Document Detail


Distinctive signatures of recursion.
MedLine Citation:
PMID:  22688640     Owner:  NLM     Status:  MEDLINE    
Abstract/OtherAbstract:
Although recursion has been hypothesized to be a necessary capacity for the evolution of language, the multiplicity of definitions being used has undermined the broader interpretation of empirical results. I propose that only a definition focused on representational abilities allows the prediction of specific behavioural traits that enable us to distinguish recursion from non-recursive iteration and from hierarchical embedding: only subjects able to represent recursion, i.e. to represent different hierarchical dependencies (related by parenthood) with the same set of rules, are able to generalize and produce new levels of embedding beyond those specified a priori (in the algorithm or in the input). The ability to use such representations may be advantageous in several domains: action sequencing, problem-solving, spatial navigation, social navigation and for the emergence of conventionalized communication systems. The ability to represent contiguous hierarchical levels with the same rules may lead subjects to expect unknown levels and constituents to behave similarly, and this prior knowledge may bias learning positively. Finally, a new paradigm to test for recursion is presented. Preliminary results suggest that the ability to represent recursion in the spatial domain recruits both visual and verbal resources. Implications regarding language evolution are discussed.
Authors:
Maurício Dias Martins
Publication Detail:
Type:  Journal Article; Research Support, Non-U.S. Gov't; Review    
Journal Detail:
Title:  Philosophical transactions of the Royal Society of London. Series B, Biological sciences     Volume:  367     ISSN:  1471-2970     ISO Abbreviation:  Philos. Trans. R. Soc. Lond., B, Biol. Sci.     Publication Date:  2012 Jul 
Date Detail:
Created Date:  2012-06-12     Completed Date:  2012-10-15     Revised Date:  2013-04-16    
Medline Journal Info:
Nlm Unique ID:  7503623     Medline TA:  Philos Trans R Soc Lond B Biol Sci     Country:  England    
Other Details:
Languages:  eng     Pagination:  2055-64     Citation Subset:  IM    
Affiliation:
Department of Cognitive Biology, University of Vienna, Althanstrasse 14, Vienna 1090, Austria. mauricio.martins@univie.ac.at
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:
Algorithms
Behavior / physiology*
Cognition / physiology*
Computational Biology
Humans
Language*
Memory, Short-Term / physiology
Verbal Learning
Visual Perception / physiology

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  How semantic biases in simple adjacencies affect learning a complex structure with non-adjacencies i...
Next Document:  Processing multiple non-adjacent dependencies: evidence from sequence learning.