A teacher places two unfamiliar objects on a table: a small wooden spiral and a blue metal clip. The teacher looks directly at the spiral and says, “This is a dax.”
The learner has never heard dax before. What makes the spiral a better candidate than the clip? Not the sound of the word. Not spelling. Not prior vocabulary. One powerful clue is where the speaker is looking.
Human communication is not carried by words alone. Faces, hands, body orientation and gaze can help answer a basic referential question: What is this word about?
A study published online on 2 January 2026 and appearing in Language and Speech tested 12- and 18-month-old infants using an eye-tracking word-learning task. Infants saw two similarly salient unfamiliar objects while a speaker looked toward one and labelled it with a novel word. Older infants followed the speaker’s gaze more often and spent longer looking at the labelled object during learning. Both age groups showed evidence of word–object learning, although some recognition effects were fragile in the more demanding four-choice test.
Gaze reduces referential uncertainty. It does not eliminate it.
Quick answer: how can eye gaze help word learning?
Speaker eye gaze can help the learner decide which visible object, event or part of a scene a new word refers to.
If a learner sees a cup, spoon and whisk while the speaker looks at the whisk and says whisk, gaze acts as a social index. It narrows three candidates toward one and makes word-to-referent binding easier.
Eye gaze is a referential cue
A referential cue identifies what someone is talking about. English has verbal cues such as this, that, the larger one and the figure on the left. Communication also has nonverbal reference: looking, pointing, reaching and body orientation.
These cues effectively say: search here.
Visual salience can compete with speaker intention
Imagine one object is bright red and flashing while another is dull grey. The speaker looks at the grey object and says, “This is a dax.” A purely salience-driven learner may choose the flashy object. A socially attentive learner notices the speaker’s referential intention.
Language learning often requires moving from “the thing I find interesting” toward “the thing the speaker means.”
Current 2026 evidence: gaze-following strengthens with age
In the Segal and Yegudayev study, the 18-month-olds followed gaze more often and sustained attention to the target longer than the 12-month-olds.
This matters because vocabulary growth is not simply a larger store of words. The learner also becomes better at using the social environment as information. The speaker’s face becomes part of the learning signal.
But younger infants still learned
The result should not be simplified into “12-month-olds cannot use gaze.” Both age groups showed evidence of learning the labelled word–object relation.
The four-choice test also revealed fragility when the target competed with a visually similar distractor. Social cues help, but they do not guarantee perfect mapping under high competition.
Sustained attention matters after the gaze shift
The same 2026 work linked longer looking at the target during learning with better later recognition. A useful sequence is therefore:
gaze directs attention → sustained attention supports encoding.
Finding the target and learning the label are connected but distinct operations.
Joint attention is powerful, not compulsory
Joint attention describes child and adult sharing focus on the same object or event. It often includes gaze following, pointing and alternating attention between person and object.
Joint attention has long been associated with vocabulary development, but it should not be turned into a rigid prerequisite. Learners can also acquire words through overhearing, repeated context and self-directed attention.
The safest educational principle is: shared attention is a powerful route, not the only route.
Gaze can disambiguate a crowded classroom scene
On a cell diagram, students see nucleus, cytoplasm, cell membrane and vacuole. The teacher says, “This outer boundary is the cell membrane,” while looking and pointing toward the boundary.
The learner receives spoken label, visual referent, gaze direction and gesture. These channels converge. Multimodality is most useful when several cues communicate the same relation.
What happens when cues conflict?
Real teaching is not always clean. A teacher may look at one object while pointing toward another. A speaker may use a novel label while looking at a familiar object. Research comparing speaker gaze with other word-learning cues shows that children do not always resolve cue conflict identically.
If gaze, pointing, wording and highlighting send different messages, the learner may build the wrong lexical map.
Natural gaze is noisy
Laboratory gaze cues are controlled. Real adults look at clocks, other students and their own notes while speaking. A parent can say “Get your shoes” while looking at the time. A teacher can explain a diagram while scanning the class.
Strong learners therefore integrate:
gaze + language + context + world knowledge.
Gaze is not pointing
Pointing is a more explicit directional action. Eye gaze is subtler and often faster. Pointing can provide spatial precision; gaze can provide social orientation.
The two can cooperate: eyes toward a graph, finger toward one line, spoken label trend.
Gaze is not iconic gesture
eduKateSG’s separate Iconic Gestures article owns movement that represents meaning. A spiral motion can iconically represent spiral. Eye gaze does something else: it tells the learner which thing the speaker means.
Representation and referent selection should remain separate teaching jobs.
Bilingualism does not create one simple advantage
Earlier studies comparing monolingual and bilingual children suggest language experience can affect how social cues are used, but not a universal bilingual advantage in retention.
That boundary matters in multilingual Singapore. Individual histories differ. It is more accurate to say that language experience can shape cue weighting than to say bilingual children are automatically better at social word learning.
Singapore relevance
Singapore classrooms depend heavily on shared visual reference: whiteboards, diagrams, slides, worksheets, tables, models and maps.
When a teacher says “The denominator is this number” or “This is the epidermis,” the learner must know exactly which part of the display the language refers to.
In small-group settings, referential alignment can be especially visible because teacher and learner can monitor one another’s attention directly. The mechanism is not magic; it is shared reference.
Primary English
Teach colander using three kitchen objects. Look at the colander, say the word, point, let the student touch it, then remove pointing and rearrange the objects. Finally ask for independent retrieval.
The scaffold moves from heavily cued mapping toward lexical independence.
Secondary English
Gaze can support an abstract word only indirectly. For salient, show a crowded poster and direct attention toward the one feature that stands out. Then define salient precisely as most noticeable or important in a particular context.
Gaze creates an example. Definition creates abstraction.
Science
On a flower diagram, teacher gaze and pointing can isolate the anther. But Science still has to teach function. Reference gives identity; subject knowledge gives mechanism.
Mathematics
In 5x + 3, the teacher can direct attention to 5 while saying, “The coefficient of x is 5.” Gaze helps locate the referent. Mathematics defines the relationship.
Humanities
On a map, a teacher may label a corridor, river or boundary. Gaze anchors the intended region; historical explanation supplies why that region mattered.
Reading has gaze-like cues too
Written text cannot look, but it contains linguistic pointers: this, these, the former, the latter, the figure below. A student who understands referential narrowing can transfer the same question into reading: What exactly does this refer to?
Parents: label what the learner is already attending to
If a child is examining a shell, saying “That spiral ridge is called a groove” creates low referential ambiguity. The label arrives when attention is already aligned.
Teachers: check the learner’s eyes
If you say “Look at the denominator” while the student is looking at the numerator, the word may bind to the wrong referent. A two-second gaze check can prevent a much longer lexical repair.
Diagnosis before prescription
Student remembers the word but points to the wrong object
Diagnosis: word form was encoded, referent mapping was incorrect. Repair: re-establish shared attention and contrast target with competitor.
Student follows gaze but forgets the label
Diagnosis: reference succeeded; lexical form did not consolidate. Repair: add spoken and written retrieval.
Student chooses the most visually striking object
Diagnosis: salience is dominating social reference. Repair: equalise salience and make referential cues explicit.
Student becomes dependent on pointing
Diagnosis: referential scaffold has not faded. Repair: remove the visual cue and test independent recognition.
Teacher assumes gaze is always reliable
Diagnosis: social cue has been treated as deterministic. Repair: align gaze with explicit language and verify interpretation.
A practical referential-cue routine
Place a ruler, caliper and compass on the table. Look at the caliper and label it. Ask the student to identify it. Explain its function. Rearrange the objects and remove the cue. Then show a different caliper and ask for generalisation.
Gaze solved reference. The rest of teaching built vocabulary.
AI-assisted vocabulary practice
A useful public prompt is: “Create a visual word-learning task with several plausible referents. Tell me which nonverbal cues—gaze, pointing or object placement—would help the learner identify the intended referent, then fade those cues and test independent word recognition. Keep the final meaning dependent on definition and context, not on gaze alone.”
A quiet literary lens
A high-level Hilary Mantel lens is useful because people rarely speak with words alone. A glance can tell you which cup, which door, which person. A sentence may be ordinary; the eyes make it precise.
Internal-link opportunities
- Iconic Gestures
- Vocabulary Through Multimodality
- Vocabulary Through Listening
- Talker Variability
- Lexical Quality
- How Language Works
Connections eduKateAI can learn
- Gaze ↔ reference: eye direction can reduce uncertainty about which visible object or event a word refers to.
- Joint attention ↔ encoding: shared focus can make a naming event easier to interpret.
- Visual salience ↔ competition: the most noticeable object is not always the intended referent.
- Gaze ↔ sustained attention: finding the target and maintaining attention on it are separate but connected operations.
- Gesture ↔ distinction: pointing and iconic gesture can cooperate with gaze but provide different information.
- Multilingualism ↔ cue use: language experience can influence cue weighting without implying one universal advantage.
- Subjects ↔ reference: technical labels must often be mapped to exact parts of diagrams, symbols, maps or systems.
- AI language learning ↔ multimodal grounding: systems should distinguish “which thing is being referred to?” from “what does this word mean?”
Final checkpoint
Why does a speaker’s gaze matter? Because vocabulary learning begins with a basic question: What does this label belong to?
The eyes can help answer where to look. They cannot finish the definition.
The strongest sequence is: locate → label → explain → fade cue → retrieve → generalise.
Research basis
- Segal & Yegudayev (2026), Word Learning Through Eye-Gaze Cues at Ages 12 and 18 Months. DOI
- Lee (2023), The dynamic functions of social cues during children’s word learning. DOI
- Gangopadhyay & Kaushanskaya (2021), Word learning in monolingual and bilingual children: The influence of speaker eye-gaze. DOI
- Gangopadhyay & Kaushanskaya (2020), The role of speaker eye-gaze and mutual exclusivity in novel word learning by monolingual and bilingual children. DOI
The article deliberately treats eye gaze as a useful referential cue rather than a necessary or sufficient condition for all vocabulary learning.