Language behaviour and cognition
This strand of the project was dedicated to studies on linguistic behaviour specifically and on human cognition in general. We studied behaviour in situations when people (volunteers) were asked to produce, choose or judge inflectional variants that might be challenging.
We asked people to produce word forms for verbs or nouns that are defective, and hence whose inflectional variants are somewhat problematic. Consider a situation in which you have to choose among different word forms for the verb "stride." Should it be, for example, "had stridden," "strode," "strided," or "stroden"?
Our linguistic knowledge seems to be incomplete, no matter how experienced we are as speakers of a language. Therefore, to make inflectional choices based on incomplete information, we rely on the same cognitive mechanism as we used to rely on when we were children acquiring our first language(s). The term "statistical preemption" used in the framework of Cognitive Linguistics can explain much about how we deal with incomplete information and uncertainty. When there is no ideal variant to fill a slot in an inflectional paradigm, we tend to apply the so-called "just good-enough strategy," which leads to one particular inflectional variant in the case of one speaker, but can lead to significant variation when we look across a population of speakers.
In our inflectional choices, we rely not only on our previous experience with language, but also on norms that we believe exist in our language. Norms are conventions that, in many instances, are encapsulated in recommendations that language authorities impose on us, the lay language users. We do not have to obey them; however, they may influence us, as we may deduce from them that there is a correct way to use language (in our instance to inflect words) and an incorrect way. Defective paradigms are interesting in this respect, as in many languages, there seem to be no norms on how to deal with them. Hence avoidance is one of the major behavioural responses in the case of words with defective paradigms.
When we encounter a defective paradigm cell, we tend to avoid many inflectional variants, since we feel there is no "correct" form that we can find. However, if we opt to produce a form, we may find that other speakers have settled on a different one. Therefore, variation of different inflectional candidates to fill a gap in a paradigm may result in what looks like an overabundance of inflectional forms. This evident similarity may be deceptive, as more classical examples of overabundance tend not to include this element of individual uncertainty: after all, any of the variants might be equally acceptable.
In managing both defectivity or overabundance, speakers often have to activate multiple inflectional variants in their minds, and consequently, they suppress all but one that they eventually choose. This typically happens within the timeframe of tens or maybe hundreds of milliseconds. However, speakers appear to be slower in producing or choosing variants to fill defective slots than other variants. This is not necessarily the case when we choose among overabundant variants that are often all acceptable. Therefore, incomplete linguistic information slows us down as we produce language.
Publications
Nikolaev, A., Bermel. N. (2023). Studying negative evidence in Finnish language corpora. Word Structure 16(2-3), 206-232.
Nikolaev, A., & Bermel, N. (2022). Explaining uncertainty and defectivity of inflectional paradigms. Cognitive Linguistics 33(3), 585–621.