With no explanation, label text_A→text_B with either "strengthener" or "weakener".
text_A: An older gentleman in brown gives a look of astonishment to the camera. The man is smiling
text_B: The man has tears in his eyes.
weakener.