Language and culture in perception: a three-pronged investigation of phylogenetic, ontogenetic and cross-cultural evidence

Brown and Lenneberg (I954) and Rosch Heider (1972) were among the first to conduct psychological investigations to test the Whorfian view that language affects thought. They both asked about colour categories. The debate has continued with some research supporting a relativist (Whorfian) account (Davidoff, Davies & Roberson, I999; Borodistsky, 200I), and some supporting a universalist account (e.g., Kay & Regier, 2003; Spelke & Kinzler, 2007). The present thesis adds to the debate by taking three different approaches i.e., cross-cultural, ontogenetic and phylogenetic frames in which to carry out investigations of categorization of various perceptual continua. Categorical Perception's hallmark is the effect of mental warping of space such as has been found for phonemes (Pisani & Tash, I974) and colour (Bornstein & Monroe, I980; Bornstein & Korda, I984). With respect to colours, those that cross a category boundary seem more distant than two otherwise equally spaced colours from the same category. Warping is tested using cognitive methods such as two-alternative-forced-choice and matching-to-sample. Evidence is considered for the continua under investigation i.e. colour and animal patterns. Experiments 1 and 2 find evidence of categorical perception for human-primates and not for monkeys. Experiment 3 finds that Himba and English human adults categorize differently, particularly for colours crossing a category boundary, but also show broad similarity in solving the same matching-to-sample task as used with the monkeys (experiment I) who showed clear differences with humans. Experiment 4 and 5 tested Himba and English toddlers and found categorical perception of colour mainly for toddlers that knew their colour terms despite prior findings (Franklin et al., 2005) indicative of universal colour categories. In experiment 6, Himba and English categorical perception of animal patterns was tested for the first time, and result indicate a cross-category advantage for participants who knew the animal pattern terms. Therefore, a weak Whorfian view of linguistic relativity's role in obtaining categorical perception effects is presented. Although there is some evidence of an inherent human way of grouping drawn from results of experiment 1 and 3, results in all experiments (1,2,3,4,5, and 6) show that linguistic labels and categorical perception effects go hand-in-hand; categorization effects are not found when linguistic terms are not acquired at test and have not had a chance to affect cognition. This was true for all populations under observation in this set of studies, providing further support for effects of language and culture in perception