Explicitness in science discourse: a Gricean account of income-related differences

Abstract

Highly explicit language use is prized in scientific discourse, and greater explicitness is hypothesized to facilitate academic achievement. Studies in the mid-twentieth century reported controversial findings that the explicitness of text differs by the income and education levels of authors’ families. If income-related differences in explicitness persist today, this may be one of the mechanisms by which parent income currently affects academic achievement. Therefore, this study reexamines the relationship between parent income and reliance on deictic terms, one proxy for explicitness, in the science writing of 100 5th grade students. Findings suggest that income-related differences in reliance on deictic terms persist in the twenty-first century, that students from low-income families are more ‘novle’ or assume a relatively less knowledgeable reader than students from higher income families, and that this novility is inversely related to science achievement. Implications for instruction are discussed.

Keywords:

• explicitness
• academic language
• deictic terms
• arridence
• novility
• discourse pragmatics

View correction statement:

Corrigendum

Acknowledgements

The authors would like to acknowledge Shannon Oxman for her feedback on the coding scheme used in this paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Meaning is always constructed in relation to a social context (Geertz 1994). All texts and words depend in some way on social context to be meaningful. Therefore, texts cannot be objectively ‘explicit’ or ‘implicit’; a text can only be more explicit or less explicit than an alternative text.

2. There are a variety of linguistic tools that can be used to raise or lower the explicitness of text. A full description of the explicitness in a given text would require investigation of a range of linguistic features, many of which would require higher inference coding. Here, the low-inference proxy ‘deictic terms’ was chosen to reduce the possibility that the analysis would be skewed by middle class bias (Jones 2013).

3. This means that two texts could be equally explicit, yet if the first text is markedly longer, it might include more deictic terms. Relative explicitness then does not depend on the raw number of deictic terms, but rather the degree to which the writer relies on deictic terms in place of noun phrases. In order to avoid confusing word count for explicitness, this study measures reliance on deictic terms by calculating a deictic term to total words ratio for each student.

4. ‘One’, ‘thing’, and ‘stuff’ were included here as deictic terms because these were words, like ‘they’ and ‘it’, that relied more heavily than other nouns on the physical context for meaning.