Words and possible words in early language acquisition

Area of Science:

• Cognitive Science
• Developmental Psychology
• Linguistics

Background:

• Infants must identify words and their possible forms from continuous speech.
• Word segmentation and identification are crucial for language acquisition.
• The interplay between structural and statistical information processing in infants is not fully understood.

Purpose of the Study:

• To investigate the developmental order of infants' sensitivity to possible words versus actual words.
• To determine how segmentation cues affect infants' ability to extract word structure and statistical patterns.
• To compare the abilities of 12- and 18-month-old infants in processing artificial languages.

Main Methods:

• Exposing infants (12 and 18 months) to an artificial language with conflicting statistical and structural information.
• Varying the presence or absence of segmentation cues in the familiarization stream.
• Analyzing infants' ability to identify possible words and actual words under different conditions.

Main Results:

• Eighteen-month-olds identified possible words with segmentation cues but struggled with continuous streams.
• Eighteen-month-olds identified actual words from continuous streams using statistical relationships.
• Twelve-month-olds identified possible words with segmentation but failed to extract statistical patterns from continuous streams.

Conclusions:

• Sensitivity to word structure (possible words) is more primitive than statistical learning for word identification.
• Statistical learning for language acquisition becomes effective later in development.
• Structure-extraction mechanisms independent of extensive statistical sampling play a significant role in early language acquisition.