In the present research, we investigated the influence of native-language adjective-noun word order on category accessibility for nouns and adjectives by comparing Portuguese speakers (in whose language nouns precede adjectives) with English speakers (in whose language adjectives precede nouns). In two studies, we presented participants with different numbers of verbal or pictorial stimuli, and subsequently they answered questions about noun- and adjective-conditioned frequencies. The results demonstrated a primacy effect of native-language word order. Specifically, although both populations showed a speed advantage for noun-conditioned questions, this tendency was significantly stronger for Portuguese than for American participants. We discuss the important role of native-language syntax rules for the categorization and representation of information.
Decades of inquiry have gone into investigating the notion that language shapes human experience, on the premise that the world is highly complicated and that language assists in making it palatable and understandable. In the present work, we address a related, but more specific, proposition: that the various ways in which languages grammatically organize experience lead to different ways of cognitively organizing experience.
Whorf's (1956) linguistic relativity principle is the idea that differences across languages can lead to corresponding differences in cognition. Although early studies (e.g., Heider & Olivier, 1972) were unsupportive of Whorf, more recently linguistic relativity has taken on a new face in a new generation of empirical work that has successfully demonstrated the important role of language in cognitive processes. Characteristic of this recent work is the proposition that, even if language does not determine thought, there still exists much room for the influence of language. These are commonly referred to as the strong and weak forms of the Whorfian hypothesis, respectively (see Boroditsky, 2001). Strong linguistic relativity purports that language severely limits thought, such that subtleties of experience not labeled by one's language are literally imperceptible. Its weak counterpart implies crosslinguistic differences in cognitive tendency, rather than potentiality. It is this weak form that is the focus of much recent work.
For example, Boroditsky, Schmidt, and Phillips (2003) investigated the effects of gendered articles (through which certain objects are arbitrarily given either feminine or masculine designations) on object perception. They presented speakers of Spanish and German with objects of opposing genders between the two languages. Participants' descriptions were congruent with each object's grammatical gender in their native language. For example, key was described as stereotypically masculine (e.g., "heavy, jagged, metal") by Germans, but as feminine (e.g., "golden, intricate, little") by Spaniards. Similar crosslinguistic differences have been demonstrated for perceptions of causality as a function of causal language rules (Fausey & Boroditsky, 2008), the ability to remember and match arrays of objects as a function of available counting words (Frank, Everett, Fedorenko, & Gibson, 2008), stereotypic judgments as a function of available stereotype labels (Hoffman, Lau, & Johnson, 1986), and perceptions of colors as a function of available color words (Roberson, Davies, & Davidoff, 2000).
The Present Studies: General Rationale and Approach
Unlike many earlier investigations, the present study examines not what happens when speakers of a language lack particular labels, but how speakers of different languages with different syntax rules behave in situations in which multiple labels apply. For example, a red shirt is both a red thing and a shirt. A person organizing a pile of clothes might sort it into the "red" pile or into the "shirt" pile, depending on whether the task was doing laundry or putting clothes into drawers. …