Libfixes are parts of words that share properties with both blends, compounds and affixes. They are deliberate formations, often with a jocular character, e.g. nerdalicious ‘delicious for nerds’, or scientainment ‘scientific entertainment’. These are not one-off formations – some libfixes have become very productive, as evidenced by high type frequency in a single corpus. Libfix constructions are particularly interesting for a network analysis for three reasons: they do not always have discrete morpheme boundaries, they feature a wide variety of bases (including phrases, as in give-me-a-break-o-meter), and they may be the source of back formations such as infotain. In this paper, we present a corpus-based analysis of eight English libfixes (cracy, fection, flation, gasm, licious, (o-)meter, tainment, and tastic), detailing their formal and semantic properties, as well as their differences and similarities. We argue that libfixes are most fruitfully analysed in a Bybeean network model, in which nodes are connected on the basis of phonological similarity, which allows for both fully compositional and non-compositional constructions to be linked without an exhaustive analysis into morphemes.
Towards the automatic generation of Arabic Lexical Recognition Tests using orthographic and phonological similarity maps
