The effect of orthography on the recognition of pronunciation variants.

Abstract

In conversational speech, it is very common for words' segments to be reduced or deleted. However, previous research has consistently shown that during spoken word recognition, listeners prefer words' canonical pronunciation over their reduced pronunciations (e.g., pretty pronounced [prɪti] vs. [prɪɾi]), even when the latter are far more frequent. This surprising effect violates most current accounts of spoken word recognition. The current study tests the possibility that words' orthography may be one factor driving the advantage for canonical pronunciations during spoken word recognition. Participants learned new words presented in their reduced pronunciation (e.g. [trɒti]), paired with one of three spelling possibilities: (1) no accompanying spelling, (2) a spelling consistent with the reduced pronunciation (a reduced spelling, e.g., "troddy"), or (3) a spelling consistent with the canonical pronunciation (a canonical spelling, e.g., "trotty"). When listeners were presented with the new words' canonical forms for the first time, they erroneously accepted them at a higher rate if the words had been learned with a canonical spelling. These results remained robust after a delay period of 48 hours, and after additional learning trials. Our findings suggest that orthography plays an important role in the recognition of spoken words and that it is a significant factor driving the canonical pronunciation advantage observed previously.

Free Allophonic Variation in Native and Second Language Spoken Word Recognition: The Case of the German Rhotic

“…These studies, which have once again focused on reduction processes entailing full and reduced variants, and the subsequent asymmetric matches with orthographic representations, have led to somewhat mixed conclusions. On the one hand, building on their results in a novel word learning paradigm assessing /t/ and /d/ tapping and /nt/ flapping, Charoy and Samuel (2020) claim that orthography is indeed the driving force behind the canonical-form advantage. By contrast, Bürki et al (2018) and Viebahn et al (2018) assessed the recognition of variant forms for schwa deletion in French and argue for a less prominent role of orthography in spoken word recognition, especially in comparison to the effects of variant frequency and input variability during learning.…”

“…This potential confound has brought about more recent research questioning to what extent the advantage for canonical forms may be driven by orthographic influences ( Bürki et al, 2018 ; Viebahn et al, 2018 ; Charoy and Samuel, 2020 ). These studies, which have once again focused on reduction processes entailing full and reduced variants, and the subsequent asymmetric matches with orthographic representations, have led to somewhat mixed conclusions.…”

“…In consequence, the converging results across different types of variation have led to hypothesizing that both citation forms (e.g., ce[nt]er) and variants from dominant language varieties are more accessible to listeners because of their prototypical status (i.e., the memory inequality hypothesis in Sumner et al, 2014 ). However, the advantage for canonical forms has also led to the question as to what extent this advantage may relate to orthography (e.g., Bürki et al, 2018 ; Viebahn et al, 2018 ; Charoy and Samuel, 2020 ), as canonical variants tend to match the orthographic representation better than other variants.…”

The role of phonology-to-orthography consistency in predicting the degree of pupil dilation induced in processing reduced and unreduced speech

“…Similar results were found in Ranbom and Connine (2007), where the more frequent pronunciation of reduced variant [ɾ] elicits slower response latencies than the less frequent pronunciation of careful spelling-like variant /nt/. While Pitt (2009) argues that the reason that /nt/ is easier to recognize than [ɾ] may be because /nt/ is perceptually more distinct than [ɾ], Ranbom and Connine (2007) and others argue that the reason that unreduced forms are easier to process than reduced counterparts could be due to the orthographic form of the words, specifically the consistent relationship between the unreduced pronunciation and its orthographic form (Charoy & Samuel, 2019; Racine et al, 2014; Viebahn et al, 2018). Crucially, these spontaneous speech studies suggest that the surface form (i.e., phonetic realization) of words could be substantially inconsistent with its spelling, meaning that the inconsistency between the way yesterday is pronounced and spelled should be greater for the reduced pronunciation [jɛʃeɪ] than for the unreduced one /jɛstɚdeɪ/, and the inconsistency could affect perception of spoken words.…”

The role of phonology-to-orthography consistency in predicting the degree of pupil dilation induced in processing reduced and unreduced speech

“…Similar results were found in Ranbom and Connine (2007) where the more frequent pronunciation of reduced variant [R] elicit slower response latencies than the less frequent pronunciation of careful spelling-like variant /nt/. While Pitt (2009) argues that the reason that /nt/ is easier to recognise than [R] may be because /nt/ is perceptually more distinct than [R], Ranbom and Connine (2007) and others argue that the reason that unreduced forms are easier to process than reduced counterparts could be due to the orthographic form of the words, specifically the consistent relationship between the unreduced pronunciation and its orthographic form (Charoy & Samuel, 2019;Racine, Bürki, & Spinelli, 2014;Viebahn, McQueen, Ernestus, Frauenfelder, & Bürki, 2018). Crucially, these studies suggest that the surface form (i.e., phonetic realization) of words could be substantially inconsistent with its spelling, meaning that the inconsistency between the way yesterday is pronounced and spelled should be greater for the reduced pronunciation [jESeI] than for the unreduced one /jEst@~deI/, and the inconsistency could affect perception of spoken words.…”