A compositional shape code explains how we read jumbled words

We raed jubmled wrods effortlessly, yet the visual representations underlying this remarkable ability remain unknown. Here, we show that well-known principles of neural object representations can explain orthographic processing. We constructed a population of neurons whose responses to single letters matched perception, and whose responses to multiple letters was a weighted sum of its responses to single letters. This simple compositional letter code predicted human performance both in visual search as well as on explicit word recognition tasks. Unlike existing models of word recognition, this code is neurally plausible, seamlessly integrates letter shape and position, and does not invoke any specialized detectors for letter combinations. Our results suggest that looking at a word activates a compositional shape code that enables its efficient recognition. SIGNIFICANCE STATEMENT Reading is a recent cultural invention, but we are remarkably good at reading words and even jubmeld words. It has so far been unclear whether this ability is due to a representation specialized for letter shapes, or is inherited from basic principles of visual processing. Here we show that a large variety of word recognition phenomena can be explained by well-known principles of object representations, whereby single neurons are selective for the shapes of single letters and respond to longer strings according to a compositional rule.

• Publication year

  2019

• Venue

  bioRxiv

• Publication date

  2019-05-30

• Fields of study

  Biology, Computer Science, Psychology

• Identifiers
  DOI 10.1101/653048
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Reading Increases the Compositionality of Visual Word Representations

  2019cited by this paper
• Stimulus orientation and the first-letter advantage

  2018cited by this paper
• Orthographic processing: A ‘mid-level’ vision of reading: The 44th Sir Frederic Bartlett Lecture

  2018cited by this paper
• Representation of multiple objects in macaque category-selective areas

  2018cited by this paper
• Symmetric Objects Become Special in Perception Because of Generic Computations in Neurons

  2017cited by this paper
• Is nevtral NEUTRAL? Visual similarity effects in the early phases of written-word recognition

  2017cited by this paper
• Neurophysiological dynamics of phrase-structure building during sentence processing

  2017cited by this paper
• Object attributes combine additively in visual search

  2016cited by this paper
• Ocular-following responses to white noise stimuli in humans reveal a novel nonlinearity that results from temporal sampling

  2016cited by this paper
• Look before you seek: Preview adds a fixed benefit to all searches

  2016cited by this paper
• Neural representation for object recognition in inferotemporal cortex.

  2016cited by this paper
• Influence of lexical status and orthographic similarity on the multi-voxel response of the visual word form area

  2015cited by this paper
• Coding of relative size in monkey inferotemporal cortex

  2015cited by this paper
• Responding to nonwords in the lexical decision task: Insights from the English Lexicon Project.

  2015cited by this paper
• Illiterate to literate: behavioural and cerebral changes induced by reading acquisition

  2015cited by this paper
• Dynamics of 3D view invariance in monkey inferotemporal cortex

  2015cited by this paper
• Texture discriminability in monkey inferotemporal cortex predicts human texture perception

  2014cited by this paper
• GloVe: Global Vectors for Word Representation

  2014cited by this paper
• Features in visual search combine linearly.

  2014cited by this paper
• A Toolbox for Representational Similarity Analysis

  2014cited by this paper
• The face inversion effect in infants is driven by high, and not low, spatial frequencies.

  2014cited by this paper
• Transposed-Letter Effects Reveal Orthographic Processing in Baboons

  2013cited by this paper
• Does linear separability really matter? Complex visual search is explained by simple search.

  2013cited by this paper
• ICA model order selection of task co-activation networks

  2013cited by this paper
• Generalization between canonical and non-canonical views in object recognition.

  2013cited by this paper
• Models of visual word recognition

  2013cited by this paper
• Reading through a noisy channel: why there's nothing special about the perception of orthography.

  2012cited by this paper
• How to say "no" to a nonword: a leaky competing accumulator model of lexical decision.

  2012cited by this paper
• Orthographic Processing in Baboons (Papio papio)

  2012cited by this paper
• Turning visual search time on its head.

  2012cited by this paper
• Extra-large letter spacing improves reading in dyslexia

  2012cited by this paper
• A letter visual-similarity matrix for Latin-based alphabets

  2012influential reference
• Serial position effects in the identification of letters, digits, symbols, and shapes in peripheral vision.

  2012cited by this paper
• Alphabetic letter identification: effects of perceivability, similarity, and bias.

  2012influential reference
• Cortical representation of the constituent structure of sentences

  2011cited by this paper
• Global Image Dissimilarity in Macaque Inferotemporal Cortex Predicts Human Visual Search Efficiency

  2010cited by this paper
• The spatial coding model of visual word identification.

  2010cited by this paper
• Responses to Compound Objects in Monkey Inferotemporal Cortex: The Whole Is Equal to the Sum of the Discrete Parts

  2010cited by this paper
• How Learning to Read Changes the Cortical Networks for Vision and Language

  2010cited by this paper
• R34D1NG W0RD5 W1TH NUMB3R5.

  2008cited by this paper
• The overlap model: a model of letter position coding.

  2008cited by this paper
• The uncrowded window of object recognition

  2008cited by this paper
• The Diffusion Decision Model: Theory and Data for Two-Choice Decision Tasks

  2008cited by this paper
• Spatial Summation Can Explain the Attentional Modulation of Neuronal Responses to Multiple Stimuli in Area V4

  2008cited by this paper
• Moving beyond Coltheart’s N: A new measure of orthographic similarity

  2008cited by this paper
• Parts, Wholes, and Context in Reading: A Triple Dissociation

  2007cited by this paper
• The English Lexicon Project

  2007cited by this paper
• Hierarchical coding of letter strings in the ventral stream: dissecting the inner organization of the visual word-form system.

  2007cited by this paper
• Raeding Wrods With Jubmled Lettres

  2006cited by this paper
• Distractor familiarity leads to more efficient visual search for complex stimuli

  2005cited by this paper
• Multiple Object Response Normalization in Monkey Inferotemporal Cortex

  2005cited by this paper
• The neural code for written words: a proposal.

  2005cited by this paper
• A new SPM toolbox for combining probabilistic cytoarchitectonic maps and functional imaging data

  2005cited by this paper
• A diffusion model account of the lexical decision task.

  2004cited by this paper
• The visual word form area and the frequency with which words are encountered: evidence from a parametric fMRI study

  2004cited by this paper
• Does the huamn mnid raed wrods as a wlohe?

  2004cited by this paper
• Inferotemporal neurons represent low-dimensional configurations of parameterized shapes

  2001cited by this paper
• Eye movements in reading and information processing: 20 years of research.

  1998cited by this paper
• The Psychophysics Toolbox.

  1997cited by this paper
• Visual search and stimulus similarity.

  1989cited by this paper
• An interactive activation model of context effects in letter perception: Part 2. The contextual enhancement effect and some tests and extensions of the model.

  1982cited by this paper
• On the Use of Visual Cues during Visual-word Recognition

  year unknowncited by this paper

• No citing papers are available for this paper.