Dexterous object manipulation with the freedom of choice of contact points has primarily been shown to elicit a feedback-dependent force control as opposed to the classically studied memory representation-based force control for constrained grasping. We probed this dichotomy in sensorimotor integration by introducing uncertainty about the grasp type during movement preparation within the context of motor planning. An inverted T-shaped instrumented manipulandum was presented to 24 participants, and they were asked to reach, grasp, and lift it while minimising the peak roll. On certain catch trials within a constrained block, with the participants unaware, the target cues for finger placement were extinguished at movement onset, essentially changing a constrained grasp to an unconstrained grasp. The role of the catch trials was to dissociate the information in the planning phase from the information in the execution phase. For the rest of the trials within this block, called the test trials, visual feedback of contact cues persisted throughout the task. Based on the evidence of Bayesian processes in sensorimotor processing and learning in humans, we hypothesised that the kinetic signatures of the initial uncertainty would be modulated by the preceding stimulus within the test trials. Our findings suggest that the kinetic measures were not statistically different between the test trials preceded by test trials and the test trials preceded by catch trials. This result points towards the preparatory states being non-dynamic in that they were invariant to the structure of the uncertainty and the prior distribution of the stimuli.