We are delighted to invite you to the upcoming WA Wednesday Lunch Talk, which will take place on November 12th at 13:15 in the Aula.
This session will feature two presentations:
- Dr Halszka Bąk: “The “norm” in norming databases of affective language”
- Prof. Ronald Kim: “The synchronic status of Sievers’s Law in Gothic”
When you sit down to resolve a research problem at the intersection of language and affect (or emotion), you will typically not create your own stimuli. Rather, you will reach for one of the apparently abundant databases of affective language with appropriate norms compiled for research purposes. However, how often do you ponder the construct validity behind the affective dimensions you rely on to select the words for the stimulus material for your study? The study I am working on and will be presenting here is a meta‑analysis which evaluates what we know about the nature of basic affective dimensions (valence, arousal, dominance) based on the most commonly used affective language norms. In it, I looked at the key parameters that determine construct validity behind the affective dimensions of language. This included factors like: languages existing norms come from, the role of translation in the compilation of these databases, the populations the norming studies’ participants come from, the way affective dimensions are constructed within study designs. The preliminary results show fundamentally inconsistent reporting of study and translation procedures and marked variability in the constructs of affect. Among the existing norms, Germanic languages and female college-age populations are over‑represented. Correlations between the norms of valence often reach ceiling values, but those for arousal and dominance range from low to moderate. Overall, the illusion of an abundance of affective language resources obscures serious problems with the norms many empirical studies rely on. The analysis concludes with recommendations for future directions in affective norming studies, focusing on a recommended standard of reporting in those studies.
Sievers’s Law in Germanic refers to the realization of yod + vowel sequences as nonvocalic after light syllables (*V̆CjV), but vocalic after heavy syllables (*V̄CijV, *VCCijV). The resulting allomorphy has left important traces in the nominal and verbal inflection of Old English and other older Germanic languages, but in Gothic, phonologically the most archaic Germanic language, SL is widely assumed to have still been a synchronic phonological process. Recent studies have analyzed SL in terms of syllable structure or “prosodic optimization”, setting up repair rules or constraints to generate the attested outcomes. However, it is argued from the existence of variant forms and widespread occurrence of superheavy syllables that SL was in fact no longer part of the synchronic grammar of Gothic, and that the relevant alternations were morphologically conditioned just as in the other Germanic languages. This finding not only contributes to our understanding of Gothic phonology, but highlights the importance of clearly distinguishing synchrony from diachrony and taking full account of variation, loanwords, and other often overlooked evidence in premodern languages.
We look forward to seeing you there!