From proper name to primary interjection: The case of gee!

This paper is concerned with interjections that have evolved from nouns. Looking at a subclass of interjections, so-called expletives like gee!, jeeze!, gosh!, crikey!, it will be shown how primary interjections regularly evolve from religious names (here Jesus, God, Christ) via secondary interjections (Jesus!, God!, Christ!). Four stages can be distinguished: I proper name Jesus II secondary interjection Jesus! III phonological modification of Jesus! IV primary interjection gee!The “standard” model for subjectification in grammaticalization, Traugott and Dasher’s Invited Inferencing Theory of Semantic Change (2002), which focuses on the speaker as initiator of change, suffices to explain the gradual process through which the secondary interjection Jesus! has evolved through the repeated use of Jesus in invocations, oaths and prayers. Moreover it is claimed that the syntactic position of Jesus in these uses — sentence-initial and separated from the rest of the sentence — played a crucial role in the emergence of the interjection. Invited inferencing, however, fails to explain the emergence of gee!. It will be shown with corpus data that gee! emerged rather abruptly. The meaning of gee! is different from the meaning of Jesus! from its very beginning, and — in contrast to the process through which the secondary interjection Jesus! emerged — there are no transitional stages attested where both the meaning of Jesus! and the meaning of gee! can be inferred. It will be argued that this change was hearer-based and that semantic stability and relationships between words are determined by aspects of phonological transparency. As a result of the clipping of Jesus! (speakers did not want to “take the name of the Lord in vain”), which yielded gee!, the semantic base of gee! became phonologically obscured. Accordingly the hearer had to infer the meaning of gee! from the context of its occurrence alone — consequently gee! differed in meaning from Jesus! straightaway. The OED will be used as a corpus for the diachronic study (cf. Fischer 1992 and Hoffmann 2004 for a discussion of how the OED can be used as a historical corpus). The synchronic use of the forms discussed will be illustrated with examples from the British National Corpus and from the GoogleGroups newsgroups (cf. Bergh 2005 for a discussion of whether and how the Internet can be used as corpus).