CHAPTER 8. Summary, discussion and conclusion 133 and complexity and equally e ective pertaining to accuracy. In light of the way the approaches di er regarding the process of form-use-meaning mapping, these results do not come as a surprise. Indeed, an SB approach is cognitively more demanding than a DUB approach, preventing SB students to develop uency and complexity as e ectively as DUB students. is section has tried to explain the e ectiveness of a DUB program regarding the development of uency and complexity over an SB program in terms of competition, but this leaves the question of the e ect of the programs on accuracy. An SB program focuses on written skills (See chapter 4 on reading skills and chapter 5 on writing skills), and SB students receive a great deal of explicit instruction on accuracy. In contrast, DUB students do not receive a great deal of feedback on morphological accuracy to reduce anxiety and facilitate speaking (See chapter 8 on speaking). Yet, both programs appear to be equally e ective in facilitating accuracy. is might be explained by the learning process with a much higher degree of repetition, facilitating entrenchment and automatization, and the nature of L2 exposure, which is more authentic in the DUB condition. is will be discussed in the next section. ENTRENCHMENT AND AUTOMATIZATION e main di erence between the programs is that DUB students are predominantly exposed to oral and meaningful, rather authentic French, while SB students are predominantly exposed to written textbook French, which is usually less meaningful and less authentic. Vandeweerd and Keijzer (2018) compared input from L2 French textbooks with input from an oral corpus and concluded that the relative amount and the frequency of authentic input in the oral corpus was signi cantly higher than in the textbook corpus. In chapter 6 it was argued that frequency of exposure to authentic language in a DUB program favors L2 development, as DUB students were found to outperform SB students with regard to producing longer, lexically based chunks. ese prefabricated sequences are consolidated in the learner’s brain by input repetition and by output repetition. According to Wray and Perkins (2000), they contribute to minimizing the e ects of a limited memory as these sequences allow short-cuts in processing: A veword prefabricated sequence is treated as one linguistic item (Gustafsson & Verspoor, 2017) and will be equally demanding to retain as one single word if this sequence’s form-meaning mapping is deeply entrenched and consolidated in the learner’s brain by repeated exposure and use. In a DUB approach, repeated exposure to and use of authentic language are extremely important, which means that form-use-meaning mappings of individual words and larger prefabricated sequences have a greater chance of being consolidated in the learner’s brain to a degree that automatization is reached. In an SB approach, only the shorter, more grammatically based chunks may be expected to reach this level of automatization.