Глубокое разногласие – это абнормальное расхождение во мнениях в споре. Рассмотрение подобного расхождения, развернувшегося в 2018 году в российских СМИ, при помощи методики, основанной на современных концепциях анализа аргументации, новой диалектике и логико-когнитивной теории аргументации, продемонстрировало потенциал цифровой визуализации, исключило интерпретацию решения как убедительного для сторон в содержательном смысле, но указало на компромиссный путь урегулирования разногласия как наиболее перспективный в условиях связи домогательств (харассмента) с дискриминацией, размытости социальных границ допустимости ухаживаний и отсутствия юридического определения домогательств как недопустимого поведения. We aim at demonstrating the potential of digital visualization of argumentation in searching and defining dispute resolution, and studying argumentation in the discussion (launched by a harassment conflict in 2018) with the help of the conceptions of the new dialectics and the logical-cognitive theory of argumentation. The digital visualization is done using the OVA software. The conflict and the discussion revealed legal, moral, and social aspects of the harassment problem in Russia, which affected the dispute resolution. At the first stage of the discussion analysis, the visualization allows discovering and reconstructing the arguments in relation to the parties’ divergence of opinions, and results in an argumentation map of the dispute, by means of which we establish the dispute outcomes at the second stage and determine the solution at the third stage. The advantage of the proposed method lies in the algorithm for determining the resolution of the dispute, to which digital visualization makes a significant contribution, acting as a convenient alternative of formalization. It allows identifying features of the argument, enhances the precision of argument evaluation by escaping from the risk of remaining indistinguishable in formulaic notation or flowcharts. The arguments resistant to the counterarguments form up the set of dispute outcomes, the subset of which convey the dispute resolution with respect to the type of dispute and the positions of the parties. The arguments are evaluated as sound or unsound by their replies to the critical questions formulated in relation to their structure, varying regarding the deductive, inductive, or plausible arguments. We reconstructed the discussion as two disputes about questions A: Did MP violate the norms of behavior by speaking or acting against the journalists? and B: Are actions like MP’s behavior harassments? We grouped the opinions of the participants in the discussion into four points of view: A1 – he did not violate, A2 – he violated, B3 – they are not, B4 – they are; identified three arguments in defense of each of A1, B3, and B4, four arguments in defense of A2; and visualized the parties’ positions and the outcomes of the disputes on three diagrams. The solution to the dispute A + B was the subset of four arguments that ensured the victory of A1 + B4: MP did not violate the norms of behavior for the lack of evidence of accusations, and this was not in his nature; harassments like MP’s actions are unacceptable, and since signs of courtships can be interpreted in different ways, accusations of indecent behavior must be brought and investigated immediately. The inconsistency in the dispute resolution in favor of A1 + B4 (MP did not violate the norms of behavior + such actions are harassments), convincing for the parties in the technical sense of the algorithm we employed, highlighted a deep disagreement between the parties about the admissibility of courtship. A deep disagreement is an abnormal divergence of opinions in a dispute, it ruled out the interpretation of the decision as convincing for the parties in a meaningful sense, but indicated a persuasive compromise way of resolving the disagreement.