Mirror letter priming is rightward-biased but not inhibitory: Little evidence for a mirror suppression mechanism in the recognition of mirror letters.

Abstract

Unlike other visual objects which are invariant to the left–right orientation, mirror letters (e.g., b and d) represent different object identities. Previous masked priming lexical decision studies have suggested that the identification of a mirror letter involves suppression of its mirror image counterpart reporting as evidence that a pseudoword prime containing the mirror letter counterpart slowed down the recognition of target word relative to a control prime containing an unrelated letter (e.g., ibea–idea > ilea–idea). Furthermore, it has been reported recently that this inhibitory mirror priming effect is sensitive to the distributional bias of left/right orientation in the Latin alphabet such that only the more dominant (frequent) right-facing mirror letter prime (e.g., b) produced interference. In the present study, we examined mirror letter priming with single letters and nonlexical letter strings with adult readers. In all experiments, relative to a visually dissimilar control letter prime, both the right-facing and left-facing mirror letter prime consistently facilitated, rather than slowed down the recognition of a target letter (e.g., b–d < w–d). Assessed against an identity prime, mirror primes showed a rightward bias, although it was small in magnitude and not always significant within an individual experiment. These results provide no support for a mirror suppression mechanism in the identification of mirror letters, and an alternative interpretation in terms of noisy perception is suggested.

Letters away from the looking glass: Developmental trajectory of mirrored and rotated letter processing within words

“…Thus, lexical influences will especially enhance the orientation cost for them. In adults, the effect of orientation transformations of letters occurs primarily at a prelexical stage that cascades into higher levels (Fernandes et al, 2022;Kinoshita & Liong, 2023;Pittrich & Schroeder, 2023), and hence, the orientation cost is especially larger for reversible letters when material or task promote the involvement of lexical dynamics (Brossette et al, 2022;Fernandes et al, 2022).…”

“…Thus, regarding mirror-image processing, letters must depart from other visual objects. In this vein, recent studies with fluent adult readers adopted the masked priming paradigm, 3 the gold standard used to investigate the early stages of orthographic processing, and consistently found that mirror invariance is not observed in letter processing in a fully-mature orthographic system (Brossette et al, 2022;Fernandes et al, 2022;Kinoshita & Liong, 2023;Pittrich & Schroeder, 2023). Fernandes et al (2022) capitalized on the fact that reversible letters of the Latin alphabet include mirror-image and plane-rotation pairs with the same 180 • difference (e.g., d-b, and d-p, respectively) to test the claim that mirror letter priming effects reflect the operation of mirror invariance.…”

“…Indeed, when letters were embedded in strings, only 5th-graders were able to perform above chance in a samedifferent matching task (e.g., vibil-vidil;erfos-ertos: Brault Foisy et al, 2017). Furthermore, successful performance in these tasks, especially on unmasked same-case letter pairs, does not necessarily depend on abstract orthographic representations (Kinoshita & Liong, 2023;Thompson, 2009), as performance can be based on visual shape, whereas cross-case tasks do require letter identification (i.e., abstract letter identities). Indeed, children are slower on cross-case than samecase letter pairs until 6th-grade, albeit this lag decreases along reading development (Bigsby, 1990;Reitsma, 1978).…”

Letters Directionality Effects on the Inhibition of Mirror Generalization in the Recognition of Objects: Evidence From the Negative Priming Paradigm

“…Indeed, conversely to Soares et al (2019) findings, Soares, Lages et al (2021 found that left-oriented non-reversible target words were recognized faster not only when preceded by mirror-image primes than by orthographic control primes (ϱenial-genial < zenial-genial), but also that the difference between identity primes and mirror-image primes was non-significant (e.g., genial-genial = ϱenial-genial) which was not the case for right-oriented non-reversible target words (casual-casual < rasual-casual). Facilitative rightward mirror letter effects were also observed by Kinoshita and Liong (2023) in a recent study using a go/no-go letter match task with single letters, two-to-three letters strings, and pseudowords, though only when mirror letter primes were compared with identity primes and not with orthographic control primes as in Perea et al (2011) and Soares et al (2019). Nevertheless, and regardless of the methodological options adopted in these studies, the important point to stress is that all of them suggest that the directionality (right vs. left) of the letters affects the way we recognize words irrespectively of the nature of the letters they contain (i.e., whether they have reversible or non-reversible letters) and it should not be disregarded when examining the cost of suppressing the mirror-generalization mechanism in reading.…”

“…Based on previous research (e.g., Ahr et al, 2016Ahr et al, , 2017Ahr et al, , 2018Borst et al, 2015;Foisy et al, 2017;Kinoshita & Liong, 2023;Soares et al, 2019), we anticipate that negative priming effects should be observed when identical pairs of animals presented in mirror positions (vs. non-mirror positions) were preceded by reversible right-oriented letter pairs (e.g., "p-q") but not by reversible left-oriented letter pairs ("q-p"). In addition, since Soares, Lages, et al (2021) have demonstrated that the visual word recognition system also has difficulties in inhibiting the mirror generalization mechanism for non-reversible letters that violate the right-orienting rule (i.e., for left-oriented non-reversible letters), it is also possible that traces of a negative priming effect could be observed when identical pairs of animals presented in mirror positions (vs. non-mirror positions) were preceded by non-reversible right-oriented letter pairs (e.g., "f-j") but non-reversible left-oriented letter pairs ("j-f").…”