The aspectual periphrasis 'estar' plus gerund in Child Spanish. In between previous knowledge and the influence of the input

In this study, we undertake a corpus-based analysis of the verbal periphrasis construction estar 'be' +gerund in Spanish. In the literature, there is evidence to support that this periphrasis is relatively late in acquisition (Jackson-Maldonado & Maldonado, 2002) and that it appears rapidly, with no apparent gradual development (Ponce, 2008). We tested the following hypotheses: 1) The acquisition of this periphrasis is influenced by Child Directed Speech frequency; 2) The acquisition of the verbal estar 'be' periphrasis is relatively late but seemingly sudden since it is supported by already acquired locative and adjectival estar 'be' constructions. To test the first hypothesis we measured both token and cumulative type frequency of use of the estar 'be'+ gerund verbal periphrases of all three caretakers against the children's periphrases, using a Pearson correlation analysis. To test the second hypothesis we applied a lineal and exponential regression method, making use of the cumulative type frequency of each estar 'be' construction. Type frequency is strongly correlated only when Child Directed Speech is correlated with dense data. However, the children's periphrases are strongly correlated with the Child Directed Speech when using cumulative type frequency (once per month). On the other hand, the prior locative construction seems to support the acquisition of this aspectual periphrasis.