Abstract: In some languages, all the sentences must include or exclude some specific substrings. Sometimes, the substrings must be replaced. It is difficult for a Turing machine to accept such languages. Traditional method is based on the storage technique. In this paper, we present a new method, the scanning substring techonology. In the method, the specific substring is treated as a single unit. Multiple status transition functions of the Turing machine are combined to scan multiple characters at one time, but the read/write head only moves one unit. Through some examples, the paper shows that the Turing machine which scans multiple characters is equivalent to a Turing machine scanning one character.