The importance of situation-specific encodings: analysis of a simple connectionist model of letter transposition effects

This paper analyses a three-layer connectionist network that solves a translation-invariance problem, offering a novel explanation for transposed letter effects in word reading. Analysis of the hidden unit encodings provides insight into two central issues in cognitive science: (1) What is the novel...

Full description

Saved in:
Bibliographic Details
Main Authors: Shin-Yi Fang, Garrett Smith, Whitney Tabor
Format: Article
Language:English
Published: Taylor & Francis Group 2018-04-01
Series:Connection Science
Subjects:
Online Access:http://dx.doi.org/10.1080/09540091.2016.1272097
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849764148870643712
author Shin-Yi Fang
Garrett Smith
Whitney Tabor
author_facet Shin-Yi Fang
Garrett Smith
Whitney Tabor
author_sort Shin-Yi Fang
collection DOAJ
description This paper analyses a three-layer connectionist network that solves a translation-invariance problem, offering a novel explanation for transposed letter effects in word reading. Analysis of the hidden unit encodings provides insight into two central issues in cognitive science: (1) What is the novelty of claims of “modality-specific” encodings? and (2) How can a learning system establish a complex internal structure needed to solve a problem? Although these topics (embodied cognition and learnability) are often treated separately, we find a close relationship between them: modality-specific features help the network discover an abstract encoding by causing it to break the initial symmetries of the hidden units in an effective way. While this neural model is extremely simple compared to the human brain, our results suggest that neural networks need not be black boxes and that carefully examining their encoding behaviours may reveal how they differ from classical ideas about the mind-world relationship.
format Article
id doaj-art-c09e486fc2cb4f50af33887e95b4d4f7
institution DOAJ
issn 0954-0091
1360-0494
language English
publishDate 2018-04-01
publisher Taylor & Francis Group
record_format Article
series Connection Science
spelling doaj-art-c09e486fc2cb4f50af33887e95b4d4f72025-08-20T03:05:13ZengTaylor & Francis GroupConnection Science0954-00911360-04942018-04-0130213515910.1080/09540091.2016.12720971272097The importance of situation-specific encodings: analysis of a simple connectionist model of letter transposition effectsShin-Yi Fang0Garrett Smith1Whitney Tabor2University of ConnecticutUniversity of ConnecticutUniversity of ConnecticutThis paper analyses a three-layer connectionist network that solves a translation-invariance problem, offering a novel explanation for transposed letter effects in word reading. Analysis of the hidden unit encodings provides insight into two central issues in cognitive science: (1) What is the novelty of claims of “modality-specific” encodings? and (2) How can a learning system establish a complex internal structure needed to solve a problem? Although these topics (embodied cognition and learnability) are often treated separately, we find a close relationship between them: modality-specific features help the network discover an abstract encoding by causing it to break the initial symmetries of the hidden units in an effective way. While this neural model is extremely simple compared to the human brain, our results suggest that neural networks need not be black boxes and that carefully examining their encoding behaviours may reveal how they differ from classical ideas about the mind-world relationship.http://dx.doi.org/10.1080/09540091.2016.1272097connectionismlearningmodality-specific encodingssymmetry breakingletter transposition
spellingShingle Shin-Yi Fang
Garrett Smith
Whitney Tabor
The importance of situation-specific encodings: analysis of a simple connectionist model of letter transposition effects
Connection Science
connectionism
learning
modality-specific encodings
symmetry breaking
letter transposition
title The importance of situation-specific encodings: analysis of a simple connectionist model of letter transposition effects
title_full The importance of situation-specific encodings: analysis of a simple connectionist model of letter transposition effects
title_fullStr The importance of situation-specific encodings: analysis of a simple connectionist model of letter transposition effects
title_full_unstemmed The importance of situation-specific encodings: analysis of a simple connectionist model of letter transposition effects
title_short The importance of situation-specific encodings: analysis of a simple connectionist model of letter transposition effects
title_sort importance of situation specific encodings analysis of a simple connectionist model of letter transposition effects
topic connectionism
learning
modality-specific encodings
symmetry breaking
letter transposition
url http://dx.doi.org/10.1080/09540091.2016.1272097
work_keys_str_mv AT shinyifang theimportanceofsituationspecificencodingsanalysisofasimpleconnectionistmodeloflettertranspositioneffects
AT garrettsmith theimportanceofsituationspecificencodingsanalysisofasimpleconnectionistmodeloflettertranspositioneffects
AT whitneytabor theimportanceofsituationspecificencodingsanalysisofasimpleconnectionistmodeloflettertranspositioneffects
AT shinyifang importanceofsituationspecificencodingsanalysisofasimpleconnectionistmodeloflettertranspositioneffects
AT garrettsmith importanceofsituationspecificencodingsanalysisofasimpleconnectionistmodeloflettertranspositioneffects
AT whitneytabor importanceofsituationspecificencodingsanalysisofasimpleconnectionistmodeloflettertranspositioneffects