smart forms into legal proceedings, clear communicative practices gain on significance. Given that online tools enable court applications to be submitted remotely and at any time convenient to the user, lay court users need to be at the centre of not only digital design but also content design and conceptualisation of communicative practices. Clear explanations and definitions alongside supportive and explicit elicitation strategies should be part of institutional court communication in order to promote trust in government bodies among citizens and encourage compliance [33]. The efficiency of the legal system, the public trust in the justice system and access to procedural justice rely on communicative practices embedded in legal proceedings [31]. Drawing inspiration from the Swedish model of language planning across institutions [cf 33], there is a clearly defined function linguists can fulfil across jurisdictions and different types of legal cases: linguistic expertise is crucial for ensuring comprehensibility of court communication and supporting narrativisation and elicitation strategies.
Further research needs to investigate the applications submitted by court users in order to explore the actual effectiveness of online court forms in eliciting relevant information. The significant role language plays in online courts and remote justice settings cannot not be underestimated. In fact, it is not only the digitisation of legal proceedings, but also engaging the lay user with AI (e.g. for predicting the outcomes of legal proceedings) that relies heavily on the users' comprehension and clear elicitation strategies. There are thus many new and exciting research frontiers which are opening for applied linguists within the administration of court proceedings and management of the justice system.