2 From two words to three

Bowerman (1975) also followed the same children as they went on to utter their first three-word utterances (see Table 2).

Interestingly, Bowerman notes how there are virtually no new construction types in this three-word phase, rather these utterances are best characterised as combinations of two-word utterances blended together to express new relationships or concepts. For example, the agent-action and action-object types that were separate in the two-word phrase have been cut-and-pasted together to form an agent-action-object category in the three-word phase.

During this phase, it is sometimes apparent that children want to communicate an idea that is more sophisticated than their grammar or cognition can package. One way in which children work around this problem is by sequentially laying out smaller fragments of what would be a larger adult utterance, referred to as replacement sequences (see Table 3). For example, ‘dog there … doggie sitting’ instead of the more integrated and blended form ‘that dog is sitting there’.

The increasing number of three-word utterances in children’s speech is accompanied by more sophistication in the way that they use two-word utterances as well. The overall developmental picture at this stage is one of gradual progress, where children consolidate the gains they have made with simpler language and engage with incremental exploration of more complex forms.