The present paper examines the development of the variation between a marked and an unmarked infinitival complement clause in three types of cleft constructions in 20th century English. Data from corpora of written and spoken British (BrE) and American English (AmE) evidence a significantly divergent development of these clefts types in speaking when compared to writing. The written corpora show a steady increase in the frequency of clefts, and a decrease of the to-infinitive paired with an increase of the bare infinitive, thus a reversal of preferences in both varieties in all three types of clefts. This erosion of to as an (optional) grammatical marker leads to a higher degree of syntactic integration and can be interpreted as the next step in the grammaticalization process of clefts in written English. The spoken corpora also show a significant increase of clefts in spoken BrE in the last quarter of 20th century. While the infinitive marker is still persistent in spoken BrE, especially in wh- and th-clefts (less so in all-clefts), it is virtually absent in spoken AmE. More importantly, though, the majority of wh-clefts in the spoken material are difficult to classify in terms of a canonical, bi-clausal structure because they show a lack of syntactic integration to varying degrees, the only link between the wh-clause and the focus being their co-reference relation. This suggests a different development of the construction in spoken English. Wh-clefts fulfil specific metapragmatic functions in terms of managing temporality in interaction, and have recently been analysed as projector constructions that foreshadow upcoming discourse and are structurally not fixed, but interactively produced, thus emergent and open. Thus, the paper argues that wh-clefts in spoken English represent a case of pragmaticalization, i.e. the grammaticalization of discourse functions, in which the initial wh-clause develops into a topic marker or discourse-marker-like phrase.