Drift as a Driver of Language Change: An Artificial Language Experiment

Area of Science:

• Linguistics
• Cognitive Science
• Psychology

Background:

• George Zipf's observation: more frequent words are older.
• Frequent words are less often replaced/regularized.
• Two hypotheses: stronger selection or stronger drift for frequent words.

Purpose of the Study:

• To experimentally test hypotheses on why frequent words change less.
• To investigate the roles of selection and drift in language change.

Main Methods:

• Participants learned a miniature language with varying noun frequencies.
• Experimental conditions manipulated drift and selection pressures.
• A model accounted for participant heterogeneity to measure regularization rates, selection, and drift.

Main Results:

• Drift alone sufficiently explains the higher regularization rate in low-frequency nouns.
• Quantified rates of noun regularization, selection strength, and drift strength.

Conclusions:

• Stochastic drift is a significant driver of language change.
• Experimental evidence supports drift's role over selection in explaining word frequency patterns.