Event Title
Loading...
Publication Date
2020
Start Date
14-11-2020 1:20 PM
End Date
14-11-2020 1:25 PM
Description
Preschoolers often appear to not understand the logical relationships expressed by words like “or”, and “if/then”. Despite the assumption that errors in logic are due to children not understanding, there might be principled reasons behind children not demonstrating adult-like behavior. If a type of error is consistent and pattern-like we might expect that a similar process is occurring behind such errors for all logical connectives, yet research to date has struggled to show a unified approach that accounts for this. Semantic modelling using multiple exhaustification can account for errors in disjunction (or) but fails to explain similar errors in conditionals (if/then). Mental model theories and heuristic analyses of conditionals account for errors in if/then syllogisms but do not account for disjunctive errors. What if instead these consistent errors were created by the lack of a suitable context for the experiment? Through comparing two groups of children, where one is exposed to a child-alien who is new to earth and learning English, and the other is exposed to a human who sometimes says silly things. We predict that we can elicit more predictable adult-like responses from the children in the child-alien group. If it is the case that we can reliably and predictably create an environment that shows children do comprehend these kinds of words, this has important implications for both language acquisition and research methodology.
See full abstract linked below.
Video transcript
Iwanchuk, Skordos-mACOL2020-abstract.pdf (194 kB)
Full abstract
Children's Understanding of Logical Connectives
Preschoolers often appear to not understand the logical relationships expressed by words like “or”, and “if/then”. Despite the assumption that errors in logic are due to children not understanding, there might be principled reasons behind children not demonstrating adult-like behavior. If a type of error is consistent and pattern-like we might expect that a similar process is occurring behind such errors for all logical connectives, yet research to date has struggled to show a unified approach that accounts for this. Semantic modelling using multiple exhaustification can account for errors in disjunction (or) but fails to explain similar errors in conditionals (if/then). Mental model theories and heuristic analyses of conditionals account for errors in if/then syllogisms but do not account for disjunctive errors. What if instead these consistent errors were created by the lack of a suitable context for the experiment? Through comparing two groups of children, where one is exposed to a child-alien who is new to earth and learning English, and the other is exposed to a human who sometimes says silly things. We predict that we can elicit more predictable adult-like responses from the children in the child-alien group. If it is the case that we can reliably and predictably create an environment that shows children do comprehend these kinds of words, this has important implications for both language acquisition and research methodology.
See full abstract linked below.