How Common Are Nominalizations in Science Textbooks? Implications for Science Teachers of Emerging Bilinguals

• In Event: Writing and Literacies Across Disciplinary Contexts: Implications for Students Learning English as an Additional Language

Abstract

The sophisticated language demands the new standards place on teachers and students attending American schools may particularly impact Emerging Bilinguals (Bunch, Walqui, & Pearson, 2014; Lee, Quinn, & Valdés, 2013). Yet, the specific linguistic challenges of science texts at different grade levels remain largely unknown (Schleppegrell, 2004). With the purposes of developing descriptions of the language of science texts and identifying the knowledge teachers may need to analyze that language, we (1) analyzed the complexity of identifying nouns as nominalizations, and (2) measured the types and frequencies of nominalizations in middle school science and social studies textbooks. Nominalizations are nouns derived from other word-classes, i.e., verbs, adjectives, or adverbs (Fang, 2008). They are one of the main features of scientific English (Halliday & Martin, 1993) and a source of difficulty for students (Chenhansa & Schleppegrell, 1998; Fang, 2008; Fatonah, 2014). However, nominalizations have not been investigated using large corpora (Liu, Fang, & Wei, 2014) and the few studies conducted have neither addressed the features of textbooks used at the K-12 level nor the implications of these results for teachers.

We compiled a corpus of 12 science and 12 social studies textbooks adopted in California—five publishers, three grade levels per publisher, i.e., 6th, 7th, and 8th grade (Appendix A). The corpus consisted of 1,667,906 words. Using the Stanford log-linear part-of-speech tagger (POS Tagger), we identified 550,000 nouns in our corpus. Then, following Holtz’s (2009) methodology, nominalizations were automatically identified using a suffix-derived approach: i.e., adjective-derived (e.g., -ness) or verb-derived (e.g., -tion). When the derivations were not clear, determinations were made using the Oxford English Online Dictionary (OED, 1987). The data were analyzed using a mixed-effects logistic regression model that treated textbook chapter as the level for repeated measures. The model predicted the binary outcome of a noun being a nominalization with the fixed factor being the subject area.

There was a statistically significant difference between the percentage of nouns that were nominalizations in each of the subject areas: 11% in science vs. 9% in social studies. The percentage of nominalized nouns that stemmed from verbs or adjectives was not significant. In addition, the process of classifying nouns as derived from nominalizations was not simple and required a nuanced understanding of linguistic theory. The complexity of identifying nominalizations challenges the conception that using suffixes to classify nouns as nominalizations is straightforward.

A question remains as to what linguistic knowledge science teachers would need to use linguistic-based strategies accurately and effectively (Fang, 2008; Bunch, 2013; Turkan et al., 2014). Nevertheless, helping students identify how, where, and why nominalizations are used in science texts may support comprehension (Chenhansa & Schleppegrell, 1998; Fang, 2008). In addition, because many nominalizations used in science texts have Latin roots, Spanish-speaking Emerging Bilinguals may have a cognate advantage (Kelley & Kohnert, 2012) when reading science. We are investigating these claims in the second phase of our project.

Authors

• Diego Roman, Southern Methodist University

• Allison Briceno, Texas Woman's University

• Waldo Hasperué, Universidad de la Plata, Argentina

• Kristen Biedermann, Dallas Independent School District

• Greses Perez, Plano Independent School District

• Deni L. Basaraba, Bethel School District 52