Space - Language as Spotlight (Topological Relations) Flashcards
Topological Relations
Linguistic way of abstracting spatial relationship between a figure and ground for communicative purposes, and simplest spatial relation is coincidence of figure and ground with simplest notions in English being: at, on, in (prepositions
Can be understood as support and containment
Baillargeon (1998) - prelinguistic spatial categories
Studied babies knowledge of spatial relations, showing infants have rich and developing knowledge before knowing specific words to talk about space in language
Gave infants different images to see at which point they are surprised by what they see
For support: distinction between support and non-support in first year, gradual learning of quantitative details e.g., how much overlap between objects needed if one is to support the other
For containment: fewer systematic developments but (1) at 5.5mo, know containers must have bottoms; (2) at 8.5mo, take into account width and compressibility of object when determining whether it can be put into a container
Suggested infants first acquire spatial concepts non-linguistically, and that they then learn which words map onto existing learned concepts
Universality of Spatial Concepts - Landau and Jackendoff (1993)
Proposed that there are universal spatial concepts, mirrored by small number of prepositions in English language
No Universal Spatial Concepts - Bowerman and Choi (2001)
Cross-linguistic data crucial (Majid, 2018) and suggests variation in what notions conceptualised as spatial in first place
Part-whole relationships treated spatially in Germanic languages, but not others e.g., muscles in my leg are sore, which contrasts with languages that use possessive relationships which have no spatial information e.g., muscles of my leg are sore
English treats ‘smeary things’, holes, and cracks, as if they are figures that cannot be located, but other languages do not e.g., butter is on the knife vs knife is buttery
No Universal Spatial Concepts - Bowerman and Pederson (1992)
Cross-linguistic naming study systematically investigating whether meanings such as “on” and “in” are universal > pps of dozens of unrelated languages saw pictures of various objects and answers “where is the X” > either maximal examples of support e.g., cup on table and containment e.g., apple in bowl, with potentially ambiguous stimuli between two maxima
Substantial variation in meanings of basic concepts: (1) English used “on” for all but “apple in bowl”, including ambiguous stimuli such as “picture on wall”; (2) Dutch included preposition for support relationship where one thing is defying gravity; (3) Berber showed overlap in how stimuli spatially conceived; (4) Spanish one preposition for everything
McDonough et al (2003) preferential looking
Used dynamic support/containment stimuli: for English, need a verb e.g., “put in / put on”, but in Korean, doesn’t separate spatial prepositions, but different verbs used for different types of relations that cross-cut English distinctions, instead distinguishing something to effect of tight/interlock and loose
Preferential looking task with 9, 11, and 14 month old English and Korean children, and adults
Familiarised with series of videos with different objects either all tight-fit or loose-fit relations, shown two items at a time as not wanting infants to abstract from any specific objects and instead focus on relational properties
Shown test screen where either congruent or incongruent with familiarisation scenes, which was counterbalanced
Both English and Korean infants (9-14mo) discriminated categorically between tight-IN and loose-IN relations, with Korean adults, but not English adults, doing the same
Suggests children that haven’t learned spatial language are sensitive, but by adulthood, lose sensitivity based on learned spatial language (language as spotlight)
McDonough et al (2003) adult odd-one-out task
Koreans immediately picked out tight/loose relation, but English unable to pick out relevant difference > even in task where opportunity to think about tightness of fit, dimension not as salient to English adults as Korean adults
Since all stimuli were containment, long-term exposure to and use of English obscured differences between tight and loose fit (language as spotlight)
English adults gave explanations for bases of selections in relation to general object properties, whereas Koreans focused on relational properties, which are more salient in Korean language
Different foci of attention parallels crosslinguistic differences in Choi and Gopnik (1995) on early lexical development: English-learning children focus primarily on acquiring nominal words (nouns), but Korean-learning children acquire many relational words
Choi (2006) - support for McDonough et al (2003)
Supports McDonough et al (2003) as found that by 3, sensitivity to distinction between tight and loose fit diminishes greatly in English-speaking children, although ability to distinguish between the two seen as early as 5 months
McDonough et al (2003) Criticisms
None of studies explicitly assess whether it is learned language that results in a loss of sensitivity to tight/loose dimension > no interference tested, nor training, and perhaps change from infants to adults either solely cultural, or culture moderating influence of language > difficult to tease apart (Majid, 2018)
McDonough et al (2003) Criticism - Norbury et al (2009)
Classify tight-loose as a whole dimension of “fit”, and state that McDonough et al (2003) averaged over tight- and loose-fitting events, providing composite index of speakers’ sensitivity to “fit”, but previous experiments reported tight- and loose-fitting events independently, and although English adults didn’t reliable demonstrate attention to “fit” in any conditions, attention to “fit” more pronounced when pps familiarised to tight- than to loose-fitting events > English adults may represent “fit” but in asymmetric fashion
Asymmetry towards tight, not loose, also seen in Korean data
Suggests sensitivity to “fit” not a consequence of language acquired > however, source of asymmetry only supported by analogous evidence of existing asymmetries in other domains, but no specific evidence base for why this exists, although argue close-fitting nature may be more perceptually salient
Support by Choi (2006): at ~3y, Korean children maintained sensitivity to “fit” in context of both tight and loose, but English weakened, although closer examination reveals “fit” attention more likely if familiarised to tight- than to loose-fitting events
