It would be useful to do the reverse of the great speech to text/captioning feature. If one could start with a CSV or Excel file that has start and end time column values compatible with Premiere nomenclature (e.g., 00;00;12;24), and a third column with the corresponding verbiage text for each segment, then allow import of that into a sequence in which an artificial voiceover with captioning is created at those in and out points without using a mike. Value: for making a rough script without going through initial voice over creation (with all of its problems, hems and haws, ums, etc.) that can be the stepping stone for a subsequent human-voiceover with a refined script.