The French Ministry of National Education has announced that a questionnaire on sexual violence will be distributed to millions of students, ranging from preschool to high school. The questionnaire can be completed either anonymously or with the student's name, depending on their choice. It is presented as a systematic tool to identify cases of sexual violence. The goal is to encourage more children who have experienced such violence to speak out. However, collecting a child's testimony and ensuring it is genuinely heard are two different processes. According to the Independent Commission on Incest and Sexual Violence Against Children (CIIVISE), children do speak out, but adults do not always protect them. The issue is not only the silence of children but also how their statements are received by institutions and adults around them. A questionnaire may not change how adults perceive children as credible interlocutors. This issue of credibility is not unique to children. When adult women disclose sexual violence, their testimonies are often discredited. For children, the doubt is even more systematic: they are often suspected of lying about everything. Since 2022, research has shown that this testimony is caught in a double relationship of domination: the generational one between adults and children, and the institutional one between professionals and users. These relationships of domination are themselves fueled by gender, age, class, race, and representations related to disability. Doubt towards the collected testimony is not only an individual posture but also a method, established as a shared professional culture. However, while some doubts allow protective action by requiring verification, cross-checking, and securing a decision, other doubts paralyze protective action. These latter doubts revolve around three recurring disqualifying figures: the "lying child," the "child who makes mistakes," and attributing the narrative of violence to external causes. The first figure, the "lying child," has roots in the history of psychiatry and continues to influence professional routines in the form of "typical cases." For example, adolescent girls perceived as practicing a religion may be suspected of lying when they report rape, assuming they invented the accusation to hide a premarital sexual relationship they cannot assume. The second figure, the "child who makes mistakes," is attributed to supposed cognitive immaturity, allowing institutional expertise to correct the child's experience under the name of a truth that escapes the child. For example, young children who have not yet had vaginal or anal penetration may be assumed to be mistaken due to a lack of anatomical knowledge. The third consists of attributing the narrative of violence to external causes (such as parental separation or suggestions from the protective parent), which suspends the institution's trust in the child. Disability, whether affecting the victim or the aggressor, saturates these frameworks of interpretation and tends to be pathologized. On the side of the aggressor, the violence is reclassified as a simple "lack of boundaries," shifting the responsibility of prevention to the child, who is then required to develop their own ability to protect themselves from danger. On the side of the victim, the narrative is systematically viewed through the prism of their disability, which becomes a master status, erasing the narrative of the violence itself. These disqualifying figures operate both before and after the testimony is collected, in the way the testimony is subsequently interpreted. For the testimony to be "heard" and trigger protective measures, the child's statement must meet three conditions: place (speaking to the right person), time (speaking at the right moment), and posture. The child must often assess alone when and where to reveal violence (a time alone with a professional, a formal time of an audition, an individual space outside collective places, etc.). Institutions do not only listen to what is said but also observe how the child positions themselves vis-à-vis their aggressor and family system. Explicit rejection of the aggressor appears as a pivotal variable for the credibility granted to the child. As if a child who continues to love or does not explicitly reject the person harming them becomes less credible. A school questionnaire, regardless of its quality, cannot alone recreate the conditions of place, time, and posture necessary for the testimony to be received properly. It may initiate a revelation but does not guarantee the subsequent institutional journey of this testimony. This observation relates to a broader sociological fact: institutions hold a monopoly on the legitimate production of truth. They decide which testimonies are credible and which are not. However, institutions are familialist, meaning the family is an object to be protected in itself. They tend to prioritize conflict avoidance and maintaining family ties, sometimes at the direct expense of protecting minors in danger. The education ministry's questionnaire can, to a certain extent and under certain conditions, allow the formulation of a revelation: it thus partially meets the requirement of timing by establishing a specific time. However, a detection tool does not change how children's testimony is subsequently received. It does not alter the disqualifying figures that structure professional doubt, the familialist nature of institutions, or the tendency to treat each case individually rather than as a symptom of a generalized power relationship between adults and children. Philosopher Miranda Fricker proposed the concept of testimonial injustice to describe this phenomenon: the act of granting less credibility to someone's testimony due to their social status or identity, regardless of the content of what they say. The figures of the "lying child" or "child who makes mistakes" are concrete manifestations of this: they act as filters that reduce the credibility of the child before their testimony is even genuinely examined. This deficit of credibility resides in the reception, not the collection: a questionnaire does not change the judgment categories of professionals who will subsequently decide whether to believe what was said. However, this risk is not a fate, it points precisely to the areas where progress can be made.