Harassment Limitation Runs from Cessation of the Course of Conduct: s18B(2)(a) Applied to Long-Running Domestic Abuse

1) Introduction

X against Y (Court of Session) [2026] CSOH 54 (Outer House, Lord Young, 24 September 2025) is a civil action in which the pursuer (X) sought damages for a long course of domestic abuse characterised as harassment under section 8 of the Protection from Harassment Act 1997. The defender (Y), a party litigant at proof, denied the allegations.

The background included criminal proceedings in which the defender was convicted (and non-harassment orders imposed) in respect of certain assaults and threatening/abusive behaviour towards the pursuer, but acquitted of other charges. The civil proof ranged across a relationship spanning more than 20 years. The court had previously granted anonymity (Rules of Court 102 and related statutory reporting restrictions) and the opinion anonymised lay witnesses and locations.

The key issues were:

  • What conduct was proved and whether it amounted to a course of conduct constituting harassment under the 1997 Act.
  • Whether alleged rapes were proved on the civil standard.
  • The effect of the defender’s criminal convictions on the civil fact-finding (including s10 of the Law Reform (Miscellaneous Provisions) (Scotland) Act 1968).
  • Limitation: whether older incidents were time-barred or whether time ran from the end of the course of harassment (Prescription & Limitation (Scotland) Act 1973, s18B).
  • Quantum: solatium, interest, employability loss, and treatment costs.

Procedurally, the pursuer sought late amendment to add post-2019 incidents (2020–2024), which was refused due to lateness and prejudice to a party litigant defender.

2) Summary of the Judgment

Lord Young held that the defender had harassed the pursuer contrary to s8 of the Protection from Harassment Act 1997 by a course of conduct including physical assaults and sustained verbal abuse, continuing into 2018. The court:

  • Rejected the pursuer’s rape claim as not proved on the balance of probabilities.
  • Found proved multiple assaults (including choking/neck-grabbing incidents) and a pattern of humiliating/verbal abuse as part of a single campaign of coercive control.
  • Applied s10 of the 1968 Act so that the defender’s criminal convictions shifted the evidential burden in the civil case on the corresponding incidents.
  • Rejected the defender’s limitation argument, holding that for harassment the limitation period runs from cessation of the course of conduct (s18B(2)(a)), not from each incident.
  • Awarded £200,040 comprising solatium £65,000, interest on past solatium £58,240, loss of employability £75,000, and therapy costs £1,800, plus judicial interest from decree.

3) Analysis

A. Precedents Cited

Harassment as a course of conduct; “oppressive and unacceptable”

The court reaffirmed that harassment under the 1997 Act is assessed at the level of the course of conduct, not each discrete incident, relying on Marinello v City of Edinburgh Council 2011 SC 736 (para [12] cited). This matters acutely in long domestic relationships where abusive behaviour may be repetitive, diffuse, and difficult to date precisely.

Lord Young adopted English authority on the qualitative threshold: in Majrowski v Guy's and St Thomas's NHS Trust [2006] UKHL 34 Lord Nicholls’ boundary between everyday irritations and harassment was quoted, and the shorthand “oppressive and unacceptable” (as later used in Veakins v Kier Islington Ltd 2009 EWCA Civ 1288) was treated as a practical touchstone rather than an impermissible gloss. The opinion also noted Sheriff Braid’s application of a gravity test in Dickie v Flexcon Glenrothes Limited (unreported, 4 September 2009).

Civil standard of proof for grave allegations

The court reiterated orthodox Scots civil evidence principles: even where allegations resemble serious criminal conduct, the standard remains the balance of probabilities. The opinion drew together well-known formulations from:

  • Mullan v Anderson 1993 SLT 835 (careful and precise proof),
  • 1st Indian Cavalry Club Ltd v HM Customs & Excise 1998 SC 126 (cogency of evidence), and
  • C v G 2017 CSOH 5; 2017 CSIH 72 (careful scrutiny),

while emphasising that “careful scrutiny” should not silently elevate the standard of proof.

Effect of criminal conviction in later civil proceedings: s10 of the 1968 Act

For incidents that matched the defender’s criminal convictions, Lord Young applied s10 of the Law Reform (Miscellaneous Provisions) (Scotland) Act 1968 to shift the burden: once the pursuer showed alignment between the criminal libel and the civil allegation, it became for the defender to prove he did not commit the act.

The defender argued that the court required the “full judicial context” (trial transcript/sentencing remarks). Lord Young rejected that and relied on Towers v Flaws [2015] CSIH 97 (para [27]) as directing attention to comparison of the criminal charge and the civil pleadings. The practical consequence is significant: civil pursuers can deploy convictions efficiently, without satellite litigation over criminal transcripts, where the overlap is clear.

Corroboration concepts used as supporting logic

Although corroboration is not a civil requirement, the opinion used corroborative features (medical records, photographs, witness testimony, and convictions) to test reliability. It also referenced the “single complainer mutual corroboration” principle from HMA v Taylor 2019 JC 71 as an available route to corroborative support (had it been required) by cross-support between similar assaults. The reference functions as an evidential reassurance device: repeated similar assaults can mutually strengthen inference about occurrence and pattern.

Interest on solatium in historical abuse cases

On interest methodology, Lord Young was referred to JM v Fife Council [2008] CSIH 63 and F v Chalmers [2025] CSOH 23, where courts separated historic “initial abuse” suffering from later-life sequelae to avoid unfair interest accrual over decades. The court declined to apply that structuring here, distinguishing those institutional/childhood abuse cases by the shorter timescale and “homogenous” suffering.

Broad-brush employability loss

For employability loss, the court took a broad approach consistent with Blamire v South Cumbria HA [1993] PIQR Q1, appropriate where exact earnings evidence is limited but impairment is established.

B. Legal Reasoning

1. Pleading discipline, amendment control, and proof management

Lord Young enforced a clear boundary: findings would be tethered to averments on record. However, the opinion rejected an over-demanding requirement for date-specific pleading in long-running domestic harassment, noting that even criminal indictments often libel broad periods. This is a pragmatic approach: where the defender’s position is a wholesale denial, the prejudice from lack of precise dates is limited.

The refusal of late amendment to add 2020–2024 incidents illustrates judicial sensitivity to fairness where a defender is a party litigant: late expansion of the factual matrix risks procedural imbalance and adjournment pressure.

2. Evidence law: opinion, character, and unauthenticated recordings

The opinion contains three practical evidential rulings:

  • Improper “expert-like” opinion: the defender’s witness (a psychiatrist) was not permitted to offer quasi-clinical opinions about the pursuer’s mental health or “coaching” in criminal proceedings without acting as an independent expert with duties to the court.
  • Character evidence: the defender’s “good character” evidence was treated as collateral and inadmissible (citing Walker & Walker on Evidence). The decision reinforces that harassment/abuse fact-finding turns on proved acts, not generalised reputation.
  • Private recording: although an objection fell away procedurally, Lord Young made clear he would not rely on a recording/transcript not properly proved in evidence (provenance not agreed; maker not called). This is a caution against attempting to smuggle contested material into submissions.

3. Credibility findings driving the factual matrix

The court accepted the pursuer as credible and generally reliable, with allowance for trauma-related memory fragmentation. By contrast, the defender was found not wholly credible, particularly where he advanced serious allegations (e.g., that injury photographs were fabricated) without an evidential foundation or even putting the allegation squarely to the pursuer.

4. Rape not proved: evidential insufficiency, not moral disbelief

The opinion’s treatment of rape allegations is careful and legally structured:

  • The court accepted that sex occurred in circumstances the pursuer now experienced as non-consensual.
  • However, the court distinguished reluctant consent from passive non-consent, echoing the conceptual framework familiar from criminal jury directions (and aligning with “free agreement” language in the Sexual Offences (Scotland) Act 2009).
  • With no corroborative support and limited contextual exploration, the court could not determine on the balance of probabilities whether there was no consent and, additionally, could not exclude the possibility of the defender’s reasonable belief in consent as a constituent feature of “rape” as alleged.

Importantly, the rejection rested on the pursuer not discharging the burden on a legally complex inference (consent/reasonable belief), not on a finding that she was lying.

5. Assaults and verbal abuse proved: building a “course of conduct”

On assaults, the court relied on a cumulative evidential structure:

  • GP records recording assault/domestic violence injuries (1997, 2001) as contemporaneous support.
  • Photographs of bruising after the September 2006 incident, with rejection of manipulation allegations as unsupported.
  • De recenti accounts and witness observations (including the parties’ child “B” arriving shortly after the 2017 incident).
  • Criminal convictions as burden-shifting anchors via s10 of the 1968 Act for relevant charges (assault by neck-seizing/struggling and threatening/abusive behaviour).

Verbal abuse and humiliation—criticisms of appearance, derogatory remarks, and a specific cutting remark in 2018 witnessed by a relative staying in the home—were treated as part of the same campaign when their function was to hurt, humiliate, and undermine the pursuer’s autonomy.

The court also treated the defender’s 2008 conduct (moving out while the pursuer was abroad expecting childcare support) as not merely separation conduct, but as a control tactic aimed at maximising distress and reinforcing dominance when the pursuer asserted independence.

6. Drawing the line: not every criticism within a marriage is harassment

Lord Young explicitly cautioned against converting ordinary (even obnoxious) domestic criticism into a delict. Where criticisms concerned parenting/household/renovations, and the court lacked a basis to assess their merit, the opinion refused to treat them as harassment unless they bore the hallmark of oppression (especially when embedded in violence and coercive control). This boundary-setting is doctrinally important for s8 claims within intimate relationships.

7. Limitation: one start date for harassment—cessation of the course

The most pointed doctrinal clarification is the limitation analysis:

  • The defender argued incidents before 2 December 2017 were time-barred.
  • The court held this misconceived: s18B(2)(a) of the Prescription & Limitation (Scotland) Act 1973 fixes the start of the three-year period at the date when the course of conduct ceased.
  • Because harassment was found to have continued into 2018, proceedings raised in December 2020 were in time, and s19A discretion was unnecessary.

In practical terms, the decision underscores that litigants and advisers must plead and prove the continuity of the course, not attempt to slice the claim into incident-by-incident limitation arguments.

8. Quantum: moderate-severe PTSD/assault solatium; orthodox interest; Blamire-style employability

The court accepted expert psychological evidence diagnosing PTSD caused by the defender’s abuse, with a guardedly optimistic prognosis and anticipated therapy. Solatium was assessed by reference to the Judicial Studies Guidelines, viewed as falling within “moderately severe” PTSD and assault categories, producing £65,000 (80% past). Interest was calculated at 4% on past solatium from September 1997 to decree, with a refusal to adopt the split-interest methods used in far older abuse cases.

On employability, the court made a broad-brush award (£75,000) reflecting part-time counselling earnings capacity impaired by PTSD, with a discount for contingencies and potential residual earnings; and awarded modest therapy costs (£1,800).

C. Impact

The decision is likely to influence Scottish harassment litigation in three main ways:

  1. Limitation strategy for long-running domestic harassment: it reinforces that the limitation clock runs from cessation of the course (s18B(2)(a)), making proof of late-stage incidents (even if fewer) potentially decisive for timeliness.
  2. Use of criminal convictions in civil harassment claims: it confirms a practical route to s10 burden-shifting without requiring criminal transcripts, where the civil allegations align with the libel (Towers v Flaws [2015] CSIH 97 applied).
  3. Boundary discipline: while adopting a robust view of violence-plus-humiliation as harassment, the opinion guards against over-expansion by refusing to treat all domestic criticism as delictual harassment absent the “oppressive” quality.

More broadly, the case illustrates how s8 of the 1997 Act can operate as a civil framework for patterns of domestic abuse—particularly coercive control expressed through violence, intimidation, humiliation, and autonomy-suppressing tactics—even where criminal proceedings only partially capture the historical conduct.

4) Complex Concepts Simplified

“Course of conduct” (harassment)
Harassment is not one bad incident; it is a pattern involving at least two occasions. The court looks at the overall pattern and whether, taken together, it is oppressive and unacceptable.
“Oppressive and unacceptable” / gravity threshold
Courts distinguish harassment from ordinary conflict. Behaviour must be seriously wrongful in quality—beyond mere annoyance or ordinary relationship friction.
Balance of probabilities, even for grave allegations
Civil courts decide what is more likely than not. They scrutinise serious allegations carefully, but do not apply the criminal “beyond reasonable doubt” test.
s10 of the 1968 Act (effect of criminal conviction)
If a defender has been convicted of an offence and the civil claim concerns the same conduct, the civil court can treat the conviction as shifting the burden: the defender must then show the act did not occur (in the sense relevant to the civil case).
Limitation for harassment (s18B(2)(a))
The three-year time limit runs from when the harassment course ends, not from each individual incident. So an older pattern can still be actionable if it continued into the three-year window.
Solatium and interest
Solatium is compensation for pain, suffering, and loss of amenity. Interest is added to reflect the time value of money for past suffering.
Blamire approach
Where exact earnings loss is hard to calculate, the court may award a broad lump sum based on likely earning capacity and realistic contingencies.

5) Conclusion

[2026] CSOH 54 stands as a detailed application of the Protection from Harassment Act 1997 to a long-term domestic context, illustrating how a mixed pattern of assaults, intimidation, and humiliation can constitute a single harassing course of conduct. The court’s key doctrinal message is that limitation for harassment runs from cessation of the course (s18B(2)(a)), not incident-by-incident, and that criminal convictions can materially shape civil fact-finding through s10 of the 1968 Act without the need for criminal transcripts where overlap is clear.

At the same time, the opinion demonstrates judicial restraint: rape was not held proved on the available evidence, and ordinary domestic criticisms were not automatically treated as harassment absent an oppressive quality. The outcome—substantial damages for PTSD and long-term abuse—signals the civil courts’ willingness to provide meaningful redress where a sustained coercive pattern is proved to the civil standard.