While most existing gestural interfaces focus on the static posture or the dynamic action of the hand, few have investigated the feasibility of using the forces that are exerted while performing gestures. Using the photoplethysmogram (PPG) sensor of off-the-shelf smartwatches, we show that, it is possible to recognize the force of a gesture as an independent channel of input. Based on a user study with 12 participants, we found that users were able to reliably produce two levels of force across several types of common gestures. We demonstrate a few interaction scenarios where the force is either used as a standalone input or to complement existing input modalities.