You can also use indices ( see the section on Indices) in "perldoc Tk::Text". Essentially, you find the start indice of your word, count the number of characters in the word, delete ($indice + $n chars), the insert at the same $indice.

If you show a working code example, we can probably show you how.