Universals and Varia-on in Spa-al Referencing: Are Spa-al Mental - - PowerPoint PPT Presentation

SLIDE 1

Universals and Varia-on in Spa-al Referencing: Are Spa-al Mental Models Universal?

Marco Ragni Cogni-ve Computa-on Lab University of Freiburg ragni@informa-k.uni-freiburg.de

SLIDE 2

My focus

Language Background Knowledge Cogni-ve processing (WM)

Spa-al Mental Models

SLIDE 3

Reasoning about spa-al rela-ons I

Which rela-on holds between D and E?

Premises: Ques+on: Model: E is in front of A A* is to the leP of B B is to the leP of C D is in front of C

E A B D C *Just for representa-onal reasons leRers instead of words are used…

SLIDE 4

A is to the leP of C B is to the leP of C E is in front of A A is to the leP of B D is in front of C

E A B D C C E A B D

Which rela-on holds between D and E?

E A B D C

Spa-al reasoning about rela-ons:

Ambiguous representa-on

Johnson-Laird & Byrne, 1991; Knauff et al., 1995; Rauh et al., 2005

SLIDE 5

Reasoning about rela-ons II

Bob is taller than Debbie. Is Charly taller than Debbie? Charly is taller than Adam. Bob is smaller than Adam. Adam is taller than Bob. Is Charly taller than Debbie? Bob is taller than Debbie. Charly is taller than Adam.

Nejasmic, J., Bucher, L. and Knauff, M., 2015. The construc-on of spa-al mental models—A new view on the con-nuity effect. The Quarterly Journal of Experimental Psychology, 68(9), pp.1794-1812.

Discon+nuity and figure effect Con+nuity effect

SLIDE 6

Universals?

Ø Hypothesis: Cogni-ve spa-al processes that are not directly related to language regions in the brain are rather “universal”

Ø Due to “working memory” processes of parsimonious representa-ons Ø Due to the localiza-on of these processes Ø Building upon the Hard- and SoPware idea: Brain, Knowledge, Language

Ø Hypothesis: We can gain some insight if we look which regions are ac-ve in the brain when spa-al rela-onal informa-on is processed?

SLIDE 7

Meta-analysis

Frontal Lobe Parietal Lobe Prefrontal Cortex Inferior Frontal Gyrus vlPFC Superior Parietal Cortex Inferior Parietal Cortex MC PMC (dorsolateral) (pars oper.) (pars triang.) (pars orb.) medial lateral AG TPJ / SMG Author Year BA 4 6 8 9 10 11 44 45 46 47 5 / 7 5 / 7 39 40 Prado et al. 2012

• ()

()

• spatial

Goel & Dolan 2001 (40%)

• Goel et al.

1998

• Knauff et al.

2002

• (60%)
• Knauff et al.

2003

• (70%)

Ruff et al. 2003

• *

Goel et al. 2004

• (60%)
• Goel & Dolan

2001 (60%)

• Knauff et al.

2003

• (70%)

Fangmeier et al. 2006

• l (30%)
• Fangmeier &

2009

• l (50%)
• Knauff

Wendelken & Bunge 2009

• (30)

(60%)

• Prado et al.

2009

• (50%)
• Prado et al.

2010

• (40%)

(40%)

• Shokri-Kojori et al.

2012

• (50%)
• Acuna et al.

2002

• *

(50%)

• visual

Hinton et al. 2010

• (30%)
• Prado et al.

2013 (80%) non-spatial Goel et al. 1998

• ()

Knauff et al. 2003

• (70%)

Knauff et al. 2003

• (70%)

Goel et al. 2009

• ()

Brzeziczka et al. 2011

• (50%)

(80%)

• !

Ragni, Franzmeier, Wenczel & Maier (2014)

What is the role of these regions in the spa-al reasoning process?

SLIDE 8

TMS Methods

Spa-al mental models and the rSPL

Transcranial Magne-c S-mula-on

SLIDE 9

Ø Effect of s-mula-on site: Vertex > SPL s-mula-on Ø Effect of model type: incorrect > correct models preferred > alterna-ve models Ø SPL S-mula-on of correct models affected the accuracy, while incorrect models remained unaffected Ø Working memory capacity moderates effect of TMS

Ragni, Franzmeier, Maier & Knauff (2016). Uncertain rela-onal reasoning in the parietal

• cortex. Brain and cogni-on, 104:72–81.

Localiza-on of Spa-al Mental Models

SLIDE 10

Open ques-ons

Ø Cogni-ve processing: Meta-analysis shows the involvement of the SPL in the model construc-on and varia-on process TMS-Study supports the role of the SPL for the manipula-on of “spa-al mental models“ Ø Is language and background knowledge more relevant for the construc-on of models? Ø Are PMM stable across cultures? Not for topological informa-on (Knauff & Ragni, 2011)? Ø Neuro-experiments so far only for specific rela-ons and only in English/German/Japanese. What is about other languages? Especially languages that use same words for different situa-ons? (Bloom, 1999)