40 Adelbert Denaux mentation) approach to style, on the other, is broadly a difference between interactive and determinist theories respectively. […] … to define style as a series of socially or topically irrelevant redactional changes”36. Hence, according to Henderson, one can distinguish between three methodological approaches to style definition: on the one hand a classical, rhetorical and sociolinguistic approach, on the other hand a redaction-critical and neo-rhetorical approach (that looks for the argumentation of the text). Methodologically, Henderson’s view on style implies an adapted vocabulary. Concerning the Didache, he starts from a multilingual context and introduces the concepts of Code-switching, i.e. “alternations of linguistic varieties within the same conversation”37 and “style-switching”, i.e. “which goes beyond the style-switching of monolinguals and allow[s] individuals a flexibility of expression that could not be obtained in a single system”38. As is well known, Luke’s Greek is characterised by a remarkable adaptability or versatility: his preface (Lk 1,1-4) is modelled on classical patterns; the infancy story (Lk 1,5-2,52) is written in de highly semitizised Greek; the subsequent chapter (Lk 3) he seems to use a good literary Koine Greek39. Jonathan W. Watt has used the model of code-switching, developed in modern socio-linguistic studies, to explain this phenomenon in Luke-Acts. He states that: “Semitic style, or register, does not appear consistently throughout Lk and Ac, and it has been impossible to ascertain with certainty which Semitic features Luke produced, created, borrowed from other speakers, or retained from his sources. Yet their presence in Lk and Ac gives the overall effect of vividly creating a Jewish orientation for many of the accounts. The Semitisms in this ancient document function quite like the code- switching observed in modern bilingual situations”40. He thinks that the increased frequency of Semitisms in many sections of Luke-Acts may well correlate with I.H. Henderson, Style-Switching, 196. 36 I.H. Henderson, Style-Switching, 199. 37 I.H. Henderson, Style-Switching, 199. More particularly, he points to two categories of 38 “style-switching” in Did: Greek/Aramaic (Hebrew) CS, and quotation and “non-quotation”. Moreover, he refers to the variation between “passages-tu” and “passages-vous”, that J.-P. Audet, La Didachè (Paris 1958) has distinguished and interpreted from a redaction-critical point of view. This has already been remarked by A. Plummer, The Gospel According to St. Luke 39 (ICC; Edinburgh 1896; 1922 [= 1964]) xlix: “The author of the Third Gospel and of the Acts is the most versatile of all N.T. writers. He can be as Hebraistic as the LXX, and as free from Hebraisms as Plutarch. And, in the main, whether intentionally or not, he is Hebraistic in describing Hebrew society, and Greek in describing Greek society”. J.M. Watt, Code-Switching in Luke and Acts (Berkeley Insights in Linguistics and 40 Semiotics, 31; New York etc. 1997) 92.