Motor control is essential in daily life as well as in sports activities. It involves commands from central and peripheral neural systems to muscles and joints and may be achieved by open-loop and/or closed-loop systems (for reviews, see Magill and Anderson 2007). Both types of systems include a movement control center, movement commands, and movement effectors. In open-loop control system, all information is contained in the commands sent from the control center to the effectors and output activity cannot be changed once the commands have been sent. In contrast, in a closed-loop control system, feedback containing both exteroceptive and proprioceptive information is sent to the movement control center. The control center can modify the commands based on the feedback. The updated commands are then sent to the effectors.

There are two main differences between open- and closed-loop control systems. First, feedback is included in the closed-loop control system sent from the effector to the movement control center, but not in the open-loop control system; second, in the closed-loop control system, feedback information exists in updated commands sent from the movement control center to the movement effectors. Basically, whether the planned movement relies on an open-loop or the closed-loop system depends on duration of the movement. Ballistic movements, for example, are too rapid for the possibility of modification by proprioceptive feedback.

Focused on simple, self-paced, limb positioning movements, Adams formulated his closed-loop theory (Adams 1971). In Adams’ theory, movement is initiated by a processed named memory trace, which contains the necessary information about the motor commands. Once all necessary information is collected in the memory trace, commands will be sent to initiate the movement without involvement of any feedback. The performance is then evaluated with respect to a reference called perceptual trace which is derived from memories of earlier movements. This process compares feedback information about the actual movement with the goal movement. If any discrepancy is detected, motor commands will be modified and sent to the effectors until a match is achieved upon which the movement will be stopped.

Schmidt (1975) pointed out that Adams’ theory has difficulty explaining two problems. The storage problem concerns the limited memory storage of various movements; it would be very difficult to store representations of all possible movements. The novelty problem relates to the difficulty of accounting for the production of novel movements, which are not yet stored in the perceptual or memory trace. To solve this inconsistency, Schmidt proposed his schema theory for discrete actions. A schema is defined as a set of rules which act as a basis for decisions. A schema is formed by related information gathered from previous experiences. For example, after seeing lots of cars in real life as well as on TV and in books, you will likely develop a set of rules which form your concept of “car”. This set of rules constitutes a schema. Using this schema, one can recognize if an object is a car, even if they have never seen this specific one before.

According to Schmidt’s schema theory, only a limited number of generalized motor programs (GMP) concerning fixed movement structures are stored in memory. The performer has to apply certain movement parameters to an abstract GMP before a specific movement can be executed. This process is governed by the motor response schema, which has two components. One component is the recall schema, which has the responsibility of applying specified response characteristics to a certain GMP and executing a particular action. The second component is the recognition schema, which monitors performance and modifies movement commands based on feedback.

According to Schmidt’s schema theory, movement features may be fixed or variable. Fixed features are held to be associated with invariant spatiotemporal motor patterns including sequencing of actions and relative timing. In contrast, examples of variable parameters include overall duration and overall force. It should be noted that some studies (Shea and Wulf 2005; Leuthold and Jentzsch 2011; plus many others) refer to relative force as a GMP, however, Schmidt (2003) has pointed out that relative force is not invariant. Rather, relative force is a parameter that scales muscles. Variable features allow the adaptation of individual movements to fit the requirements of any given situation.

A typical example to illustrate motor schema involves each person’s written signature, which remains relatively constant whether it is large or small, or written by hand or foot (Schmidt and Lee 2013). The spatial and temporal characteristics of a written signature are GMPs, whereas the size of the handwriting and the effector used are parameters that need to be assembled.

Previous studies have shown that some factors that are beneficial for the learning of GMPs can have negative effects on parameterization learning. This supports the independence of the two processes (Wulf and Schmidt 1989; Wulf et al. 1993, 1994; Whitacre and Shea 2000; Shea and Wulf 2005). For example, Wulf and Lee (1993) asked participants to sequentially press four buttons at a fixed ratio of intervals (relative timing, i.e., 1: 2: 1.5) in different goal movement time (overall timing, i.e., task A: 200–400–300 ms; task B: 250–500–450 ms; task C: 300–600–450 ms, respectively.). They separated overall errors by calculating differences between actual and target movement time, and differences between actual and target intervals. The results suggested that learning of relative timing benefited from random practice, whereas blocked practice improved the learning of absolute timing. The dissociation of errors in relative and absolute timing agrees with the concept that GMP and parameterization learning are independent.

Schmidt and Lee (2013) assert that following stimulus identification, the GMP is retrieved from long-term memory in the response selection stage. Parameterization is held to occur during the following motor programming stage. However, there is conflicting evidence regarding the ordering of GMP and parameterization.