1-splicing vs. 2-splicing: separating results

Splicing is a binary operation on strings defined by T. Head in 1987 which is inspired by the cut and paste phenomenon on DNA molecules [2]. Based on this operation, splicing systems, or H systems, were proposed as a generative device of languages: starting from a set of words, by repeated application of the splicing operation, new words may be generated. When we consider Păun’s definition, two types of H systems may be defined, according to the fact that one word or two words are produced as a result of the splicing operation, called 1-splicing and 2-splicing operation, respectively. While the power of splicing languages was investegated in comparison with classes of Chomsky hierarchy it wasn’t known if 1-splicing and 2-splicing languages represent the same class. In this paper we present non trivial classes of regular languages which are generated by using the 1-splicing operation (over a finite H system) but which cannot be generated by using the 2-splicing operation. Thus, we show that 1-splicing operation is more powerful than the 2-splicing operation.