In 1822 French philosopher, political economist and printer Pierre Leroux, while aprenticed to the Didots, issued a 16-page pamphlet entitled Nouveau procédé typographique qui réunit les avantages de l'imprimerie mobile et du stéréotypage. In this pamphlet Leroux speculated on the development of a composing machine that would combine the process of polyamatype invented by Henri Didot in 1817, with the ability to write documents. Didot's process of polyamatype involved casting 140 letters into a stereotype plate at one time rather than typesetting each of these characters individually and then casting them into a stereotype plate. The machine that Leroux wanted to build would combine the writing directly with the typesetting and stereotype plate-making, giving the writer control over the publishing process. Like other inventors Leroux called his machine a "pianotype". However, Leroux mentioned few details as to how his machine would actually work.
Was it a coincidence that Leroux published his pamphlet the same year that William Church patented the first machine for type composition? It is certainly unlikely that Leroux, in France, could have read Church's English patent very quickly, but Leroux might have read a reference to Church's patent in a magazine or newspaper.
In 1845 when various mechanical typesetting machines such as those by Young and Delcambre, Rosenborg, and Gaubert's had appeared on the scene, and were being used to a certain extent, Leroux reprinted his pamphlet in his journal, La revue indépendante, Vol. 6, pp. 264-291 in an article entitled D'une nouvelle typographie. In this he reminded his audience of his original concept for an authorial composing machine.
Tresch, The Romantic Machine: Utopian Science and Technology after Napoleon (2017). Chapter 8: Leroux's "Pianotype: The Organogenesis of Humanity" pp. 223-225.