Limitation = contradictory research shows development may occur earlier than Martin and Halverston suggest.
Longitudinal study (2009) obtained data about children’s language from 9 to 21 months.
Results = children tended to label themselves ‘boy’ or ‘girl’ at around 19 months, (when first learn how to start communicating. )
This shows that gender labelling may be possible way earlier than Martin and Halverston suggest but children are not able to communicate it.
Therefore this theory has limited explanatory power, and generalisability as research shows contradictory results.