SpatioTemporal Metaphors and Time Estimation Panos Athanasopoulos University

Spatio-Temporal Metaphors and Time Estimation Panos Athanasopoulos (University of Reading) p. athanasopoulos@reading. ac. uk Collaborators: Emanuel Bylund (Stockholm University) Alina Schartner (Newcastle University Ifigeneia Athanasiadou (Aristotle University of Thessaloniki) Trolle Carlsson (Stockholm University) Tin Carlsson (Stockholm University) 1

Do speakers of different languages think differently? 2

Sapir-Whorf Hypothesis • Benjamin Lee Whorf (1897 -1941) “The linguistic relativity principle…means, in informal terms, that users of markedly different grammars are pointed by their grammars toward different types of observations and different evaluations of externally similar acts of observation, and hence are not equivalent as observers but must arrive at somewhat different views of the world. ” 3

“Language infects and inflects our thought at every level. The structures of grammar enforce a discipline on our habits of thought” Daniel Dennett “No one is really sure how Whorf came up with his outlandish claims” Steven Pinker “utterly boring, even if true” Geoffrey Pullum 4

Modern ‘neo-Whorfian’ approaches • Evidence for Linguistic Relativity - colour (Athanasopoulos et al. , 2010; Gilbert, et al. , 2006; Kay & Kempton, 1984; Roberson, Davies & Davidoff, 2000) - number (Casasanto, 2005; Frank, et al, 2008; Gordon, 2004; Pica, Lemer, Izard, & Dehaene, 2004; Spelke & Tsivkin, 2001) - motion (Athanasopoulos & Bylund, 2013; Gennari et al. , 2002; Papafragou & Selimis, 2010) - space (Levinson, 1996; Levinson et al. , 2002; Li & Gleitman, 2002; Majid et al. , 2004) - time (Boroditsky, 2001, 2008; Chen, 2007; January & Kako, 2007; Miles et al. , 2012; Núñez & Sweetser, 2006) 5

Time • “a concept around which our whole existence revolves“ • “a system to sequence events” etc. 6

Time as an abstract concept • Crosslinguistic differences in the encoding of time • Grammatical (e. g. , tense, aspect) and lexical (e. g. , adverbials) devices • Focus of today’s talk: Time metaphors Outline: - Spatio-temporal metaphors - Crosslinguistic differences in time perception - The conditions of such differences - Time perception in bilingual speakers 7

Talking about the time that is yet to happen… PAST Swedish: framtid (‘front time’) ie ahead of us, to come 8 FUTURE

Talking about the time that is yet to happen… Aymara: qhipuru (‘behind time’) ie can’t be seen, unknown FUTURE • Commonality of Swedish and Aymara: Succession on horizontal axis 9 PAST

Talking about the time that is yet PAST to happen… Chinese: shàng (‘up’) earlier, past xià (‘down’) later, future; • Chinese uses vertical metaphors in addition to horizontal metaphors FUTURE 10

Talking about time… • Commonality between these ways of talking about time: Spatial reference ”A long rope” ”A long meeting” ”They moved the car forward two meters” ”They moved the meeting forward two hours” Source domain: SPACE (concrete) Target domain: TIME (abstract) (e. g. , Lakoff & Johnson, 1980) 11

Talking about time… • Different types of spatio-temporal metaphors are used to talk about duration: English: Duration = Waiting for a long time Distance long night long party Greek: Duration perimeno poli (’much’) ora = Quantity megali (’big’) nychta parti pou kratise poly [’party that lasted much’] Distance (Germanic languages); Quantity (Spanish, 12 Greek)

Talking about time… Black bars indicate the proportion of Google ‘hits’ for expressions meaning long time, and white bars for expressions meaning much time in English and Greek. Casasanto, et al in prep 13

Thinking abouttime… Talking about • What are the implications of these linguistic encodings of time? • Do we think of time in terms of space? • If so, do speakers with different spatial time metaphors think differently about time? 14

Thinking about time… Duration • What are the effects of these linguistic encodings on time perception? • One way of investigating this is to have speakers of these languages looking at animations that depict different symbolic figures, and estimate their duration (Casasanto et al. , 2004; 2005; 2008; 2013) 15

line = distance 16

17

container = quantity 18

19

Thinking about time… Duration • In the test, the participant is given two different kinds of information: 1) temporal information (the duration of the stimulus) 2) spatial information (the length/growth of the stimulus) • If the spatial metaphors that we use to talk about time actually influence our thinking about time, then we would expect an effect of spatial information on time perception 20

A 21

B 22

A 23

B 24

Thinking about time… Duration • Speakers of languages with length metaphors would be influenced by line length when estimating the duration of line animations, i. e: They would tend to think that longer lines have a longer duration • In contrast, speakers of quantity metaphor languages would be influenced by the degree to which the containers are filled, i. e: They would think that the more a container is filled the more time has passed 25

Experimental design Two measures are calculated: 1) Accuracy of duration estimation 2) Spatial interference 26

Spatial interference Estimated duration (ms) Estimated duration – Line/Filling container growth 4000 3500 3000 2500 2000 slope = 0 1500 1000 0 100 200 300 400 Length/Growth (pixels) 500 600 27

Spatial interference Estimated duration (ms) Estimated duration – Line/Filling container growth 3800 3600 3400 3200 3000 2800 2600 2400 2200 2000 slope = 1. 39 0 100 200 300 400 Length/Growth (pixels) 500 600 28

Casasanto (2005, 2008, 2013) • Speakers of English and Indonesian (distance languages) • Speakers of Greek and Spanish (quantity languages) 29

Casasanto (2005, 2008, 2013) • Crosslinguistic differences in spatial interference in the line task: Slope English Indonesian Greek Spanish 1. 49 1. 40 . 13 . 47 • Crosslinguistic differences in spatial interference in the container task: Slope English Indonesian Greek Spanish . 18 . 51 1. 16 1. 24 30

Casasanto (2005, 2008, 2013) • Speakers of English and Indonesian were influenced by line distance when estimating time, but not by filling container growth • The opposite pattern was found for speakers of Greek and Spanish • These findings seem to indicate that time perception indeed differs across language groups, and it does so in a way that corresponds to spatiotemporal metaphors for duration 31

The conditions of language-specificity • In Casasanto et al. (2005, 2008, 2013), lines and containers were preceded by a prompt that indicated the task, i. e. TIME DISTANCE • Bylund & Athanasopoulos (in prep) removed the linguistic label of the prompt, leaving only the symbol 32

Bylund & Athanasopoulos (in prep) • Spatial interference, containers [+ linguistic label] Slope Spanish Swedish 1. 42 . 54 p <. 05 • Spatial interference, containers [– linguistic label] Slope Spanish Swedish . 96 . 65 p >. 1 33

Bylund & Athanasopoulos (in prep) • Spatial interference, lines [+ linguistic label] Slope Spanish Swedish . 70 1. 21 p <. 05 • Spatial interference, lines [– linguistic label] Slope Spanish Swedish . 81 1. 03 p >. 1 34

Bylund & Athanasopoulos (in prep) • Crosslinguistic differences in spatial interference are reduced in the [– linguistic label] condition • The linguistic prompts trigger a set of perceptual distinctions learnt through and associated with language, thus leading the individual to attend to perceptual attributes in a language-specific way 35

Accounting for the influence of metaphor on thought • Associative learning: when people use a linguistic metaphor for time, they activate the corresponding mental metaphor. In doing so, they would strengthen this particular associative mapping. • As people use the dominant and less-dominant metaphors in their language according to their distributional statistics, they activate one mental metaphor more frequently than the other(s). • This should strengthen one mental metaphor, and at the same time weaken the alternative mapping(s). 36

Does language shape the way we think? • If specific space-time associations are strengthened by frequency of use, then bilinguals might be influenced by the language they use most often 37

Bylund & Athanasopoulos (in prep) • L 1 Spanish – L 2 Swedish adult bilinguals, living in Sweden Age of L 2 acquisition: 11. 5 (7. 8) years Frequency of L 1 use: 21. 8 % weekly Frequency of L 2 use: 78. 2% weekly Length of residence: 20. 4 (6. 1) years • Experimental conditions: Lines Linguistic labels (Swedish: tid) 38

Bylund & Athanasopoulos (in prep) • Spatial interference, lines Slope Spanish-Swedish bilinguals Swedish mono . 92 1. 21 p <. 05 • Spatial interference, lines Slope Bilinguals: Spanish users Bilinguals: Swedish users Swedish mono . 65 1. 18 1. 21 p <. 05 p >. 1 39

Summary • Frequency of language use affects spatial interference • Spatial interference in bilinguals using Spanish more frequently converges with Spanish patterns (i. e. , L 1 patterns) • Spatial interference in bilinguals using Swedish more frequently converges with Swedish patterns (i. e. , L 2 patterns) • How early in language development do mental space-time associations appear? Frequency of exposure? Learning context? 40

Language and Thought Linguistic relativity: • People who talk about time differently also think about it differently • What is the extent of the influence of linguistic structure on cognitive processes, and what conditions suppress or promote this influence? Conceptual representation • Learning and using a specific language can shape mental representations by strengthening specific space-time associations 41

Thank you! 42

Bylund & Athanasopoulos (in prep) • Spatial interference, containers Spanish-Swedish bilinguals Swedish mono r 2 . 61 . 38 Slope . 68 . 54 p <. 05 • Spatial interference, containers Late bilinguals Early bilinguals Swedish mono r 2 . 88 . 45 . 38 Slope 1. 01 . 44 . 54 p <. 05 p >. 1 43

Modern ‘neo-Whorfian’ approaches Linguistic diversity: Languages encode reality in different ways Thinking for Speaking: Speakers structure information differently when they prepare content for speech Linguistic relativity: Speakers of different languages think and perceive the world differently Indexed by verbal and co-verbal behaviour Indexed by non-verbal cognitive behaviour 44