"LANGUAGE CAN BOOST OTHERWISE UNSEEN

OBJECTS INTO VISUAL AWARENESS"

PRIYANKA BHATT, COGNITIVE
PSYCHOLOGY PAPER PRESENTATION
 INTRODUCTION
Can words change what we see, literally?

Traditional theories of perception argue that what we see is primarily driven by our visual input—
bottom-up processing. However, newer theories suggest that perception is not purely passive. Instead,
it may be influenced by higher-level cognitive processes, like language.
This study by Lupyan and Ward (2013) asks: Can hearing a word (like “zebra”) help us see something
we otherwise wouldn’t?
They test this using continuous flash suppression (CFS), a technique that renders images invisible to
the conscious mind, and explore whether verbal labels can “boost” such images into awareness. If
successful, it challenges the traditional view of perception and supports a more interactive model of
cognition—where language, memory, and vision influence one another.
 THEORETICAL BACKGROUND
Why this study matters
The Modular View (Fodor, 1983): Perception and cognition (including language) are separate
modules. Vision is impenetrable to cognitive influences.
Challenge to Modularity: Prior studies have shown that language seems to affect what we see,
but critics argue that these studies didn’t really change what people saw, only how they thought
about what they saw, or how they remembered it.

What makes Lupyan & Ward (2013) different?

Uses CFS, which suppresses conscious visual experience.
Looks at basic detection (not categorization or memory).
Tests whether top-down activation from language can help break through
suppression—literally helping us “see” what’s not otherwise visible.
 RESEARCH PROBLEM
Can words change what we see?
"How do verbal labels affect visual perception and awareness? Can language
fundamentally change what we see?"

This question challenges our understanding of perception. Does language

merely interpret what we see, or does it actively shape our visual experience? If
language can influence perception, it has profound implications for how we
understand the mind and the world around us
 HYPOTHESIS
“VALID VERBAL LABELS WILL INCREASE THE DETECTION OF
SUPPRESSED IMAGES.”
“INVALID LABELS WILL IMPAIR DETECTION.”

When information associated

plant sunflower with verbal labels matches
stimulus-driven activity,
language can provide a boost to
perception, propelling an
otherwise invisible image into
awareness.
 METHODOLOGY
Continuous Flash Suppression (CFS)
is a method used to make visual
stimuli invisible to conscious
awareness.

It works by showing high-contrast dynamic noise (like flashing patterns) to one eye, while
showing a low-contrast image (e.g., a drawing or photo) to the other.
The brain prioritizes the noisy image, suppressing the actual target image from conscious
perception for several seconds.
This allows researchers to test whether non-visual cues (like words) can still influence whether
we “see” the hidden image.
In this study, participants were randomly assigned to only one of the three experiments.
 METHODOLOGY
Undergraduate students (18-35) from the University Normal/Corrected Vision
of Wisconsin–Madison, primarily psychology
students. No History of Neurological Disorders

Participants received course credit for their Right-handed

participation.

Native English Language Speakers

Randomly assigned to 3 groups of 40 participants
each making up an experiment participants were not
exposed to all three experiments. Each person was
Total Sample Size: n = 120
assigned to just one of the three experimental
designs—either with simple line drawings, naturalistic Independent Trials
photographs, or geometric shapes. Laboratory-based Experiment
 EXPERIMENTAL DESIGN: A TRIO OF CONDITIONS
In each experiment, there were trials where an object was present and trials where it was absent.
This is a standard control to measure false alarms.

Participants had a simple task: to indicate whether they saw any image within the CFS display.

GROUPS
No-Label Baseline: Participants heard a neutral sound (white noise) or nothing at all.
This provides a baseline to compare the effects of valid and invalid labels.🔇
Valid Label: Before the CFS display, participants heard the correct name of the object
(e.g., hearing "zebra" before seeing a zebra).🗣️✔️
Invalid Label: Participants heard the incorrect name of the object (e.g., hearing
"pumpkin" before seeing a zebra).🗣️❌
EXPERIMENTAL DESIGN: A TRIO OF CONDITIONS
CUE → CFS DISPLAY → DETECTION RESPONSE
Design: 3 (Label Type) × 2 (Object Presence) within-subjects where participants
experienced:
Measured (DVs):
Label Type (IV1): Detection Accuracy (Hit rate, False alarms)
Valid Label (e.g., “zebra”) Response T ime (RT)
Sensitivity (d′)
Invalid Label (e.g., “pumpkin”)
No Label (white noise)

Object Presence (IV2):

Present vs. Absent
 MEASUREMENTS
Response T ime (RT):
T ime taken to respond after stimulus onset, measured in milliseconds (ms).
Accuracy:
Hit Rate: Correctly identifying object-present trials.
False Alarm Rate: Incorrectly reporting an object when none was shown.

Signal Detection Sensitivity (d′):

A statistical measure that combines hit rate and false alarm rate to assess how well participants
distinguish signal from noise, independent of response bias.

Eye Tracking (1000 Hz):

Monitored fixation stability and attention during trials, ensuring valid visual input.
 EXPERIMENTS
Experiment 1: Object detection

Stimuli: Simple grayscale line drawings of common objects.

Number of Stimuli: Limited set of 8 objects, each presented multiple times.

Valid Labels Boost Detection: Participants were significantly more likely to detect
an object when they heard the correct label beforehand. 📈
Invalid Labels Hinder Detection: Conversely, participants were less likely to detect
an object when they heard an incorrect label. 📉
Response Times Affected: Valid labels led to faster detection times, while invalid
labels led to slower detection times.
 EXPERIMENTS
Experiment 2: Naturalistic Stimuli

Replication with Complex Images: The researchers replicated the findings from
Experiment 1 using a larger set of more complex, naturalistic images.

Stimuli: Grayscale photographs of objects from a wider range of categories.

Number of Stimuli: 160 different images, each presented only once.

This was done to address potential concerns about participants learning to associate
specific labels with specific locations on the screen.

The same “boosting” and “hindering” effects were observed

Generalizability: This suggests that the effect of language on perception is not
limited to simple line drawings
 EXPERIMENTS
Experiment 3: Label-Shape Congruency Matters

Stimuli: Geometric shapes varying on a continuum from a square to a circle.

Labels: "Square," "Circle," or a neutral sound.

This experiment was designed to test whether the effect of labels depended on the visual similarity
between the label and the stimulus.

The effect of labels depended on the match between the label and the shape of the object.
Example: Hearing the label "square" helped participants detect more square-like shapes, while
hearing the label "circle" helped them detect more circle-like shapes.

Fine-Tuning of Perception: This suggests that language can fine-tune our perception, making us
more sensitive to features that are relevant to the label
 KEY FINDINGS
Lupyan and Ward found that valid labels increased object detection rates, while invalid labels
decreased them. This suggests that language can influence what we see, even when the visual
stimulus is suppressed from conscious awareness.

Experiment 1: Object Detection Valid Labels: ↑ Detection Rates (X% vs. No-Label, p < 0.05)
Invalid Labels: ↓ Detection Rates (Y% vs. No-Label, p < 0.05)
Response Times: Faster w/ Valid, Slower w/ Invalid (Z ms & W ms, p < 0.05)

Experiment 2: Naturalistic Stimuli Replication: Confirmed findings w/ complex images

Generalizability: Effect extends beyond simple shapes

Experiment 3: Shape Categorization Label-Shape Congruency: Labels influence perception

based on shape match
Square" labels → ↑ detection of square-like shapes
"Circle" labels → ↑ detection of circle-like shapes
KEY FINDINGS
IMPLICATIONS
Challenges the modular view of perception
Shows that language can influence visual
awareness—contrary to theories that treat
vision as cognitively impenetrable.
Supports interactive models of cognition
Findings align with top-down processing and
predictive coding theories, where perception is
shaped by expectations and prior knowledge.
 IMPLICATIONS
Demonstrates that language can 'boost' perception
Hearing the correct label makes participants more likely to detect suppressed
images, even when the visual signal is weak.

Extends to real-world contexts where Insights could inform:

User interface design (e.g., verbal prompts improving visual search)
Education (e.g., learning environments where labels guide attention)
Advertising & perception science (e.g., priming attention with product names)

Raises broader cognitive questions

How do beliefs, culture, and vocabulary shape not just interpretation—but our
raw perceptual experience?
 CRITICAL ANALYSIS
Strengths
Innovative Methodology (CFS): Effectively isolates language's influence.
Overall, Lupyan and Ward's study
Multiple Experiments: Enhances generalizability.
is a valuable contribution to our
Clear Theoretical Framework: Provides a strong foundation.
understanding of the relationship
Limitations between language and perception.
While there are still many
Unclear Neural Mechanisms: Further research needed.
unanswered questions, their
Stimulus Simplicity: Limited complexity of objects/shapes.
research provides a strong
Individual Differences: Not explored in the study. foundation for future
Laboratory Setting Limitation
FUTURE RESEARCH RECOMMENDATIONS
Explore Neural Mechanisms: Use fMRI or MEG to localize brain regions involved in label-induced
perceptual boosts (e.g., visual cortex, language areas).

Cross-Linguistic & Cultural Studies: Test whether effects vary across languages and cultural
vocabularies —does a word in one language boost perception differently in another?

Investigate Individual Differences: Examine how cognitive style, attentional capacity, or language
proficiency influence susceptibility to label-driven perception.
Vary Label Types & Emotional Valence: Compare effects of concrete vs. abstract words, positive vs.
negative labels—do emotionally charged words change what we see more strongly?

Real-World Applications:
Education: Do verbal cues improve attention in visually noisy classrooms?
UX Design: Can spoken prompts guide users toward key visual elements?
THANKS FOR LISTENING!
THE END!