Please use social media carefully. You may have already participated in the execution of a “witch” without realising it. Every repost, comment, like, and even every scroll slightly pushes a person further towards being seen, discussed, and judged.

Using the interface of social media as its main form, Evil Woman Trends simulates a common contemporary phenomenon: the Digital Witch Hunt. This project explores the question: How are these “witches” manufactured within Chinese social media, and how are digital witch hunts and executions collectively carried out by “us”?

Since early modern Europe, witch hunting has referred to excessive moral judgement mainly directed at women and presented as justice. Today, these witch hunts happen on social media platforms where everyone can participate. Any woman can become a “witch”. She may be a celebrity or an ordinary person. Often, she has done nothing truly wrong, yet a facial expression, a relationship, a style of clothing, or even an emotion can become “evidence” against her.

The website contains posts in different visual forms, each representing a different execution of a “witch”. On social media, harm is created through small and repeated interactions. Scrolling, liking, and reposting may seem harmless, but algorithms amplify emotions, encourage collective attacks, and make them appear justified. Therefore, every action in this project affects the condition of the “witch”. Most interactions accelerate her collapse, while supportive actions are too weak to stop the process.

On digital platforms, witch hunts are no longer carried out only through obvious violence, but through the everyday interactions of countless ordinary users. When harm towards victims is broken down into lightweight actions that only require moving a finger, are we still able to realise that we ourselves may also be participating in this violence?

In contemporary mainland Chinese long-form TV series, “evil women” (恶女) have become increasingly popular with audiences, often attractions more attention than traditionally “angelic” female characters. Yet their portrayals appear to be repeatedly shaped through highly uniform narrative templates.

This project asks whether such representations still reduce the complexity and diversity of women as both characters and people, and whether they contribute to the formation of a new stereotype of the female figure.

Evil Woman Dictionary is published in volumes, each covering 3 years. For each period, 5 representative “evil women” characters with high cultural visibility and discussion are selected.

The dictionary compiles 100 personality trait words selected from Anderson’s (1968) study Likableness Ratings of 555 Personality-Trait Words, arranged from higher to lower likableness ratings. These 5 characters are then used as material examples, appearing under the trait words that best describe them. Through this structure, the Evil Woman Dictionary takes shape.