Datos de investigación asociados al artículo: http://hdl.handle.net/10396/36111, Dataset del paper "Lexical Blending and Dehumanisation in Misogynistic Hate Speech: A Cognitive, Corpus‑Assisted Study of foid" en el que se han estudiado los mecanismos de deshumanización hacia las mujeres en el blend léxico "foid" a través de la combinación de la lingüística del corpus y de la lingüística cognitiva., El acceso a este dataset se ha restringido debido al contenido especialmente sensible del mismo.

Datos de investigación disponibles en: http://hdl.handle.net/10396/36049, Incel discourse constitutes a form of online hate speech against women, mainly located in the digital manosphere and specific forums. It is characterised by extremist misogyny and recurrent self-victimisation, whereby incels frame themselves as powerless and dominated by women. This study examines the relationship between incel language and emotion by analysing dehumanization strategies through a cognitive corpus-assisted approach. The Incelcor corpus, consisting of 229 posts retrieved from incel forums, was compiled for this purpose. The analysis focuses on the blend foid (female + android). Corpus-assisted discourse analysis was combined with cognitive linguistics to examine the behaviour of the blend when it acts as subject/object and the interaction with its modifiers. Besides, cognitive modelling has been addressed in order to explore the granular dimension of the term under study. With the linguistic evidence and instances, a usage-based dictionary entry has been proposed distinguishing the two folded dehumanization strategies identified in the blend. Linguistic evidence reflects that incels made use of victimization in their speech, and their perpetration of linguistic and symbolic violence lies on these mechanisms of dehumanization and victimization. Granularity played a key role in the study of this salient term as it allowed to zoom on the lexical and discursive levels, also capturing elaboration through the study of collocational patterns. There are limitations that should be acknowledged like the corpus size, the specialized dataset and the absence of experimental measurements that empirically linked language and emotions.