Visual word processing and experiential origins of functional selectivity in human extrastriate cortex

Chris I. Baker, Jia Liu, Lawrence L. Wald, Kenneth K. Kwong, Thomas Benner, Nancy KanwisherProceedings of the National Academy of Sciences 2007 · DOI

TL;DR

This study uses high-resolution fMRI to identify a specific brain region in the ventral visual cortex that is highly selective for visual word forms. Crucially, they demonstrate that this selectivity is not innate but is shaped by extensive experience with reading, as evidenced by stronger responses to Hebrew words in Hebrew readers compared to non-readers.

What Problem Does This Paper Address?

The human brain exhibits remarkable specialization, with distinct regions dedicated to processing specific categories of visual information, such as faces (FFA) and places (PPA). A key question in neuroscience is how these category-selective regions arise: are they genetically pre-programmed, or do they develop through experience? The authors tackle this by investigating a region in the extrastriate cortex that responds to visual words, often called the Visual Word Form Area (VWFA). Since reading is a relatively recent human invention (a few thousand years), it's unlikely that a dedicated brain region for word recognition evolved through natural selection. Therefore, visual words serve as an ideal test case to explore the role of experience in shaping cortical selectivity. The paper aims to determine how specific this word-selective region is and, more importantly, whether its selectivity is a product of learned experience.

Methods

The researchers employed high-resolution functional magnetic resonance imaging (fMRI) with voxel sizes of 1.4 x 1.4 x 2 mm. This resolution was chosen to be sensitive enough to detect the small, specialized regions of interest. They used two main fMRI experimental designs:

  • Blocked Design: Participants viewed blocks of stimuli (e.g., English words, consonant strings, Hebrew words, line drawings) for 15 seconds each, interleaved with fixation blocks. This design is good for identifying regions that respond more strongly to one category over another.
  • Event-Related Design: Stimuli were presented individually on interleaved trials, allowing for a more precise estimation of the hemodynamic response function to each stimulus type.

Participants were presented with various visual stimuli, including English words, consonant strings, Hebrew words, digit strings, Chinese characters, and line drawings of objects. They also performed tasks such as passive viewing, a one-back task (detecting repeated stimuli), and a motion detection task. Statistical analyses involved comparing the blood oxygen level-dependent (BOLD) signal changes between different stimulus conditions to identify regions of selective activation.

Datasets

The study used a dataset of visual stimuli including:

  • English words
  • English consonant strings
  • Hebrew words
  • Digit strings
  • Chinese characters
  • Line drawings of objects
  • Photographs of faces and objects (for FFA localization)

No new benchmarks or datasets were introduced; the focus was on analyzing brain responses to existing categories of visual stimuli.

Key Insights and Ideas

  • Experience Shapes Cortical Selectivity: The most significant finding is that extensive visual experience with a specific category (like reading) can create a functionally selective region in the brain. This challenges the idea that such specialized areas must be hardwired by evolution.
  • A Dedicated Letter String Region: The study identified a consistent region in the left extrastriate cortex that responds selectively to letter strings (both words and consonant strings) compared to other visual categories like drawings, digits, and Chinese characters. This region is termed the "candidate letter string-selective region" (CLSSR).
  • Specificity Beyond Simple Features: The CLSSR's selectivity isn't just based on basic visual features. While it responds to consonant strings, its heightened response to Hebrew words in Hebrew readers suggests a more complex, experience-driven specialization.
  • Reading is a Learned Skill, Not Innate: The finding that Hebrew readers show a stronger response to Hebrew words than non-readers directly supports the idea that the brain adapts to learned skills like reading, creating specialized neural machinery for it.
  • High Resolution is Key: The authors emphasize that the small size of the CLSSR necessitates high-resolution fMRI to reliably detect and characterize its selectivity, explaining why previous studies might have yielded inconsistent results.

Key Results

  • A small region in the left extrastriate cortex (CLSSR) was identified in 85% of participants, showing selective responses to English words and consonant strings compared to line drawings, digit strings, and Chinese characters.
  • This CLSSR responded more strongly to Hebrew words in Hebrew readers than in non-Hebrew readers, demonstrating the impact of reading experience.
  • The response to Hebrew words in Hebrew readers was significantly greater than their response to English words, and also significantly greater than the response to Hebrew words in non-Hebrew readers.
  • The CLSSR did not show a significant difference in response between English words and consonant strings, suggesting its selectivity is for "letter strings" rather than strictly "words."
  • The CLSSR was found to be much smaller than the face-selective FFA, highlighting the need for high-resolution imaging.

Limitations and Future Work

  • Genetic Predisposition vs. Experience: While the study strongly supports the role of experience, the authors acknowledge that genetic factors might still predispose this region to develop letter string selectivity due to its connectivity. They cannot entirely rule out a genetic component in the location of the CLSSR.
  • Computational Role: The exact computations performed by the CLSSR and the nature of the representations it extracts remain an open question.
  • Developmental Trajectory: The study doesn't detail the developmental timeline of the CLSSR, nor does it explore its function before children learn to read or how it differs in dyslexic individuals.
  • Attention vs. Experience: While they attempted to control for attention, the authors note that distinguishing between attentional modulation and long-term experience effects can be challenging.
  • Future Directions: The authors suggest further research into the precise computations within the CLSSR, its developmental trajectory, and its role in different reading abilities (e.g., dyslexia). They also propose investigating how experience shapes other category-selective regions in the brain.

Summary updated

Abstract

How do category-selective regions arise in human extrastriate cortex? Visually presented words provide an ideal test of the role of experience: Although individuals have extensive experience with visual words, our species has only been reading for a few thousand years, a period not thought to be long enough for natural selection to produce a genetically specified mechanism dedicated to visual word recognition per se. Using relatively high-resolution functional magnetic resonance imaging (1.4 x 1.4 x 2-mm voxels), we identified a small region of extrastriate cortex in most participants that responds selectively to both visually presented words and consonant strings, compared with line drawings, digit strings, and Chinese characters. Critically, we show that this pattern of selectivity is dependent on experience with specific orthographies: The same region responds more strongly to Hebrew words in Hebrew readers than in nonreaders of Hebrew. These results indicate that extensive experience with a given visual category can produce strong selectivity for that category in discrete cortical regions.
BibTeX
@inproceedings{baker2007visual,
  author = {Chris I. Baker and Jia Liu and Lawrence L. Wald and Kenneth K. Kwong and Thomas Benner and Nancy Kanwisher},
  title = {{Visual word processing and experiential origins of functional selectivity in human extrastriate cortex}},
  booktitle = {Proceedings of the National Academy of Sciences},
  year = {2007},
  doi = {10.1073/pnas.0703300104},
}
Figure