Detecting spelling errors in compound and pseudocompound words.

Abstract

Three experiments using a spelling error detection task investigated the extent to which morphemes and pseudomorphemes affect word processing. We compared the processing of transparent compound words (e.g., doorbell), pseudocompound words (e.g., carpet), and matched control words (e.g., tomato). In half of the compound and pseudocompound words, spelling errors were created by transposing adjacent letters and in half of the control words, errors were created by transposing letters at the same location as the matched compound or pseudocompound words. Correctly spelled compound words were more easily processed than matched control words, but this advantage was removed when letter transpositions were introduced at the morpheme boundary. In contrast, misspelled pseudocompound words showed a processing deficit relative to their matched control words when letter transpositions were introduced at the (pseudo)morpheme boundary. The results strongly suggest that morphological processing is attempted obligatorily when the orthography indicates that morphological structure is present. However, the outcomes of the morphological processing attempts are different for compounds and pseudocompounds, as might be expected, given that only the compounds have a morphological structure that matches the structure suggested by the orthography. The findings reflect 2 effects: an orthographic effect that is facilitatory and not sensitive to morphological structure of the whole word, and a morphemic effect that is facilitatory for compounds but inhibitory for pseudocompounds.

Semantic transparency is not invisibility: A computational model of perceptually-grounded conceptual combination in word processing

“…In fact, for semantically opaque compounds such as ladybird or windfall, these whole-word meanings differ dramatically from their compositionallyobtained meanings (for an overview on semantic transparency/opacity, see Schäfer, 2018). However, from a processing perspective, it still makes sense to immediately initiate a compositional process whenever a compound is encountered (see Chamberlain, Gagné, Spalding, & Lõo, 2019;Günther & Marelli, 2019a). Assuming that the main purpose of language is to convey meaning, language processing would be geared towards understanding the linguistic stimuli we encounter (Libben, 2014): Before the whole-word lexical entry has been accessed, one cannot know whether the compound is familiar or not -and thus, whether there even is such an entry (see El-Bialy, Gagné, & Spalding, 2013).…”

Carport and carpet: Effects of compound and pseudocompound word structures on typing.

“…These findings held even after controlling for semantic, orthographic, phonological, and syllabic factors. Furthermore, Chamberlain et al (2020) found that compounds and pseudocompounds exhibited different effects in a spelling error detection task; compounds showed a processing advantage relative to their matched controls whereas pseudocompounds did not show this facilitation. Also, letter transpositions at the morpheme boundary removed the processing advantage for compounds relative to their controls, but created a processing disadvantage for pseudocompounds (e.g., patnry was slower than its matched control), suggesting that there is some advantage to the presence of morphemes which is removed by the transposition of the boundary letters.…”

The role of semantic transparency in the processing of spoken compound words.

“…In contrast to opaque compounds, exposure to pseudocompounds resulted in significant inhibition of the target car. These findings suggest that the processing system establishes a morphological structure for all compounds, including opaque compounds and even pseudocompounds; however, this may lead to a processing cost when the established structure is incompatible with the true morphemic structure of the word (see also Chamberlain et al, 2020).…”