Fake news has been a buzz word for at least four years. It is timely, then, to consider whether linguistic research can be meaningfully carried out on fake news. The present article covers issues such as the definition of the term, the way in which fake news can be identified, and the visual component of fake news. Fake news as a term has been widely applied to a variety of phenomena, including conspiracy theories, propaganda, satire and advertising. In addition, it has been applied to different formats in which information is disseminated, including but not necessarily limited to, news articles and social media posts. In terms of how fake news is identified, several strategies are presented, such as evaluating the credibility of a source and checking whether a certain claim has been fact-checked by organisations dedicated to the endeavor. As to the visual aspect of fake news, this is exemplified by two instances where misleading information is communicated almost exclusively by visuals. The final part of the article attempts to tie in these various aspects of fake news and argue that linguistic analysis of fake news is practically impossible and that the attempt itself is highly unlikely to lead to meaningful results. This is because, first, the term fake news does not apply to a clearly identifiable discourse category. Second, it is a label that the researcher himself cannot apply and needs to rely on the work of others. Third, the linguistic of fake news when isolated, does not meaningfully contribute to the application of fake news.