Formulaic Sequences are multi-word prefabricated§phrases of either literal (e.g. good morning) or§non-literal (e.g. kick the bucket) reading. A§property of such sequences is variation. Formulaic§sequences can be subject to varying degrees of§lexical substitutions, grammatical variations, and§insertions. This book investigates the boundaries of§variation: the limits of lexical substitution and§insertions for formulaic sequences, i.e. how much§variation can occur before the sequence stops being§fixed and becomes context-dependent. The boundaries§between one formulaic sequence and another and the§boundaries between a literal and non-literal reading§are also explored. The Longman Idioms Dictionary§(1998) and the British National Corpus (BNC) provided§the data for the research, and the technique and§analysis used contribute to lexicography and inform§cognitive models of storing and organizing language. ukryj opis