Spelling errors on verb forms are notorious in Dutch. They even have their own name: dt-fouten (‘dt-errors’). The cluster dt refers to a verb form ending that triggers many errors, the d being a stem-final letter and the t the orthographic marking of an inflectional suffix. These errors are highly persistent and also produced by highly skilled writers (e.g., adults and older teenagers). Paradoxically, the spelling rules are easy to describe, as they reflect the application of two spelling principles: a phonological and a morphological principle. The former is the basis of the Dutch spelling system: Spell what you hear. Sometimes, however, this principle is overruled by the morphological principle. The latter guarantees that morphemes have a constant spelling across all forms in which they occur.

Together, these principles generate homophonous inflected verb forms, i.e., forms that are pronounced in the same way but spelled differently. For example, the 1st and 3rd person singular present tense forms vind (‘find’) and vindt (‘finds’) are both pronounced as [vɪnt] but have a different ending (⟨d⟩ vs. ⟨dt⟩). The former coincides with the stem vind and ends in ⟨d⟩ because a [d] is heard in the infinitive vinden (‘to find’). The ⟨d⟩ is pronounced as [t] due to final devoicing. The letter ⟨t⟩ is attached to the stem in the 2nd and 3rd person forms¹, as a [t] is heard after the stem in verb forms with the same grammatical function and a stem that does not end in the sound [t] (e.g., hij loopt, ‘he runs’). A second verb type yields homophones in the 3rd person singular present tense and the past participle: gebeurt (‘happens’) – gebeurd (‘happened’).

Even though the spelling rules are taught as early as the fourth year of primary school and errors are signaled throughout high school, many errors occur on these verb homophones (e.g., Assink, 1985; Sandra et al., 1999). This is surprising, as early teaching and repetition lead one to expect error-free performance. Consequently, the errors are laden with stigma: Anyone who makes them must lack intelligence or be lazy (Verhaert & Sandra, 2016). However, psycholinguistic research (e.g., Frisson & Sandra, 2002; Sandra et al., 1999; Sandra & Van Abbenyen, 2009) revealed three factors that account for the error persistence. First, since verb homophones account for only 5 to 10 % of all verb forms (see Verhaert et al., 2016), their (morpho-graphic) spelling rules, as described above, must be applied rarely. This low occurrence frequency prevents an automatization of the spelling rules. Hence, rule application requires working memory resources. Second, working memory overload may prevent that the spelling rule is applied in time. Third, if this happens, the more accessible higher-frequency form may be activated by an automatic retrieval process. Consequently, many spelling errors are made when the lower-frequency homophone is correct (Sandra et al., 1999; Sandra & Van Abbenyen, 2009). This effect of homophone intrusions has been called the homophone dominance effect (Sandra et al., 1999). It was established not only in experimental research but also in perception studies (e.g., Verhaert et al., 2016) and natural writing contexts (e.g., in social media, e.g., Schmitz et al., 2018; Surkyn et al., 2020).

The factors mentioned above explain why even good spellers still occasionally make these errors. However, not all spellers know the rules or are able to apply them. Despite their descriptive simplicity, they require grammatical awareness to identify, for instance, the subject of a sentence. In a recent study, Chamalaun and colleagues (2021) showed that adolescents who could identify the grammatical function of verb homophones spelled 51 % of these forms correctly, whereas those who had problems with grammatical analysis spelled only 10 % correctly. Hence, spellers also make errors on verb homophones because they fail to determine their grammatical function. At the same time, a proper grammatical knowledge did not guarantee correct verb spelling: Still 49 % verb forms were misspelled. Grammatical knowledge is not a license for correct rule application. Hence, cognitive limitations but also poor rule knowledge or poor rule application cause errors. Finally, adolescents’ spelling attitudes may have an impact on the errors. Some are likely to care more about errors than others and more meticulously monitor their spelling output.

In this study, we examine adolescents’ knowledge of Dutch verb spelling rules and their spelling attitudes. We will do so in relation to two social variables: gender (boys vs. girls) and educational track (general vs. technical education). The aim of our survey is to gain a better understanding of the social differences established in our previous research (Surkyn et al., 2019, 2020, 2021). In these studies, we examined the spelling of Dutch verb homophones in adolescents’ private, informal social media writing, more specifically, on Facebook Messenger and WhatsApp. Although all teenagers fall prey to the same type of errors, boys produce significantly more verb spelling errors in their chat posts than girls, and adolescents following a curriculum in technical education (which is more practice-oriented than a curriculum in general education) make significantly more errors than their peers in general education.

We argued that the gender effect is not likely due to a difference in rule knowledge, as there is no reason to assume that boys have less knowledge of the verb spelling rules than girls. Rather, boys and girls are more likely to have a different attitude towards standard spelling norms. Sociolinguistic research has reported that women are more concerned with “sociolinguistic norms that are overtly prescribed” (Labov, 2001, p. 293) and avoid stigmatized forms more than men (e.g., Tagliamonte, 2011, p. 32). Spelling errors on Dutch verb forms are actually highly stigmatized as a result of the clear-cut spelling rules outlined above. Hence, we concluded that the gender difference might be related to a higher error awareness and norm sensitivity in female chatters (see Surkyn et al., 2019, 2020, 2021).

Furthermore, we suggested that differences between educational tracks are most probably related to differences in rule knowledge. The curriculum in general education (GE) of Flemish secondary education has a strong theoretical orientation and prepares students for higher (university) education. The curriculum in Flemish technical education (TE) has both a practical and a theoretical orientation. Students in TE proceed to higher education (but generally much less to university than GE students) or they can start their professional career after graduation. Those in more theory-oriented tracks of GE are likely to have a better knowledge of verb spelling rules and better metalinguistic skills, as their official curriculum focuses more on reflection on linguistic concepts in general and spelling principles in particular than the curriculum for the more practice-oriented tracks of TE (VVKSO, 2014). Consequently, teachers in GE insist more on the importance of correct verb spelling, and in many cases students’ verb spelling errors (in, for example, their tests or exams) are sanctioned. This different educational focus on verb spelling might result in a more outspoken attitude towards error avoidance. Moreover, in view of their official school curriculum, GE students are likely to be exposed more frequently to (formal) standard writing (e.g., novels or proof-read texts) and thus also to correctly spelled verb forms, which might positively affect the quality of their orthographic representations (of, for example, verb forms) (The Lexical Quality Hypothesis, e.g., Perfetti, 2007; Perfetti & Hart, 2002). The quality of an orthographic representation refers to the accurateness with which a word’s graphemes are represented in the mental lexicon (e.g., the vowel sound in the word feel is spelled as ⟨ee⟩). Students in technical education, by contrast, might be subject to a proportionately higher exposure to misspelled verb forms. Indeed, taking into account their school curriculum, we may expect their experience with standard writing to be more limited and their relative exposure to (nonstandard) social media writing consequently to be higher.² In addition, teenagers from different educational tracks probably differ in terms of their exposure to Dutch. Previous research by Hilte et al. (2018) on the same social media corpus revealed that Dutch is more often not the home language of the students in more practice-oriented tracks than of those in more theory-oriented tracks.

By examining adolescents’ rule knowledge and spelling attitudes, we are able to test our earlier interpretations. Hence, we investigate (1) whether boys and girls indeed master the verb spelling rules equally well but have different spelling attitudes, and (2) whether theoretically oriented adolescents have a better knowledge of the verb spelling rules and, as a possible result, attach more importance to correct verb spelling. We will do so by conducting a two-part survey. In the first part, participants’ knowledge of Dutch verb spelling rules is probed by means of basic verb conjugation exercises. In a second part, we use ten attitude statements about correct verb spelling in different writing situations to gauge the spelling attitudes of the adolescents.

By means of a two-part survey, we wanted to investigate the role of Educational Track and Gender with respect to knowledge of the verb spelling rules (Part I) and spelling attitudes (Part II). In our previous studies on verb spelling errors by adolescents in private social media writing (Surkyn et al., 2020, 2021), we found that fewer errors are made by girls than by boys and by students from theoretically oriented tracks than by their peers in technical education. We argued that the gender difference is most likely related to a difference in spelling attitudes rather than rule knowledge. We assumed that girls attach more importance to correct verb spelling because women tend to be more careful to avoid stigmatized errors (e.g., Tagliamonte, 2011). We attributed the differences between educational tracks to both a better command of the verb spelling rules and a stronger focus on correct spelling (i.e., a stricter spelling attitude) in GE students. Our survey, administered to 451 high school students, allowed us to check the validity of these interpretations.

We found no evidence that gender plays a role in knowledge of verb spelling rules. In simple conjugation exercises boys and girls performed equally well on simple present forms (exercise 1) and past participles (exercise 2). This absence of a gender difference was also observed in participants’ perception of their own rule mastery: Boys’ and girls’ self-assessments did not differ. However, they differed in the importance they attached to correct spelling. In general, girls seem to be less tolerant towards verb spelling errors than boys in both formal and informal writing contexts. We found a significant relationship between gender and the responses for four of the five statements regarding formal situations: the requirement of correct verb spelling in (1) school tasks, (2) messages to teachers, (3) messages from teachers of Dutch, and (4) messages from teachers of other courses. With regard to informal writing situations, significantly fewer girls considered spelling errors on verb forms acceptable in conversations via WhatsApp and Facebook Messenger. Hence, our previous findings that girls make significantly fewer verb spelling errors than boys in chat posts (Surkyn et al., 2019, 2020, 2021) indeed seems to be related to a stronger focus on correct verb spelling on the part of girls.

As hypothesized, our analyses revealed that adolescents in the more theory-oriented tracks of general education have a better knowledge of the verb spelling rules than students in the more-practice oriented tracks of technical education. GE students’ performance in the conjugation exercises was significantly more often faultless than that of their TE peers (simple present: 75.32 % vs. 30.18 %; past participle: 74.03 % vs. 38.46 %). The self-assessments on rule mastery indicate that TE students are well aware of their poor command of the verb spelling rules. This is in line with earlier findings, which shows that theoretically oriented students make fewer verb spelling errors in dictation tasks and writing assignments than their peers in more practice-oriented tracks (Van den Bergh et al., 2011; Van der Horst et al., 2012). Still our results are quite striking in view of the fact that the spelling task was less demanding than in these previous studies: The verb spellings were elicited in a most straightforward and simple way. Furthermore, we found clear attitudinal differences between GE and TE students. Both in formal and informal writing contexts, GE students care more about correct spelling than TE students. This was the case for all attitudinal statements. Thus, the survey results are consistent with our earlier interpretation of the role of educational track in verb spelling errors in social media writing: The many errors produced by TE students are related to both a lack of rule knowledge as well as a more tolerant attitude towards verb spelling errors.

We can conclude that the results of this survey confirm our interpretation of the gender and educational differences that we observed in adolescents’ online messages (Surkyn et al., 2019, 2020, 2021). Hence, this study provides us with further insight into these earlier attested gender and educational differences since it substantiates conclusions that remained tentative until now.

Moreover, our study shows that many adolescents seem to lack a good command of the verb spelling rules. Only 56.25 % of all 16–20-year-olds were able to complete the simple conjugation exercises on the present tense without mistakes and only 59.00 % were able to do so in the past participle exercise.¹⁴ Considering the clear-cut spelling rules, the low degree of difficulty of the exercises, and the high educational attention for verb spelling in primary and secondary education, these percentages are remarkably low. This relatively poor rule mastery was accompanied by problems in describing the rules (or a mnemonic) and a low confidence in spelling correctness, which were both correlated with the faultless performance in the exercises on the simple present and past participle. Hence, besides working memory overload (Frisson & Sandra, 2002) and a lack of skill in grammatical analysis (Chamalaun et al., 2021), problems with rule knowledge itself seems to be at the basis of many verb spelling errors by adolescents.

Despite often having poor rule knowledge, the majority of adolescents indicated to care about correct verb spelling in formal writing contexts, i.e., roughly 90 % across the five attitudinal statement pertaining to such situations. A recent study of Hilte et al. (2019) also showed that (Flemish) adolescents consider standard language use, which includes correct spelling, important in formal communicative situations, even though they do not think of themselves as having a high standard language proficiency. Note that adolescents’ attitudes were measured by means of a self-report task, both in Hilte et al. (2019) and the current study. Hence, the responses might be influenced by social desirability effects (see e.g., Anderson, 2019; Fazio et al., 1995; Mitchell & Tetlock, 2015): The school context might stimulate the students to pay lip service to the predominant classic standard language ideology (see also Hilte et al., 2019).

In this study we made an explicit distinction between (1) participants’ knowledge of the verb spelling rules for the simple present and the past participle and (2) their attitudes towards errors against these rules. The simple and straightforward design of the survey leaves no room for ambiguity: We cannot but conclude that about half of the adolescents in the final two years of secondary education simply do not master the rules. While the present study is restricted to specific verb spelling rules and consequently does not allow for generalizations with respect to the overall spelling practices of these youngsters, this might serve as a wake-up call for teachers and educationalists. With respect to the social variables, we conclude that theoretically oriented students have a better knowledge, whereas no evidence was found that boys and girls differ in terms of rule knowledge. In spite of poor spelling achievement for these particular verb spelling tasks, most adolescents (± 90 %) consider correct verb spelling important in formal writing contexts. However, error tolerance is much higher in informal contexts (± 50 %). The latter may partly explain the high frequency of verb spelling errors in Flemish adolescent social media writing (see Surkyn et al., 2020, 2021), but the present study reveals that rule mastery definitely is an issue too. Finally, there is a correlation between spelling attitudes and social profile: Theoretically oriented students attach more importance to correct verb spelling, both in formal and informal writing contexts. Girls consider correct verb spelling more important than boys in formal writing contexts and in messages on WhatsApp and Facebook Messenger.

We thank the three anonymous reviewers for their helpful and constructive feedback on the previous versions of this paper.