Written Language & Literacy - Volume 28, Issue 1, 2025

English orthography has largely been left to evolve freely across its thousand-year history with little top-down design or reform. Yet modern English spelling has become standardized, even if its standardized spellings are sometimes internally inconsistent. The relative freedom with which English orthography has been allowed to evolve makes it an interesting test-case for exploring whether spelling standardization might arise through a process of positive frequency-dependent selection: To maximize the chance of being understood, it pays for a writer to adopt spelling variants that occur with higher frequency, leading to those forms becoming increasingly entrenched. To evaluate this possibility, we assembled a diachronic dataset of English spelling variation, SpellEng, which we make freely available. SpellEng reports on the frequency of 112,080 spelling variants (across 32,264 lemmata) over the entire history of written English. We use this dataset along with a model of frequency-dependent selection to estimate the extent to which English spelling standardization can be explained by such a process. Our findings suggest that the past 1000 years of English spelling have been characterized by both positive and negative frequency-dependent selection. We discuss how linguistic dynamics and external events have shaped these periods of stability and change.

Bird (1999a, 1999b) was one of the first linguists to draw attention to the similarities between the newly emerging Roman script orthographies of African languages, where tone is often under-represented, and Semitic orthographies, where most vowels are typically under-represented. This paper demonstrates that it is not only the orthographies but the linguistic functions of tone in African languages and vowels in both Modern Hebrew and Modern Standard Arabic that are similar; not just at a general level but also in their details. In seeking to illustrate the similarities between a wide range of African languages and the two Semitic languages, the paper catalogues pairs of heterophonic homographs as being either lexical (Section 3.1) or grammatical (Section 3.2) in nature, as well as cases of interaction (Section 3.3). The paper also acknowledges a number of lexical tone distinctions among African languages that have no parallels in the vowel structures of Hebrew and Arabic (Section 4). Given the striking similarities in how the linguistic functions of African tone and Semitic vowels are represented (or not) in the different writing systems, any researchers concerned with developing orthographies for previously unwritten tone languages, such as those in Africa, would do well to heed the extensive literature on Hebrew and Arabic literacy acquisition and to identify what lessons can be extracted when developing tone orthographies.

In this paper, we first summarize the results of studies that have investigated the acquisition, reading, and spelling of noun phrases (NPs) in German and French. Then, we present the assumptions made in those studies. In the main part of the paper, we provide a detailed comparison of German and French noun phrases. There are numerous differences that have never been discussed in this context before; for example, the behavior of articles in German and French and the consequences for graphematic representation, such as sentence-internal capitalization in German.

It has been proposed that morphological knowledge facilitates the integration of sublexical and lexical codes in both word reading and spelling and mediates the transition from reliance on sublexical cues to lexical cues during early literacy development (Bahr et al., 2020). However, the specific stage at which morphology begins to influence reading and spelling remains unclear, as does the question of whether there are differences across languages (Landerl et al., 2022). In the present study, we explore the extent to which morphology may explain, in addition to the already known predictors of phonemic awareness (PA) and rapid automatic naming (RAN), individual variations in word reading and spelling within two groups of Swedish-speaking children identified with word reading difficulties (n = 77) and typical reading (n = 59), respectively, in Grade 2. All children participated in standardized assessments of PA, RAN, morphological awareness (MA), vocabulary, word reading, and spelling at the beginning of Grade 2. Hierarchical regression analysis was used to investigate whether MA explained any additional variance within the dependent variables of word reading and spelling when controlling for word-reading group, PA, and RAN. As independent variables, word-reading group, PA, and RAN were included at the first step and MA was added at the second step. MA was found to have only a modest influence on the variances for word reading and spelling when controlling for group, PA and RAN. In conclusion, this study of beginning readers and spellers in Swedish, confirms previous findings that PA makes the more substantial contribution to the variation within early word reading and spelling, with MA contributing only modestly. These results support the view that phonics should provide the primary instructional foundations, with MA serving as a complementary component during the first years of literacy instruction.

Developmental models of reading and spelling skills are generally based on assessments of early literacy skills in young children. The literacy development of children is contingent on the concurrent development of linguistic knowledge, strategies, and functional skills. In contrast, developing literacy in adults is characterized by heterogeneous spoken proficiencies during the acquisition of L2 reading and writing skills. In the present study, we investigate reading fluency and spelling skills in adults with emergent L2 literacy skills within an L2 German language course that included literacy instruction. We confirm that the reading and spelling skills of learners are inter-related as mediated through a shared common knowledge base. The study participants could be divided into two recognisable clusters (low- and high-performing learners), despite being tested at approximately the same point during the course. Both lexical comprehension and lexical literacy reliably determined spelling performance, even though lexical literacy was less informative in terms of predicting reading skills and overall literacy profiles. We discuss these results within the framework of Kim’s (2020) Interaction Dynamic Literacy Model and conclude with some considerations concerning the practical utility of establishing learner profiles.