This article aims to investigate the linguistic means of transphobic discrimination observed on the Internet. The languages analysed are English and Polish, since they both offer their speakers direct and indirect ways that discrimination manifests itself, yet Polish seems to enable a more noticeable means because of the presence of grammatical gender in the language. The paper discusses twelve samples of Computer-Mediated Communication, using the methodological tools offered by Critical Discourse Analysis and Queer Theory. Based on the analysis of the samples, the article shows that even though transphobia lies mainly in lexical choices of the speaker, it is not always direct, i.e. visible in insults and attacks on a trans person, but is oftentimes indirect, i.e. visible in the incorrect use of personal pronouns in both English and Polish, or in the incorrect use of grammatical gender in Polish. Moreover, while transphobia visible in language is not always intended by the speaker, it  can still be considered to be discriminatory. 

This paper presents a comparison between to-bearing relative clauses, adverbials and interrogatives on the one hand, vs. their to-less variants on the other, and discusses the functions associated with the presence of to. It is argued that at least three different instances of to should be distinguished. One converts relative clauses into appositive ones, which are necessarily semantically connected to the matrix clause and it makes the semantic connection override even apparent lack of appropriate syntactic connection. It attaches to relativizers, including gdzie ‘where’ and kiedy ‘when’ relative clauses. It is argued that the same segment is present in adverbials, triggering a factitive presupposition, as is the case of appositive relatives generally. The second to links the content of a kind relative, an adverbial or a wh-interrogative to previous contexts, possibly triggering a pragmatic presupposition. The third converts standard wh-interrogatives into either rhetorical or thetic questions. It is argued that while in the third instance we are dealing with a separate word and in the second with a clitic, the first to, hitherto unidentified or possibly falsely identified in relevant literature, appears to have both some characteristics of a clitic and of an affix.