LingLunch : Maria Copot

Maria Copot (LLF)
That icky feeling: the role of linguistic prescriptivism in speakers' experience of defectiveness

 Defectiveness manifests as speakers’ rejection of any form of a lexeme that could fill a given paradigmatic cell (Sims, 2015). Work on the causes of defectiveness has so far focused on system-internal factors (uncertainty, homophony, etc). While a number of conditions that correlate with being defective have been established, it has proven challenging to make deterministic predictions about where defectiveness is to be found in a system: for example, uncertainty about the correct form will sometimes lead to defectiveness while at other times to overabundance (Sims, 2006). Similarly, attempts to empirically identify defective forms by seeking words of unexpected low frequency in a corpus have not proven fruitful, since defective forms and lexemes don’t seem to have a unique frequency profile (Copot & Bonami, 2020) or unique semantic properties (Chuang et al. 2022) . One underexplored facet of defectiveness and its causes is the impact of extralinguistic factors, such as normative pressures: a societal incentive to “speak correctly” (or even a more explicit knowledge that certain word forms are to be prescriptively avoided in normatively correct speech) can be expected to play a filtering role in deciding which of the words with the necessary prerequisites will be treated as defectives.

We conduct an acceptability judgment experiment to establish whether defective words in French bear the hallmarks of grammatical taboos: are defective forms truly not generated or blocked by the grammar, or are they instead grammatical forms that are stigmatised in usage? We compare judgments for defectives to those for slang (licit words that are prescriptively proscribed) and to ungrammatical words. To tease out the influence of prescriptive pressure, participants are asked for two types of judgement: a normative judgment ("is this usage correct? ) and a possibility judgment ("is this usage possible?"). We find that defectives are rated much more variably than either of the two item conditions, and that this variability is both the result of variation between items and between speakers. This is consistent with the idea that defective forms may be grammatical but stigmatised: participants that are more prescriptively inclined will rate the same item as worse than participants that are less so, and some defective words will be more explicitly stigmatised than others - these factors combined are expected to lead to high variability in judgments. The narrative that defectiveness results from prescriptive pressures is corroborated by the correlation between lexeme frequency and judgement for defective words: the expected low ratings for defectives are more likely to occur when the lexeme is frequent and conversely, infrequent defective lexemes are more likely to be treated as perfectly acceptable. This suggests that speakers may be relying on explicit external negative evidence to know which forms are defective.

The study has implications for how we conceptualise defectiveness and how we choose to study it, and provides further clues on open questions such as the link between defectiveness and overabundance.