Vicarious violence: conceptual evolution, empirical evidence, gender and children’s rights

1. Introduction

Using a child to threaten, control, punish or cause suffering to another person is a serious form of behaviour. It can occur in the context of domestic violence, coercive control, intense parental conflict, unjustified interference in the child’s relationship with one parent, psychological or physical maltreatment and, in the most severe cases, through the homicide of the child itself. The existence of such behaviour is not the central concern of this article. The problem begins when one seeks to determine whether the sexual configuration of the perpetrator and of the intended target should form part of the very scientific or legal definition of the behaviour.

The question has acquired importance because the contemporary concept of vicarious violence no longer circulates only in clinical, associational or political discourse. In 2026, Brazil introduced vicarious violence into the Maria da Penha Law and created the offence of vicaricide; Spain is debating an Organic Bill specifically devoted to vicarious violence; Mexico has federal legislation and recent constitutional case law on the matter; and Portugal retains the expression “violência vicariante” within the list of compulsory subjects for the continuous training of judges and magistrates exercising functions in the criminal, family and children’s courts (Brazil, 2026; Congreso de los Diputados, 2026; Lei n.º 2/2008, in its current wording; Suprema Corte de Justicia de la Nación [SCJN], 2026).

In this context, four planes that frequently intersect in public debate must be distinguished. The first is behavioural: what was concretely done to the child and to the other adult? The second is explanatory: which theory seeks to account for the occurrence and distribution of that conduct? The third is empirical: who engages in the conduct, against whom, with what frequency and gravity? The fourth is legal: which acts does the legislator decide to bring together under a particular category and with what consequences? Confusing these planes easily leads to circular reasoning. A legal category may be normatively legitimate without itself constituting a scientific finding about the nature of the behaviour it selects.

The analysis developed here starts from a position that rejects two opposing simplifications. On one hand, it would be inappropriate to ignore the historical inequality between women and men and the strongly asymmetric patterns of certain forms of male violence against women. A gender perspective is relevant to research, prevention, the organisation of services and risk assessment. On the other hand, a population-level asymmetry does not prove, without further investigation, that a given behaviour is exclusive to one sex by nature, nor does it allow the proof in the individual case to be replaced by the membership of the perpetrator or victim in a sexual category.

The aim is not to demonstrate the inverse thesis that women and men instrumentalise children in the same proportion, exercise violence with the same gravity or exhibit the same risk patterns. The data available do not support so general a conclusion. The aim is more demanding: to assess whether the definitions and methodologies used permit results different from those anticipated by the initial hypotheses, and whether the transposition of the concept into law preserves the autonomy of children’s rights.

The article accordingly addresses six questions. First, what meaning did violência vicariante carry in Portugal before the contemporary formulation associated with Sonia Vaccaro? Second, what conceptual shift occurs when the child ceases to be primarily the indirect victim of interparental violence and comes to be conceptualised as an intermediary through whom a woman is harmed? Third, which inferences are and are not warranted by the empirical bases currently in use? Fourth, how does the concept relate to the contemporary transformation of fatherhood, shared residence and divergences between feminist currents concerning sex, gender, motherhood and care? Fifth, which comparative legal models make it possible to protect victims of violence without turning sex into an answer given before the analysis of the concrete behaviour? Sixth, how can that analysis be translated into a legal-operational matrix that simultaneously preserves protection from violence, individual proof and the autonomous rights of the child?

2. Methodology and delimitation of the analysis

This work takes the form of a critical narrative review and a comparative legal documentary analysis. It is neither a systematic review, nor a meta-analysis, nor a prevalence study. No PRISMA protocol was therefore applied, nor is it the aim to estimate quantitatively the frequency of vicarious violence in the population. The methodological choice corresponds to the nature of the questions formulated, which demand conceptual reconstruction, evaluation of inferences, critical reading of selected studies and comparison of legislative and judicial options.

The documentary selection followed a hierarchy of sources. For empirical claims priority was given to peer-reviewed scientific articles, studies with an identifiable methodology and sample and official statistical publications. For the reconstruction of Sonia Vaccaro’s formulation, works and texts by the author herself were used, including the Spanish study published in 2021 and the 2023 book. For feminist theories, recourse was had to scholarly classifications, to texts by authors who identify with the currents analysed and to internal critiques of maternalist essentialism. For the comparative law, legal instruments, legislative bills, official reports, decisions of superior courts and official guidance were prioritised.

2.1. Search procedure, inclusion and exclusion

An iterative search procedure guided by the research questions was adopted, rather than a pre-registered systematic search strategy. Searches combined descriptors in Portuguese, Spanish and English, including “violência vicária”, “violência vicariante”, “vicarious violence”, “vicarious victimisation”, “revenge filicide”, “coercive control”, “parental alienation”, “shared residence” and equivalent legal terms. The initial identification of sources was supplemented by backward searching of cited references, forward searching of works that subsequently cited or developed central studies, and direct consultation of official legislative, parliamentary, judicial and institutional repositories.

Sources were included when they contributed directly to one of the article’s questions and allowed the proposition for which they were used to be verified. For empirical evidence, sufficient identification of design, sample or population was required; for conceptual reconstruction, priority was given to texts by the relevant author and to historically datable documents; for legal analysis, legislation, legislative bills, decisions and official documents were preferred. Duplicate materials, press reports where a primary source was available, comments without verifiable documentary support, and materials whose use would require generalisation beyond the design or population studied were excluded or relegated to contextual use only.

No uniform chronological range was imposed: genealogical reconstruction required examination of earlier uses of the term, whereas the comparative legal analysis was updated to 13 August 2026. This transparency does not transform the review into a systematic review, nor does it make the search fully reproducible in the sense of a closed bibliographic protocol; it does, however, make the selection criteria auditable and clarify the inferential status assigned to each type of source.

Sources of opinion or public intervention were treated as documents of the debate and not as empirical proof. Luís Aguiar-Conraria’s article on shared residence in Portugal, for instance, is relevant to document a contemporary political controversy but is not used to establish the motivations of feminist or conservative organisations. Similarly, activist texts or documents critical of vicarious violence are only drawn upon when their status is made explicit and when the claim in question can be confirmed by a primary source or is presented as the author’s own position.

The legal-documentary analysis was updated to 13 August 2026. This date is particularly important for Spain, where the Organic Bill on measures concerning vicarious violence is still in parliamentary passage and is not yet law in force. The Brazilian legislative amendments of April 2026 and the decision of the Mexican Supreme Court of 1 July 2026 were likewise verified.

The article further distinguishes between the so-called Parental Alienation Syndrome (PAS), understood as a diagnostic proposal historically associated with Richard Gardner, and parental alienation as a possible set of behaviours, relational processes or situations of unjustified rejection. Critique of the validity of a syndrome does not resolve, on its own, the empirical question of the existence of behaviours aimed at unjustifiably harming the child’s relationship with one parent. This distinction is important because Vaccaro’s work establishes an explicit link between the critique of PAS, shared residence and vicarious violence.

Finally, the analysis uses the term “vicarious violence” to designate the contemporary formulation that has gained currency in the Spanish and Latin-American sphere, and “violência vicariante” when it refers to the terminology used in earlier Portuguese documents. This choice allows the language of the sources to be preserved and makes the semantic shift — one of the central conclusions of the study — visible.

3. Before Vaccaro: indirect victimisation and violência vicariante

The claim that Sonia Vaccaro created the concept of vicarious violence in 2012 requires precise delimitation. Vaccaro and publications following her formulation attribute to her the formulation of the contemporary concept from that year onwards (Vaccaro, 2021, 2023). However, the general notion of vicarious victimisation predates that, and the expression “violência vicariante” was already in official use in Portugal five years earlier.

The international literature used, at least from the early 2000s, the expression “vicarious victimisation” to describe the exposure of children and adolescents to domestic violence. Baldry and Winkel (2003), in a study of 998 Italian adolescents, distinguished direct victimisation in the home from vicarious victimisation, defined as exposure to domestic violence. This usage did not depend on a man–child–woman relationship as a constitutive element of the category. The object was the experience of the person suffering the effects of violence directed at third parties.

In Portugal, Sani (2006) systematised research on the indirect victimisation of children in the family context, framing children exposed to interparental violence as victims of a reality that affects their development and well-being. The terminology used by Sani does not always coincide with the later legal expression, but it demonstrates that the scientific problem of child victimisation through exposure to interparental violence was already consolidated before 2012.

The clearest documentary proof appears in 2007. The III National Plan against Domestic Violence, approved by Council of Ministers Resolution no. 83/2007, established as its primary object the combating of violence directly exercised on women in intimate relationships and added the violence indirectly exercised on children who witnessed interparental violence, designating it, by reference to the doctrine, as “violência vicariante” (Presidência do Conselho de Ministros, 2007). The wording is particularly important because it combines a public policy explicitly sensitive to gender with the positioning of the child as an indirect victim.

Despacho no. 6378/2013, of the Ministry of Health, made the meaning still more explicit. The instrument created the Health Action on Gender, Violence and the Life Course and referred to exposure to violência vicariante as the indirect victimisation of children and young people in situations of domestic violence. The same document recognised the asymmetry of power between the sexes but warned that interpersonal violence takes many forms, with repercussions on the mental health of women and men, and established as a goal equity in health regardless of sex (Ministério da Saúde, 2013).

It was not possible to identify, in the Portuguese instruments consulted, the doctrinal source that gave rise to the expression used in 2007. The Council of Ministers Resolution itself limits itself to referring to what “the doctrine” designated as violência vicariante. It is therefore not methodologically sound to assert a direct line of transmission from any particular author or country. What can be affirmed is that the Portuguese terminology belongs to a prior international vocabulary of vicarious victimisation and exposure to domestic violence, documented by Baldry and Winkel (2003), among others, and that Portuguese public policy already used the term before the specific formulation later associated with Vaccaro.

The consequence of this reconstruction is significant. It cannot be sustained that Vaccaro created the generic notion of violent or vicarious victimisation. Her specific contribution lies in a different conceptual operation: the use of the term to define a form of violence against women exercised through children, siblings or other persons close to them. This change shifts the analytical centre and introduces sex and the perpetrator–victim relationship into the very concept.

The difference can be summarised as follows. In the earlier Portuguese sense, the causal chain is violence between adults, exposure of the child, harm to the child. In Vaccaro’s contemporary formulation, the defining chain is man, child or third party, woman, with the suffering caused to the woman as the element that organises the category. The two situations may coexist in the same case. A child used to reach the mother can be simultaneously a direct and indirect victim. The categories, however, are not equivalent and give rise to different empirical and legal questions.

The Portuguese expression retains current relevance. Article 74 of Lei no. 2/2008, in its current wording, includes “violência vicariante” among the compulsory subjects for the continuous training of judges and magistrates with functions in the criminal, family and children’s courts. The provision, however, does not define the term. This absence of definition makes it especially important to avoid a silent semantic shift, in which an expression used in Portugal with a meaning centred on child victimisation comes to be interpreted automatically according to a later, sexually exclusive foreign sense.

4. Sonia Vaccaro’s formulation and the limits of empirical inference

In Sonia Vaccaro’s work, vicarious violence is presented as violence against a woman carried out through an intermediary, very frequently the children. Her book “Violencia vicaria: Golpear donde más duele” (translatable as “Vicarious violence: striking where it hurts most”) organises this formulation within a theory of gender-based violence and patriarchal family, linking it to post-separation control, the critique of what the Spanish context calls “custodia compartida” — corresponding, in Portugal, to the debate on shared residence — and the opposition to PAS (Vaccaro, 2023).

The genealogy of this link predates the 2023 book. Vaccaro was already contesting PAS and its judicial use in a 2006 text. Later, a public intervention in 2018 was presented with a title that directly connected vicarious violence, the alleged PAS and “custodia compartida impuesta” (imposed shared custody). In the 2023 book, the author further claims to have concluded that PAS itself constitutes a form of vicarious violence. This sequence does not demonstrate that parental alienation, shared residence and post-separation violence are empirically the same phenomenon. It only shows that, in the author’s theoretical architecture, these debates are strongly interlinked.

The central methodological issue emerges with particular clarity in the 2021 Spanish study, presented as the first study of cases of vicarious violence culminating in the homicide of children. The research started from a base of approximately 400 judicial decisions and press reports and selected 51 child homicides classified as vicarious violence. Questionnaires, interviews with mothers and a discussion group with specialists in violence against women were also used (Vaccaro, 2021).

The study has descriptive value for characterising the cases that meet the criteria adopted. It can identify prior threats, separation contexts, history of violence, circumstances of the homicides and protection failures. Nothing in this analysis requires denying the gravity of these cases or the reality of men who kill or maltreat children to reach mothers. The difficulty arises when the sample is used to answer a question that its methodological design cannot test.

If vicarious violence is defined from the outset as violence whose final target is a woman, an investigation that includes only cases compatible with that definition will necessarily find women in that position. The result is true by construction but does not constitute an independent estimate of the sex distribution of the broader behaviour of using a child to cause suffering to another parent. To answer that question, the behaviour would have to be defined first without selecting sex, the corresponding cases identified and only then observation conducted of who engages in it, against whom, in which contexts and with what frequency.

The problem can be described as “classificatory circularity”. The definition establishes that the category contains a given sexual configuration; the data-collection systems record only cases that satisfy that configuration; the category’s statistics reproduce the definition; and those statistics may subsequently be invoked as proof that the behaviour exists only in that direction. The risk is not exclusive to vicarious violence. Any administrative or legal category can produce the same effect when one forgets that the data observed measure cases classified according to a rule, not necessarily all of the behaviour that would exist beyond it.

4.1. The mechanism of classificatory circularity

The circularity can be set out as a logical sequence. Each of the first three stages may be methodologically legitimate when the aim is to describe a category that has already been delimited. The problem arises only at the fourth stage, when statistics internal to the category are converted into independent evidence about the wider behavioural universe excluded by the inclusion criteria.

The question should be formulated in methodological terms and not through global labels about the scientific status of the concept. A concept may be politically situated and still identify real cases. A case study may produce useful knowledge without estimating prevalence. A legal category may protect a particularly exposed group without claiming to describe the whole of reality. Rigour requires only that one does not attribute to the research plane a capacity it does not possess.

This demand is consonant, in fact, with Vaccaro’s own critique of PAS. In discussing the use of controversial concepts in court, she drew attention to the need for validation, to the risk of uncritical reproduction of concepts among professionals and to the insufficiency of the argument from authority (Vaccaro, 2006). Applied in a consistent way, the same principle advises asking how vicarious violence is defined, what observation could disconfirm its hypotheses, which comparison groups are used and how far conclusions can be generalised.

The distinction between PAS and parental alienation is especially relevant at this point. Rejecting the notion of a diagnostic syndrome does not eliminate the possibility that a child is subjected to behaviours of devaluation, pressure, false information, relational interference or other practices aimed at unjustifiably harming a parental relationship. Likewise, recognising behaviours of parental alienation does not warrant the automatic conclusion that a rejection is unjustified, because the child’s refusal may result from violence, maltreatment, neglect, fear or actual deterioration of the relationship. This distinction was also developed in earlier institutional analysis by APIPDF, centred on the difference between an alleged syndrome and relational behaviours amenable to empirical assessment (Associação Portuguesa para a Igualdade Parental e Direitos dos Filhos [APIPDF], 2026). Technical work begins precisely where the use of labels ends: in discriminating between competing hypotheses.

5. Revenge filicide and the importance of defining the behaviour before sex

The literature on filicide offers a particularly useful methodological contrast because it contains research on the lethal form of a behaviour functionally close to vicarious violence without requiring, in the definition itself, any specific sexual configuration.

5.1. Three levels that should not be aggregated

To avoid excessive analogy, at least three levels of conduct should be separated. They may coexist in the same case, but they are not empirically equivalent: (a) homicide or lethal violence against the child; (b) threats, assault, exposure or other non-lethal instrumental harm directed at the child; and (c) procedural or relational use of children in dynamics of control, surveillance, pressure, obstruction or post-separation conflict. Evidence concerning sex distribution, risk factors or intention at one level cannot automatically be transferred to the others.

Myers et al. (2021) studied 62 cases of revenge filicide originating in nine countries. The authors defined the behaviour as the homicide of a child by a parental figure with the aim of causing emotional harm to the other parent. The series identified four main contexts: rejection, disputes over residence or contact, infidelity or jealousy, and adult-to-adult conflict. The perpetrators of the homicides were approximately as often men as women.

This result must be interpreted with great caution. A series of 62 cases gathered internationally does not constitute a representative epidemiological sample. It does not warrant estimating that men and women engage in revenge filicide in equal proportions in the population. Nor does it warrant concluding that the contexts, methods, gravity or risk factors are identical. What it demonstrates is more limited and, precisely for that reason, robust: the act of killing a child with the intention of causing suffering to the other parent does not, by its nature, require that the perpetrator be a man and the person reached be a woman.

The review by Freire and Figueiredo (2006) reinforces the need to differentiate filicide subtypes. The authors show that the literature includes neonaticide, homicides associated with severe mental disorder, fatal maltreatment, filicide-suicide, altruistic motivations and retaliation, among other patterns. The sex distribution varies with the definition, the age of the child, the clinical context and the population studied. To use global filicide figures to prove a thesis about vicarious violence would, on this account, be methodologically inappropriate.

The same caution applies to Portuguese data on violence in childhood. The joint publication of the National Statistics Institute and the Directorate-General for Justice Policy, based on the 2022 Survey on Safety in Public and Private Space, indicates that 12.2 per cent of people aged 18 to 74 reported emotional or physical abuse in childhood by the father or father figure and 11.8 per cent by the mother or mother figure. In child sexual violence, however, a markedly different distribution emerges, with greater prevalence among women as victims and a predominance of male perpetrators (Instituto Nacional de Estatística & Direção-Geral da Política de Justiça [INE & DGPJ], 2024).

These data do not measure vicarious violence and must not be presented as if they did. Their methodological utility is to show that the sex distribution varies according to the behaviour analysed. Data on intimate-partner violence do not allow automatic inference of the authorship of direct violence against children; data on the homicide of women in the context of gender-based violence do not allow estimation of the prevalence of the psychological instrumentalisation of children; and global filicide data do not allow determining how many homicides carried the specific intention of reaching the other parent.

The consequence is a simple rule for research: the variable whose distribution one wishes to know must be measured directly. When the object is the instrumentalisation of a child to cause suffering to another adult, the behavioural definition must be broad enough to allow the research to discover whether asymmetries exist, in which direction, with what magnitude and in which forms. An eventual strong male predominance would carry greater demonstrative value if it emerged from criteria that also allowed female cases to be identified rather than being built in as an inclusion requirement.

5.2. Behavioural possibility is not equivalent to population symmetry

This distinction is central to the article’s thesis. Identifying female perpetrators in a functionally equivalent behaviour shows that male sex is not a necessary condition of the conduct; it does not show that women and men engage in it with equal frequency, severity, persistence or lethality. Conversely, a male predominance demonstrated by comparable data may support gender-sensitive policies and risk assessment, but it does not turn a statistical predominance into the logical impossibility of other configurations. Demonstrating population symmetry or asymmetry requires comparative designs capable of observing the relevant groups under equivalent criteria.

6. Post-separation fatherhood, care and coercive control

The relationship between vicarious violence and post-separation fatherhood requires particular care, because the same apparent conduct may correspond to very different causal processes. A man who, after separation, seeks more time with his children may be reconstructing his fatherhood, adapting to a new practical division of care, responding to the fear of losing the relationship with the child or, in certain cases, using parental responsibilities to prolong a pattern of control over the former partner. Isolated observation of increased involvement does not distinguish these hypotheses.

Post-separation coercive control is a documented reality. The British Home Office’s statutory guidance expressly identifies the use of child-related and maintenance regimes to control the victim, threats of taking away or harming the children, intentional frustration of contact, the use of children to monitor the other adult and the instrumentalisation of children in family law proceedings (Home Office, 2023). The same guidance stresses that coercive control may persist or intensify after the end of the relationship.

The existence of these behaviours does not authorise, however, the logical inversion whereby any paternal claim to greater participation constitutes sufficient evidence of abuse. Longitudinal and comparative research on fatherhood shows that separation does indeed reorganise care arrangements.

Haux and Platt (2021), using representative UK data on 2,107 fathers, found an association between pre-separation involvement and subsequent contact. The association, however, was only moderate. Prior history helps explain subsequent situation but does not determine it. This is compatible with trajectories of fathers who were little involved before separation and remain in low contact, but also with trajectories of greater subsequent involvement.

Koster and Castro-Martín (2021), in a study of 1,592 fathers in the Netherlands, found significant differences according to residential regime. Fathers with shared residence and, above all, fathers with the children’s main residence showed high levels of regular participation in care, whereas non-resident fathers showed lower involvement. The authors explicitly discuss the possibility that the reorganisation of family roles after separation increases the involvement of some men.

In the Portuguese context, Marinho (2023) identifies in the news debate on shared residence the representation of the “reconstructed father” — the father who uses separation or divorce to modify his earlier way of exercising fatherhood and become closer and more caring. The concept is relevant because it shows that increased post-separation paternal involvement is not only an abstract hypothesis. It is a social representation observable in a period of transformation of family relations.

This transformation is also promoted by gender-equality policies themselves. The OECD links parental leave specifically directed to fathers with greater male participation in care and housework, greater closeness between fathers and children and potentially positive effects on women’s professional participation (OECD, 2025). The reduction of the economic and professional penalty of motherhood depends, in part, on an effective redistribution of the costs of care.

A tension must therefore be acknowledged. A public policy cannot encourage men to become more autonomous carers during the relationship and, simultaneously, treat as presumptively abusive the fact that a separated father seeks to expand or preserve care merely because a more unequal private division existed beforehand. Earlier inequality may be relevant to the assessment of family history, but it does not constitute automatic proof of subsequent motivation.

The solution is behavioural and evidentiary. To conclude for the existence of coercive control one must seek a pattern of conduct limiting the victim’s autonomy and producing serious effects, in light of the context and the body of evidence. To assess parental responsibilities the effective relationship with the child, the capacity for care, the history of violence, availability, the child’s concrete needs and the quality of co-parenting must be considered. The sex of the parent and the division of tasks existing during the relationship may enter into the analysis, but may not replace it.

7. Feminist theories, maternalism and political convergences

An analysis of vicarious violence that seeks to understand the centrality attributed to sex, motherhood and male domination cannot treat “feminism” as a homogeneous category. Feminist theories comprise distinct and, in many respects, contradictory traditions. The classification adopted here follows, for analytical purposes, Sousa (2015), who distinguishes liberal feminism, approaches of difference affirmation (within which she treats separately cultural feminism and radical feminism) and post-modern feminism. This taxonomy is an analytical tool and not an assertion that authors, organisations or actual movements belong purely and exclusively to a single current.

This caution is particularly important with respect to radical feminism. Mackay (2015), in a work written from within this tradition, characterises male violence against women as both a consequence and a mechanism for the reproduction of male supremacy and patriarchy. She explicitly rejects, however, biological determinism and the idea that men are naturally violent. In her formulation, violence is politically and socially produced, not an essence of the male sex.

The internal diversity of radical feminism is itself documented by Mackay (2022). Mackay identifies in that text as a radical feminist and, simultaneously, as queer butch/transmasc, and describes divergences reaching back to the second wave over the inclusion of trans women in feminist spaces and movements. Some radical feminists worked with trans women; others advocated spaces based on the female sex. This controversy is not central to vicarious violence, but is useful to show that radical feminism, biological essentialism and a single position on the category “woman” are not equivalent.

Three ideas that are sometimes improperly conflated must likewise be distinguished. Maternalist essentialism attributes to women a natural, intrinsic or qualitatively superior connection to motherhood and care. The political centrality of sex may, on the contrary, treat sex as a material category relevant to understanding oppressions without attributing innate psychological qualities to women. A structural theory of male domination seeks to explain inequalities and violence through social relations of power between sexual groups. An author may combine elements of these perspectives, but one does not necessarily follow from the others.

Critique of maternalist essentialism exists within feminist thinking itself. DiQuinzio (1993) analysed the exclusion problems produced when motherhood is treated as the expression of a universal feminine identity. Mezey and Pillard (2012), on the legal plane, criticised what they call the new maternalism and defended sex-neutrality in parenthood as a relevant gain for equality. In their analysis, returning a legal primacy to motherhood can simultaneously reinforce women’s overload and exclude men from the benefits and responsibilities of day-to-day care.

At this point it becomes possible to analyse a political convergence without confusing ideological genealogies. Traditional conservative maternalism may defend a privileged position for the mother because it considers motherhood and fatherhood to be distinct natural functions. An argument present in certain feminist currents or organisations may arrive at opposition to certain forms of shared residence through a completely different path — for example, understanding that legal equal treatment may supply violent men with new instruments of post-separation control. The premises are different, but the concrete solution may coincide.

The Portuguese debate on shared residence offers empirical material for observing these representations. Marinho (2023) analysed 51 journalistic pieces published during the legislative transformation that culminated in Lei no. 65/2020. Opponents mobilised predominantly representations associated with the traditional model of family and parenthood, including the child at risk, the dangerous or inadequate father and the mother at risk. Supporters mobilised more often egalitarian representations, such as the child as rights-holder, the caring or reconstructed father and the overburdened mother. Marinho also notes biases among supporters, including negative representations of motherhood and lower visibility of mothers in favour of co-parenting.

Aguiar-Conraria (2019), in an opinion article published during that debate, compared the resistance of Portuguese feminist organisations to a presumption of shared residence with historical arguments from North American conservatism against legal equality between men and women. The text does not prove empirically that the motivations are the same and must not be used for that end. Its value lies in documenting that the possibility of convergence between ideologically distinct camps was a question explicitly raised in Portuguese public debate.

Brazil offers a further example of convergence in legislative outcomes without ideological identity. Bill no. 2,812/2022, presented by deputies of the PSOL, and Bill no. 1,372/2023, presented by Senator Magno Malta, of the PL, both proposed the repeal of the Parental Alienation Law, despite originating in very different political camps (Câmara dos Deputados, 2022; Senado Federal, 2023). The relevant point is not to infer that the feminist left and the conservative right share the same theory of the family, but to acknowledge that legislative alliances or convergences may result from distinct justifications.

This distinction is crucial for the present analysis. The aim is not to identify “the responsible feminist current” for vicarious violence. The aim is to understand how certain theoretical matrices on male violence, structural domination, sexual difference, motherhood and political organisation influence the way a category is constructed and, above all, what happens when a structural theory is carried over into the definition of individual behaviour with effects in criminal and family law.

8. From theory to norm: Spain, Mexico, Brazil and Portugal

8.1. Spain: a sexually asymmetric legislative choice

The Spanish trajectory in 2026 shows with particular clarity that a sexually exclusive definition is not the only legally available option. The second draft of the Organic Bill on measures concerning vicarious violence envisaged a broader scope for the new criminal framework. In its report of 12 June 2026, the Consejo Fiscal positively valued the widening of possible passive subjects and expressly stated that vicarious violence, although it habitually occurs in the context of gender-based violence, may also arise in the context of domestic violence (Consejo Fiscal, 2026).

The Council of Ministers subsequently chose a different solution. On 14 July 2026, the Government approved the Organic Bill and defined vicarious violence as that which, with the aim of causing pain or suffering to women, is exercised by current or former husbands or male partners through children, descendants, minors, persons with disabilities or other family members and persons close to them. The Minister for Equality publicly explained the choice of the exclusive framing as gender-based violence, rejecting the more generic alternative (Gobierno de España, 2026).

The analytical relevance of this course is simple. The sexual exclusivity of the final definition does not follow from any technical impossibility of formulating the behaviour otherwise. A broader alternative was actually considered and received a favourable opinion on a specific aspect from the Consejo Fiscal. The final version therefore represents a legislative policy choice.

As of the close of the present research, Organic Bill no. 121/000106 had been presented to the Congreso de los Diputados on 17 July 2026, admitted for processing on 23 July and was before the Equality Committee in the phase for the tabling of amendments, with a deadline indicated for 17 September 2026 (Congreso de los Diputados, 2026). It must therefore be treated as a legislative bill and not as law in force.

The statistical production associated with the category requires the same caution. The Consejo Fiscal’s report referred to 68 children murdered in contexts of gender-based violence, of whom 40 were identified as victims of vicarious violence. In the political presentation of the bill, the Government used the figure of 68 as homicides by vicarious violence (Consejo Fiscal, 2026; Gobierno de España, 2026). The discrepancy may stem from updating, a change of criteria or a communicative choice, but it shows the importance of making explicit the criterion that allows movement from a broader category to a specific one. Without that transparency, the statistical debate risks conflating the reference universe with the subcategory.

8.2. Mexico: protection of women and the autonomy of children’s rights

Mexico adopted legislation on vicarious violence within the framework of the protection of women. In October 2025, the Suprema Corte validated the federal norms defining vicarious violence, finding justified the specific protection of women given historical and structural inequality, and recommended to Congress greater precision in the design of the aggravating circumstance (SCJN, 2025). Recent jurisprudential development also shows the limits that children’s rights may impose on the automatic consequences of the category.

In a 2024 decision concerning the legislation of the State of San Luis Potosí, the Supreme Court expressly acknowledged that children may also be used to cause harm to the father, although it understood that such cases could be protected through other provisions, namely those relating to family violence (SCJN, 2024). This position is conceptually relevant: the legal system may reserve the legal label of vicarious violence for a specific configuration and, simultaneously, recognise that the underlying behaviour may occur in other directions.

On 1 July 2026, in Amparo Directo en Revisión 2798/2025, the Pleno of the Suprema Corte reinforced the autonomous protection of children and adolescents. It determined that family authorities must ensure the effective participation of children in proceedings concerning residence and contact and create conditions for them to express their views freely, without pressure or interference. More significantly, it declared unconstitutional the automatic and definitive loss of patria potestad provided for in federal legislation as a consequence of vicarious violence, considering it a disproportionate sanction in light of the best interests of the child (SCJN, 2026).

The decision does not reject the Mexican category nor does it lay down any binding rule for other countries. It shows, however, that a measure designed to protect a woman victim of violence cannot turn children’s rights into a derived and automatic consequence. The child must be regarded as a person with rights of their own, and any definitive alteration of the parental relationship demands an autonomous and proportionate assessment.

Read together, the Mexican decisions of 2024 and 2026 separate two questions that are often conflated. The first is whether the legislature may reserve a specific legal category for the protection of women in the face of structural inequality: in 2024, the Supreme Court answered in the affirmative and added that the use of children to harm a father remains legally protected under other provisions. The second is whether, once vicarious violence against a woman has been established, effects on the relationship between the child and the parent may follow automatically: in 2026, the answer was negative, requiring autonomous analysis of the impact on the child, effective participation, integration of the evidence and proportionality. Taken together, the decisions do not neutralise gender; they delimit the level at which it operates and prevent specialised protection from absorbing the child’s own rights.

8.3. Brazil: vicarious violence and the creation of vicaricide

Lei no. 15,384, of 9 April 2026, amended the Maria da Penha Law, the Criminal Code and the Law on Heinous Crimes. It came to include vicarious violence among the forms of domestic and family violence against women and created the offence of vicaricide, punished with a term of imprisonment of 20 to 40 years (Brazil, 2026).

In its legal formulation, vicarious violence involves violence against a descendant, ascendant, dependent, stepchild, family member, person under guardianship or direct responsibility, or a member of a woman’s support network, with the aim of reaching her. Vicaricide punishes the homicide of certain persons with the specific purpose of causing suffering, punishment or control to the woman in a context of domestic and family violence.

The Brazilian choice is therefore unequivocally specific as to the sex of the person meant to be reached. If a woman kills a child with the demonstrated intention of causing suffering to the father, the homicide and the instrumentalisation of the child remain legally relevant but do not fulfil the offence designed to cause suffering to the woman. This can be a legitimate criminal policy choice, but it does not constitute empirical proof that the functional behaviour only occurs in that direction.

8.4. Portugal: an old term, an absent legal definition

Portugal does not at present have an autonomous criminal offence of vicarious violence or vicaricide. The expression “violência vicariante” exists, however, in public policy since 2007 and continues to form part of the compulsory training of judges and magistrates.

The Portuguese specificity lies precisely in the coexistence of an earlier documentary tradition, centred on the child as indirect victim, and the contemporary circulation of a Spanish sense that places the woman as final victim. Because Lei no. 2/2008 mentions the term without defining it, it is important that training, doctrine and case law expressly indicate which meaning they are using.

A conceptual change is not illegitimate because it is later. Legal and scientific concepts evolve. The problem would be to treat two materially different acceptions as if they constituted an evident semantic continuity, especially when the most recent sense may influence risk assessment, statistics, parental responsibilities and judicial decisions.

9. An alternative: a gender-sensitive approach, behavioural definition and children’s rights

The international comparison shows that there is not only a choice between absolute neutrality and a sexually exclusive definition. It is possible to construct a model that recognises gender asymmetries in violence, develops prevention and specialised services for women and, simultaneously, describes certain behaviours of domestic violence in a way applicable to any victim who suffers them.

Directive (EU) 2024/1385 offers a particularly clear architecture. It defines violence against women as gender-based violence directed against a woman or girl because she is a woman or girl, or that affects them disproportionately. It separately defines domestic violence as physical, sexual, psychological or economic violence occurring within the family, the household or between current or former spouses or partners. More importantly, the protection and support chapters extend to victims regardless of gender, and the definition of victim includes children who suffer harm from witnessing domestic violence (European Parliament & Council of the European Union, 2024).

The British model of coercive control is equally relevant. The statutory guidance describes concrete behaviours, including the use of children to control or coerce the victim, threats of removal or harm, frustration of contact, surveillance through the children, manipulation of family law proceedings and the use of issues relating to maintenance or residence to maintain control (Home Office, 2023). The definition of the conduct does not require that the perpetrator be a man and the victim a woman.

This does not make the British approach gender-indifferent. Domestic-abuse policy and research may recognise that certain patterns disproportionately affect women. The difference lies on the conceptual plane: population inequality informs prevention and risk, while the qualification of individual conduct begins with what the person did, the effect produced and the context in which the conduct occurred.

The Domestic Abuse Act 2021 further reinforces a principle particularly relevant to this article: a child who sees, hears or experiences the effects of domestic violence and is related to a perpetrator or victim is recognised as a victim in their own right (United Kingdom, 2021). This framing approaches, on this point, the older Portuguese sense of violência vicariante and the 2026 Mexican constitutional orientation.

A possible behavioural definition need not erase gender. It could recognise as relevant conduct the deliberate use of a child or another third party to control, punish, threaten or cause suffering to a person with whom there is or has been a family or intimate relationship. The research could then determine whether the conduct shows asymmetries by sex, which forms are most serious, in which contexts it appears, which victims are most exposed and which specific measures are required.

This structure carries an epistemic advantage: it allows the hypothesis of asymmetry to be tested rather than built into the definition. It also carries a legal advantage: it reduces the risk of excluding victims whose suffered conduct is functionally equivalent, without preventing the legislator from establishing aggravations, services or specific policies where there is sufficient constitutional and empirical foundation.

10. The child as autonomous rights-holder

The debate on vicarious violence tends to concentrate on the intention of the adult who seeks to reach another adult. That element can be decisive for understanding the conduct and, in certain legal systems, for fulfilling a legal type. Nevertheless, the focus on the adult victim cannot make the child disappear as a person directly affected.

A child may be a direct victim of physical, psychological or sexual aggression. They may be an indirect victim through exposure to violence between adults. They may be used as a messenger, source of information or instrument of surveillance. They may be pressured to take sides, subjected to one parent’s devaluation, or prevented from maintaining a safe and meaningful relationship. They may also legitimately refuse contact with a parent who has mistreated them or exposed them to violence. These situations call for different responses and cannot be resolved by a single causal narrative.

The Convention on the Rights of the Child establishes that the best interests of the child must be a primary consideration, protects the child against all forms of violence and recognises the child’s right to be heard in proceedings affecting them. It also recognises the common responsibilities of fathers and mothers in the education and development of their children (United Nations, 1989). The best interests of the child are not automatically identical to the interests of the mother or of the father.

The Mexican decision of July 2026 translates this autonomy in particularly clear terms: even in a vicarious-violence proceeding, the rights of children and adolescents must be analysed and protected autonomously, and the child’s participation must take place without pressure or interference (SCJN, 2026). This principle should inform any regime that attributes to vicarious violence consequences for the exercise of parental responsibilities.

The same logic requires distinguishing parental alienation from justified rejection. A child who rejects a parent is not necessarily being manipulated. There may be violence, neglect, fear, prolonged absence or relational deterioration. But it is equally not methodologically sound to exclude in advance the hypothesis of undue influence by the other parent. Professional assessment must compare competing explanations, gather information from multiple sources, hear the child in a manner appropriate to age and maturity and avoid letting any category replace the investigation of the facts.

This approach is also a safeguard against false positives and false negatives. A false positive of vicarious violence may lead to the wrongful attribution of abusive intent to legitimate parental conduct and produce disproportionate restrictions on the child’s relationship with one parent. A false negative may leave a woman and a child exposed to a real pattern of control and danger unprotected. Similarly, a false positive of parental alienation may discredit an allegation of abuse, while a false negative may allow the continuation of behaviours that unjustifiably destroy a parental relationship. The quality of the decision depends on the capacity to distinguish adequately between different situations, not on the abstract preference for one of the labels.

10.1. A four-question legal-operational matrix

The preceding conclusions make it possible to formulate a heuristic matrix intended to discipline legal and expert reasoning. It is not a validated psychometric instrument, a diagnosis or an automatic decision rule. Its function is to separate questions that require their own evidential basis before legal consequences are attached to a label.

11. Discussion

The analysis carried out allows five intermediate conclusions to be formulated. The first is conceptual. The expression “violência vicariante” did not begin with Sonia Vaccaro. Portuguese public policy used it in 2007 and, in 2013, expressly defined it as the indirect victimisation of children exposed to domestic violence. The international literature had also used the notion of vicarious victimisation for exposure to violence. Vaccaro’s specific contribution lies in the reformulation of the concept as violence against women through the children, not in the creation of the general idea of vicarious victimisation.

The second conclusion is methodological. An investigation that defines vicarious violence as male violence directed against women can study and adequately characterise the cases selected according to that definition, but cannot use the sex distribution of the sample itself to prove the exclusivity of the broader behaviour. The problem is not political but logical. The result was partly determined by the inclusion criterion.

The third conclusion stems from empirical comparison. Revenge filicide shows that the extreme behaviour of killing a child to cause suffering to the other parent can be engaged in by men and by women. This does not demonstrate population symmetry, nor does it eliminate the possibility that men show greater lethality or specific patterns in given contexts. It shows only that the male sex is not a necessary condition for the functional existence of the behaviour.

The fourth conclusion concerns the transformation of parenthood. Increased paternal involvement after separation is compatible with coercive control in some cases, but is also compatible with a normal reorganisation of responsibilities, with the reconstruction of paternal identity and with social policies that encourage male participation in care. Treating a change in involvement as proof of control would be to confuse a contextual factor with a demonstration of the conduct.

The fifth conclusion is comparative-legal. Spain and Brazil have moved towards specific sex-based models; Mexico has combined specific protection of women with a constitutional intervention reinforcing the autonomous rights of the child; Portugal retains an old expression that is not defined in current law; and the European Union and the United Kingdom show that it is possible to combine a gender perspective with categories of domestic violence and coercive control that do not exclude victims by sex.

These conclusions do not justify a neutrality indifferent to inequalities. Population statistics can and should guide prevention policy. If women are disproportionately victims of a given form of violence, it is legitimate to develop specialised services, gender-sensitive training, risk detection mechanisms and adequate resources. The error lies in transposing directly the prevalence of one group into the proof of an individual case, or in using statistics of one phenomenon to fill empirically the gaps of another.

The distinction between the frequency and the possibility of authorship is equally important. Showing that women may engage in a given behaviour does not show that they do so with the same frequency, gravity or impact as men. Conversely, showing a strong male predominance does not make female authorship logically impossible. Science should be able to measure both dimensions, and law should consciously decide when it intends to construct a general category and when it intends to create special protection based on a structural pattern.

The feminist discussion further confirms that there is no simple opposition between parental equality and feminism. Feminist currents decisively contributed to legal sex-neutrality, to women’s entry into the public sphere, to the critique of compulsory motherhood and to the redistribution of care. Other formulations emphasise the sexual structure of violence or the need for spaces and policies centred on women. Some may produce, in concrete matters, results similar to those of conservative-maternalist conceptions, although they start from different premises. Identifying that convergence is not equivalent to asserting ideological identity.

This tension is particularly visible in family law. Women’s economic and social emancipation also depends on men taking on more domestic work and care. Parental equality is therefore not only a male demand. It can form part of a gender-equality policy, provided it is accompanied by robust protection against violence and individual assessment where there is risk. A model that values paternal care during the relationship but presumes abusive intent when that care increases after separation contains a contradiction that must be resolved by evidence and not by stereotypes.

On the epistemic plane, the best safeguard is to formulate questions that may yield unexpected answers. Future studies on the instrumentalisation of children should define observable behaviours, include women and men, heterosexual and same-sex couples where relevant, distinguish physical, psychological, judicial, economic and relational violence, measure gravity and repetition, control for history of violence and analyse separately effects on the child and on the adult meant to be reached.

On the legal plane, proportionality requires particular caution when the consequences fall on the exercise of parental responsibilities. Deprivation or suspension of a parental relationship may be necessary to protect a child, but should not automatically follow from a category without assessment of the child’s concrete interest, save in the cases and conditions strictly provided by law and constitutionally proportionate. The Mexican experience of 2026 shows that this tension is not abstract.

Finally, statistical production must make classificatory criteria transparent. If a State decides that “vicarious violence” designates only cases in which a woman is the final victim, the statistics of that category should be interpreted as statistics of that legal definition. To study the broader behaviour of using children to control or punish another adult, a different database or a parallel categorisation that does not preliminarily exclude other configurations will be required.

12. Limitations and future lines of research

This article has limitations that should be made explicit. First, it is a critical narrative review, not a systematic review. The selection of sources was guided by the conceptual and legal questions and does not warrant a claim of exhaustive coverage of the international literature on vicarious violence, filicide or coercive control. Although the search, inclusion, exclusion and source-hierarchy criteria are now stated explicitly, the strategy was iterative and was not prospectively registered; another researcher could reproduce the decision rules and documentary checks, but not necessarily recover an identical corpus.

Second, empirical research directly devoted to the contemporary concept of vicarious violence is still relatively recent and, in many studies, the samples are deliberately made up of women who are victims of gender-based violence. These works are suited to studying the experience of that population but do not allow the prevalence of the behaviour to be compared across different sexual configurations.

Third, the legal concepts are not directly comparable across countries. Patria potestad, guarda, custody, parental responsibility and responsabilidades parentais are not perfectly equivalent institutions. The comparison undertaken is functional and limited to the dimensions required to discuss vicarious violence and the rights of the child.

Fourth, the reconstruction of the Portuguese terminology prior to 2012 makes it possible to demonstrate that the term did not begin with Vaccaro, but did not allow the concrete doctrinal source that led the 2007 Council of Ministers Resolution to use the expression “violência vicariante” to be identified with certainty. Future studies on the evolution of public policy and health may clarify that conceptual circulation.

Finally, the literature on revenge filicide is scarce and does not offer population estimates robust enough to compare men and women. The study by Myers et al. (2021) is important to demonstrate the possibility of the behaviour in both sexes, but not to establish prevalence. Especially useful would be multicentre studies with behaviourally defined criteria, systematic reviews of judicial decisions and administrative data allowing the coding of intention, relationship between adults, sex, age of children, history of violence and post-separation context.

13. Conclusion

Violence exercised on a child with the purpose of reaching another person is a serious reality that the law and public policy must not ignore. In many documented cases, men who have exercised violence against women have used sons and daughters as instruments of threat, punishment, control or revenge. Recognising those cases and improving the protection of women and children is an obligation compatible with scientific knowledge and with human rights.

Documentary and empirical analysis, however, does not support the idea that the general notion of vicarious violence or vicarious victimisation was created in 2012. Portugal officially used the expression “violência vicariante” in 2007 and, in 2013, defined it as the indirect victimisation of children exposed to domestic violence. Vaccaro’s contemporary formulation represents a relevant conceptual shift: the child comes to be defined chiefly as an intermediary through whom a man reaches a woman.

That reformulation may identify an important category of gender-based violence. What it cannot do, without further comparative research, is demonstrate that the instrumentalisation of children to reach another parent is, as a behaviour, exclusive to that sexual configuration. Studies selected from a definition that requires a final female victim characterise that category, but do not measure what was excluded from the definition.

The literature on revenge filicide, the research on post-separation fatherhood and the data on different forms of violence show why the question must remain empirically open. Recognising female perpetrators in certain behaviours does not prove symmetry. Recognising male predominance in others does not prove exclusivity. The distribution must be measured for each specific behaviour.

The comparative legal analysis also shows that a gender-sensitive perspective does not require a single normative architecture. Spain and Brazil have opted for sexually asymmetric categories; Mexico has introduced constitutional limits in the name of children’s rights; the European Union distinguishes violence against women from domestic violence and ensures protection regardless of gender; the United Kingdom describes the use of children as a possible form of coercive control through behavioural criteria.

For APIPDF, the point of equilibrium must rest on a clear distinction between structural analysis and individual decision. Gender may constitute an explanatory variable, a risk factor and the foundation of specific policies when the evidence justifies it. It should not, however, replace the assessment of conduct, evidence, risk and proportionality when concrete persons and the fundamental rights of the child are at stake.

The child is not merely the means through which a woman or a man is reached. The child is a person with rights of their own, who may be directly assaulted, psychologically instrumentalised or indirectly victimised. The best interests of the child demand that the child’s situation be assessed autonomously and that no theory about adults determine the legal response in advance.

In operational terms, legal classification should proceed through four separate steps: description of the observable conduct; demonstration of the nexus and purpose of harming another adult; integration of the violence context and relevant gender variables; and autonomous assessment of the child’s rights and of the proportionality of the measures. This sequence helps to avoid two symmetrical errors: using gender as a substitute for individual proof or, in the opposite direction, ignoring structural patterns of risk that may be highly relevant to the case.

Equality between women and men also demands a real redistribution of care. The transformation of fatherhood, parental leave directed at fathers and models of dual-employment with dual-care show that women’s emancipation equally depends on the transformation of men’s social role. Therefore, an equality policy must not crystallise in family law the same sexual roles it seeks to overcome in society.

Protecting better does not require rendering other victims invisible. It requires clear concepts, research able to disconfirm initial hypotheses, statistics transparent as to their criteria, proportionate legal responses and an approach in which children’s rights remain at the centre of the decision.