Abstract

The research looks at the ways the term ‘gaslighting’ is made, fought over, and then employed to give permission to offer guidance in relationship advice discussions on X – what used to be Twitter. Using Discursive Psychology and a four-part DP-Linguistic Construction Model, the work studies 30 discussion sets – the original posts and the answers to them – from October to December 2025. The study follows the regular groupings of how things are done, and how categorisation, position, proof and response are all used. From 450 replies which were marked – fifteen from each set – people use ‘gaslighting’ by saying straight out what it is and using set ways of explaining it; they justify it with quotes, images and what they remember; they make it more powerful with judgements of what is right and wrong, and turn it into telling-people-what-to-do advice using orders and should/ought to type of sentences. The results demonstrate that in this type of writing, ‘gaslighting’ isn’t a simple, factual description. Instead it’s a way of publicly holding people to account, and how strong that is relies on the way proof and position are put together in the discussion. The work also provides a clear method of marking – S1 to S10 – for looking at how accusations turn into advice in online chats.