Osiurak and Reynaud (O&R) claim that children do not possess the technological expertise required to innovate new solutions to problems and conclude it is debatable whether children are a good “methodological choice” to examine cumulative technological culture (CTC) (section 3.4). Indeed, children do struggle with tasks requiring creative problem-solving and we agree that sufficient technical-reasoning is required for innovation. However, the paper ignores a growing body of research that suggests some early innovative capacities and does not give adequate discussion to the early development of technical-reasoning skills. Indeed, based on recent developmental evidence, we argue that young children display some aspects of creative problem solving under limited conditions. Understanding these constraints on innovation in early childhood is key to understanding what is developing.

O&R argue based on four developmental studies (Beck et al. Reference Beck, Cutting, Apperly, Demery, Iliffe, Rishi and Chappell2014; Cutting et al. Reference Cutting, Apperly, Chappell and Beck2014; Reindle & Tennie Reference Reindle and Tennie2018; Reindle et al. Reference Reindle, Apperly, Beck and Tennie2017) that young children are poor innovators. It is true that young children's innovation is limited when they have to innovate over a short period of time (Beck et al. Reference Beck, Cutting, Apperly, Demery, Iliffe, Rishi and Chappell2014; Cutting et al. Reference Cutting, Apperly, Chappell and Beck2014) and with limited materials (e.g., a pipecleaner and string, or water and cup) (Beck et al. Reference Beck, Cutting, Apperly, Demery, Iliffe, Rishi and Chappell2014; Cutting et al. Reference Cutting, Apperly, Chappell and Beck2014; Reference Cutting, Apperly, Chappell and Beck2019; Ebel et al. Reference Ebel, Hanus and Call2019). However, young children can innovate new and effective solutions when working in small groups (McGuigan et al. Reference McGuigan, Burdett, Burgess, Dean, Lucas, Vale and Whiten2017), when they have prior experience with the task (Whalley et al. Reference Whalley, Cutting and Beck2017), when the task is open-ended and allows them to use multiple manufacturing methods (i.e., reshaping, adding, subtracting, and detaching) (McGuigan et al. Reference McGuigan, Burdett, Burgess, Dean, Lucas, Vale and Whiten2017; Voigt et al. Reference Voigt, Pauen and Bechtel-Kuehne2019), and when they have plenty of time (McGuigan et al. Reference McGuigan, Burdett, Burgess, Dean, Lucas, Vale and Whiten2017; Voigt et al. Reference Voigt, Pauen and Bechtel-Kuehne2019). In sum, young children appear to be able to explore their way to a solution but seem restricted in their ability to come up with the “right” solution in tasks that are more constrained both in terms of time, materials, and manufacturing methods (e.g., Beck et al. Reference Beck, Cutting, Apperly, Demery, Iliffe, Rishi and Chappell2014).

Intriguingly, a similar pattern has been observed when examining the development of children's hypothesis testing. When faced with a surprising event or with surprising data, young children deploy sophisticated exploration and search strategies, make appropriate inferences, and test these hypotheses (e.g., Gopnik Reference Gopnik2012; Gopnik et al. Reference Gopnik, Griffiths and Lucas2015; Reference Gopnik, O'Grady, Lucas, Griffiths, Wente, Bridgers, Aboody, Fung and Dahl2017). However, children struggle until middle childhood (and even adulthood in some contexts) to design controlled experiments that isolate causal factors (Chen & Klahr Reference Chen and Klahr1999). Explicitly testing a hypothesis and solving a specific technical problem are analogous in important ways and children seem to solve both tasks around the same time. Around 8-years-old, their problem solving in both contexts is more flexible and targeted and less reliant on imitation and exploration (Chen & Klahr Reference Chen and Klahr1999; Carr et al. Reference Carr, Kendal and Flynn2016; Lucas et al. Reference Lucas, Burdett, Burgess, McGuigan, Wood, Harris and Whiten2017). Given the cognitive overlap between designing an experiment and developing an innovative solution to a technical problem, the fact that scientific problem solving and innovation follow similar developmental trajectories suggests that domain-general developments (in addition to domain-specific knowledge) may play an important role in constraining innovation in childhood.

Some domain-general factors presumed to increase technological reasoning can be tentatively ruled out. On more constrained tasks (like the hook task) executive functioning (Chappell et al. Reference Chappell, Cutting, Apperly and Beck2013) including inhibitory demands, working memory, attentional flexibility (Beck et al. Reference Beck, Williams, Cutting, Apperly and Chappell2016), and divergent thinking (Beck et al. Reference Beck, Williams, Cutting, Apperly and Chappell2016) are not associated with innovation success rates. By implication, young children are not failing to innovate because of limits in their abilities to process information. Instead, their ability to innovate may be constrained by their ability to make connections between their prior knowledge and current tasks’ constraints (analogical reasoning, e.g., Gentner et al. Reference Gentner, Levine, Ping, Isaia, Dhillon, Bradley and Honke2016), by their ability to consider how different steps could be taken to solve a problem (advanced planning, Tecwyn et al. Reference Tecwyn, Thorpe and Chappell2014), and by improvements in children's metacognition – their ability to represent their own technical skills. This latter skill may be particularly important in allowing children to engage in more targeted forms of innovation and thus may allow children to not only explore their way to innovation but also to direct their way to innovation (see Carr et al. Reference Carr, Kendal and Flynn2016 for a similar proposal).

In conclusion, we agree with O&R that a suite of non-social cognitive factors contribute to technological reasoning and innovative thinking. We think that more work into understanding the development of these cognitive factors in children is promising. Specifically, we propose that further work examines the development of cognitive factors in both open-ended and constrained tasks. The cognitive skills required for either task may reveal multiple developmental pathways to innovation, such as via exploration or through a more directive, analogical approach.