Psychology of Learning

Module 08: Sports Psychology

Part 3: Theories of Motor Skills Learning

Looking Back

In Parts 1 & 2, we explored motor learning’s practical dimensions—how skills are acquired through feedback, refined through distributed practice, & transferred across tasks. We examined the Yerkes-Dodson Law & mental imagery. These principles describe what works in motor learning but don’t fully explain how motor learning works. How does feedback create lasting skill improvements? What changes in the brain during practice? How do separate movements become integrated into smooth sequences? Theoretical frameworks provide mechanistic explanations for motor learning phenomena.

Adams’s Two-Stage Theory: Perceptual Traces & Motor Traces

Jack Adams (1971) proposed an influential theory explaining how knowledge of results enables motor learning. His theory distinguishes two types of memory representations developed through practice: perceptual traces & motor traces. Understanding these concepts requires analogy to control system theory.

Control System Analogy: The Thermostat

Consider a home thermostat controlling temperature. You set the thermostat to 72 degrees—this is the reference input, the desired state. The thermostat continuously measures actual room temperature through sensors—this is feedback. When actual temperature differs from the reference input (72 degrees), the system generates an error signal. If the room is 68 degrees, the error is -4 degrees; the system activates heating. If the room is 75 degrees, the error is +3 degrees; the system activates cooling. The action system (furnace or air conditioner) works to minimize error, bringing actual state closer to the reference input.

Adams proposed that motor learning operates similarly, with two critical components corresponding to control system elements.

The Perceptual Trace: Your Internal Reference Standard

The perceptual trace corresponds to the reference input of control system theory. It represents the learner’s memory of what correct performance should feel like—the sensory feedback associated with successful movements. The perceptual trace is the internal standard against which performance is compared (Adams, 1971).

Think of a basketball free throw. Through repeated practice with knowledge of results, you develop a perceptual trace—a memory of what a successful free throw feels like kinesthetically. You remember the sensation of proper elbow position, the feeling of correct release timing, the proprioceptive feedback from muscles executing the movement correctly. This perceptual trace becomes your internal reference: “This is what success feels like.”

The perceptual trace develops gradually through repeated trials with KR. Early in learning, you have no perceptual trace—you don’t know what correct performance feels like because you haven’t experienced it reliably. Each successful trial, indicated by KR (“made it!”), allows you to encode the sensory feedback from that trial. Over many successful trials, these encoded memories coalesce into a stable perceptual trace representing correct performance.

The Motor Trace: Your Movement Production System

The motor trace refers to the workings of the action system of control system theory. It represents the memory of movements themselves—the motor commands & muscle activations that produce behavior. The motor trace is what generates movement (Adams, 1971).

Continuing the basketball example: The motor trace contains the motor program for executing the free throw—the sequence of muscle activations, the timing of movements, the coordination patterns. When you decide to shoot a free throw, the motor trace generates the appropriate motor commands sent to muscles.

The motor trace is relatively crude early in learning. Initial attempts produce variable, inconsistent movements because the motor trace hasn’t been precisely calibrated. Through practice, the motor trace becomes refined, producing increasingly consistent & accurate movements.

How the Two Traces Work Together

Adams’s theory proposes that motor learning proceeds in two stages, each emphasizing different traces:

Verbal-Motor Stage (Early Learning): Learners heavily depend on knowledge of results to develop the perceptual trace. KR tells you whether performance was correct, allowing you to encode sensory feedback from successful trials. During this stage, you need frequent KR because your perceptual trace is still forming—you can’t yet reliably judge whether performance felt right. The motor trace is also developing, producing increasingly consistent movements, but accuracy depends on external KR rather than internal evaluation.

Motor Stage (Later Learning): Once the perceptual trace is well-established, learners can perform accurately without KR. You’ve internalized what correct performance feels like, so you can detect & correct errors through internal feedback. You compare sensory feedback from each attempt (generated by the motor trace) against the perceptual trace (your reference standard). If feedback matches the perceptual trace, you know performance was good. If feedback deviates from the perceptual trace, you detect the error & adjust the motor trace for the next attempt.

The Role of KR Changes Across Learning

Adams’s theory makes a critical prediction: Knowledge of results is essential for developing the perceptual trace but becomes less important once the perceptual trace is formed. This explains findings we discussed in Part 1—constant KR produces dependency, while reduced KR promotes independence. Early learners need frequent KR to build their perceptual traces. Advanced learners need less KR because they’ve internalized the reference standard; they can self-evaluate using their well-developed perceptual traces.

This theory elegantly explains why beginners require constant feedback while experts perform reliably without external information. It also explains why removing KR suddenly from beginners impairs performance dramatically (no perceptual trace to guide self-correction) while removing KR from experts has minimal impact (established perceptual trace provides internal guidance). (Note: Adams’s theory sparked extensive research on stages of motor learning, optimal feedback schedules, & the transition from external to internal error detection—topics that continue to guide instructional design in motor skill training.)

Learning Movement Sequences: From Individual Actions to Integrated Wholes

Much of motor behavior involves executing sequences of movements in specific orders with precise timing. Consider activities like typing, playing piano, performing dance routines, executing gymnastics sequences, or completing surgical procedures. Each requires integrating multiple movements into smooth, coordinated sequences. How are these complex sequences learned & controlled?

Movement sequences are movements that must be performed in a fixed sequence & with correct timing. Some sequences are cyclical & repetitive, such as walking, swimming, or pedaling a bicycle. Other sequences are non-cyclical, such as typing a word, playing a musical phrase, or performing a gymnastics routine (Lashley, 1951).

Two competing theories attempt to explain how movement sequences are learned & controlled: the response chain approach & the motor program approach.

The Response Chain Approach: Sequential Associations

A response chain is a sequence of learned behaviors that must occur in a specific order, with a primary reinforcer delivered only after the final response. The behaviors in the chain stay in the correct sequence because each response produces a stimulus that serves as a discriminative stimulus (SD) for the next response (Skinner, 1938).

The response chain approach applies operant conditioning principles to movement sequences. Consider teaching a rat to press a lever, then pull a chain, then push a door to receive food. According to response chain theory:

Step 3: Push door → Food (primary reinforcer). The door-pushing response is directly reinforced.

Step 2: Pull chain → Door becomes accessible (conditioned reinforcer serving as SD for pushing). Chain-pulling is reinforced by door access & produces the stimulus (accessible door) that cues door-pushing.

Step 1: Press lever → Chain becomes accessible (conditioned reinforcer serving as SD for pulling). Lever-pressing is reinforced by chain access & produces the stimulus (accessible chain) that cues chain-pulling.

Each response in the chain produces stimuli (proprioceptive feedback, environmental changes) that serve as discriminative stimuli triggering the next response. The chain is held together by these stimulus-response-stimulus links. Reinforcement at the end of the chain maintains the entire sequence through secondary reinforcement—each response produces stimuli associated with eventual primary reinforcement.

Teaching Response Chains: Backward Chaining

The response chain approach predicts that sequences should be taught backward—starting with the final response (closest to primary reinforcement) & working backward. This is backward chaining, which we encountered briefly in Module 07. Teaching the last response first ensures that each added response produces a stimulus already associated with reinforcement, facilitating learning. Many animal trainers use backward chaining successfully to teach complex behavioral sequences.

Limitations of the Response Chain Approach

Despite some success, the response chain approach faces serious challenges when explaining rapid human movement sequences:

Timing Problem: Human reaction time to sensory stimuli is approximately 100-200 milliseconds. Yet skilled typists execute 10-15 keystrokes per second—one keystroke every 60-100 milliseconds. This means the next keystroke begins before sensory feedback from the previous keystroke could possibly be processed. If each keystroke required feedback from the previous keystroke to trigger it (as response chain theory suggests), typing would be impossible at observed speeds.

Deafferentation Studies: Research with animals shows that movement sequences can continue even when sensory feedback is eliminated. Surgically severing sensory nerves (deafferentation) removes proprioceptive & tactile feedback, yet practiced movement sequences often persist. If each response in a chain requires feedback from the previous response as a discriminative stimulus, eliminating feedback should destroy the sequence. But sequences often survive deafferentation, suggesting they don’t depend on continuous sensory feedback.

These limitations motivated an alternative theory: motor programs.

Motor Programs: Centrally Controlled Sequences

Motor programs are brain or spinal cord mechanisms, first proposed by Lashley (1951), that control a sequence of movements & do not rely on sensory feedback from one movement to initiate the next movement. Motor programs are pre-structured commands that run off automatically once initiated (Lashley, 1951).

Karl Lashley (1951) argued that rapid movement sequences cannot depend on sensory feedback loops because feedback is too slow. Instead, he proposed that the nervous system plans & organizes movement sequences in advance, creating motor programs that specify the entire sequence. Once triggered, the motor program executes automatically without requiring continuous sensory guidance.

Think of a motor program as a script or computer program. Just as a computer program contains all instructions needed to run a task without constant external input, a motor program contains all commands needed to execute a movement sequence. The program is loaded into the motor system, initiated, & then runs to completion based on pre-specified timing & sequencing.

Evidence for Motor Programs

Several lines of evidence support motor program theory:

Speed Evidence: The typing example provides compelling support. Reaction times are too slow for sensory feedback to control rapid keystroke sequences. The only plausible explanation is that entire words or phrases are programmed in advance as motor programs, with individual keystrokes executed according to pre-set timing rather than triggered by feedback from previous keystrokes. Skilled pianists executing rapid musical passages face identical timing constraints—feedback is too slow to control individual note sequences, implicating motor programs.

Deafferentation Evidence: When sensory feedback is surgically removed from animals’ limbs, many well-practiced movement sequences continue relatively normally. Cats with deafferented legs can still walk, though movements may be less precise or adaptable. Birds with deafferented wings can still fly. These findings demonstrate that sensory feedback, while useful for fine-tuning & adaptation, is not necessary for basic sequence execution. The sequences must be centrally programmed rather than dependent on peripheral feedback loops.

Reaction Time Studies: Research measuring reaction time for initiating movement sequences of different lengths provides crucial evidence. If sequences were chains (each response triggering the next), reaction time should be constant regardless of sequence length—you only need to initiate the first response. But research shows that reaction time increases with sequence length. Longer sequences require longer planning time before initiation, suggesting that the entire sequence is programmed before movement begins. This pre-programming takes more time for longer sequences, supporting motor program theory over response chain theory.

Error Pattern Evidence: When skilled performers make errors in rapid sequences, errors often preserve timing & sequencing structure. Typists might substitute one letter for another but maintain keystroke timing. Musicians might play wrong notes but preserve the rhythm. These error patterns suggest that timing & sequencing are controlled by an overarching program, with specific content details sometimes being incorrectly specified. (Note: Contemporary motor control research distinguishes between generalized motor programs specifying relative timing & force patterns versus specific parameters like overall speed or force levels—refinements that elaborate Lashley’s basic insight about centralized sequence control.)

Schmidt’s Schema Theory: Learning General Rules

Richard Schmidt (1975) extended motor program theory by addressing a critical question: How do people perform movements they’ve never practiced? A tennis player can hit serves to locations never previously attempted. A basketball player can make shots from distances never practiced. If motor programs specify exact movement sequences, how can novel movements be produced?

Schema theory proposes that by practicing different variations of the same response, people develop general rules (schemas) that allow them to perform responses they have never practiced before. Schemas are abstract representations capturing relationships between parameters rather than specific movement details (Schmidt, 1975).

According to schema theory, practice produces two types of schemas:

Recall Schema: A rule relating desired outcomes to motor program parameters. If you want the ball to travel a certain distance, the recall schema specifies how to set motor program parameters (force, timing, muscle activation patterns) to achieve that outcome. The recall schema allows you to scale movements appropriately for novel situations—throwing to a distance you’ve never thrown before, hitting a ball with a force you’ve never used.

Recognition Schema: A rule relating motor program parameters to expected sensory consequences. The recognition schema predicts what feedback should result from executing a particular motor program. This allows error detection—if actual feedback doesn’t match expected feedback, you know something went wrong & can adjust.

Variable practice—practicing the same general movement with different parameters (throwing to different distances, hitting balls at different speeds, shooting baskets from different locations)—develops richer schemas. Each practice variation provides data points relating parameters to outcomes, strengthening the schema. This explains transfer effects we discussed in Part 1—practicing variations builds general principles that apply to novel situations.

Schema theory elegantly explains how people generate novel movements, why variable practice enhances transfer, & how performers adapt movements to changing conditions. It represents a sophisticated evolution from simple motor program ideas to a more flexible, rule-based conception of motor control. (Note: Schema theory sparked decades of research on variability of practice, transfer, & the relative benefits of constant versus variable practice schedules—research that continues to inform coaching & instructional design across sports & skill domains.)

How This Material Relates to Sports Psychology

Theoretical frameworks of motor learning provide sports psychology with models for understanding how athletes acquire and refine skills. Adams’s two‑stage theory explains why beginners need frequent external feedback while experts rely on internal perceptual traces. This informs coaching progression—early emphasis on KR, later emphasis on self‑evaluation and autonomy. Sports psychologists use this to design feedback schedules that build independence and resilience in athletes.

The motor program theory and schema theory highlight the psychological processes underlying skill execution. Motor programs explain how athletes perform rapid sequences automatically, while schema theory accounts for adaptability in novel situations. For example, a basketball player can shoot from unfamiliar distances by applying generalized rules developed through variable practice. Sports psychologists encourage variable practice to strengthen schemas, enhancing transfer and adaptability under competitive conditions.

Response chain approaches and error pattern studies also tie directly to sports psychology. They explain why athletes sometimes repeat timing structures even when errors occur, and why backward chaining is effective in teaching complex routines (e.g., gymnastics or diving). By applying these theories, sports psychologists can anticipate learning plateaus, design drills that foster adaptability, and help athletes understand the cognitive mechanisms behind their performance. This theoretical grounding ensures that interventions are evidence‑based rather than intuitive.

Looking Forward

We’ve explored theoretical frameworks explaining motor learning. Adams’s two-stage theory distinguishes perceptual traces (internal reference standards) & motor traces (movement production mechanisms), explaining how knowledge of results builds internal standards enabling self-correction. The response chain approach treats sequences as stimulus-response links but cannot explain rapid sequences where feedback is too slow. Motor programs provide centrally organized movement plans that execute without continuous feedback. Schmidt’s schema theory proposes that variable practice develops general rules allowing novel movement generation. These theories explain phenomena from Parts 1 & 2 & provide comprehensive foundations for applying learning psychology to motor skill development.