We give specific recommendations on corpus construction, but we are mindful of the fact that it is not always possible to follow these recommendations.

We would therefore like to stress that you should treat them with some flexibility: if it is not feasible to follow a recommendation, then change it. To give you an idea of the variability, here are some examples of existing sketch corpora.

Feel free to explore how others have approached the task of constructing a sketch corpus.

• German sketch corpus: Urbanczik 2023, using the Leo corpus (Behrens 2006) and the Rigol corpus (Lieven & Stoll 2013)
• Inuktitut sketch corpus: Lee & Allen 2023
• Totoli sketch corpus: Bracks & Bardají i Farré 2023

References

_{Behrens, Heike. 2006. The input–output relationship in first language acquisition. Language and Cognitive Processes 21(1-3). 2–24.}

_{Bracks, Christoph & Maria Bardají i Farré. 2023. Broadening language documentation with child language data: First-hand experience from Tolitoli. Language Documentation and Conservation SP 28. 87-108.}

_{Lee, Hannah & Shanley E. M. Allen. 2023. An acquisition sketch of Inuktitut. Language Documentation and Conservation SP 28. 135-213}

_{Lieven, Elena & Sabine Stoll. 2013. Early communicative development in two cultures: A comparison of the communicative environments of children from two cultures. Human Development 56(3). 178–206.}

_{Urbanczik, Gianna. 2023. Exploring case marking in German first language acquisition using the acquisition sketch approach. Language Documentation and Conservation SP 28. 109-134.}