The Dynamic Text effect/filter has just one line for entering text and keywords. It doesn't accept carriage returns to have keywords and text on different lines. This is not a limitation of the filter `dynamictext` from MLT. In fact, Shotcut has a much taller text field and does accept carriage returns, so users can put keywords on different lines in the video. See also [this bug report](https://bugs.kde.org/show_bug.cgi?id=505449).