The Motor Equivalence Experiment (Swimming Rats) – D.A. Macfarlane

In the history of experimental psychology, few empirical dilemmas have proven as fundamentally disruptive to mechanistic orthodoxies as the problem of motor equivalence: the capacity of a living organism to accomplish a single, invariant behavioral objective using radically disparate bodily effectors and distinct muscular configurations. During the opening decades of the twentieth century, the ascendant behaviorist movement sought to exorcise all unobservable mentalistic concepts from the discipline. Led by figures such as John B. Watson, the dominant theoretical framework conceptualized complex mammalian actions as rigid, concatenated sequences of peripheral reflexes. Within this strict peripheralist paradigm, learning was understood not as the mental acquisition of an environmental layout or spatial map, but as the progressive stamping-in of specific neuromuscular bonds. Navigation through a complex maze was regarded as a linear chain of kinesthetic associations, wherein each mechanical contraction of a muscle fiber and each joint articulation served as the physical stimulus triggering the next contraction in a deterministic cascade.

This mechanistic worldview faced a profound theoretical challenge from the cognitive counter-revolution nascent within the laboratories of the University of California, Berkeley. Under the intellectual leadership of Edward C. Tolman, a divergent philosophy of purposive behaviorism emerged, asserting that organisms do not merely acquire molecular chains of muscle twitches, but instead construct holistic, molar cognitive representations of their environments. Into this fierce theoretical conflict stepped Donald A. Macfarlane, whose seminal 1930 doctoral research, published under the title The Role of Kinesthesis in Maze Learning, directly tested the foundational assumptions of peripheral kinesthetic chaining. Macfarlane devised an ingenious, conceptually brilliant experimental apparatus: an intricate maze that could be navigated by terrestrial locomotion on a dry floor, and then, without altering its topological layout, rapidly flooded with water to force the experimental subjects to swim.

The logic of Macfarlane’s intervention was uncompromising. If a rat learns a maze exclusively as a string of specific leg-muscle movements, tactile footfalls, and peripheral proprioceptive signals, then suddenly submerging the maze floor in deep water should theoretically annihilate the animal’s navigational competence. Swimming engages entirely different physiological motor systems, requires different muscular coordinations, alters head position and visual angles, and completely abolishes the terrestrial proprioceptive feedback of walking. Yet, when Macfarlane released his experienced runners into the submerged labyrinth, the animals executed the correct spatial trajectory with negligible errors, swimming directly through the correct channels to the goal box. This decisive finding established the reality of motor equivalence, dealt a fatal blow to crude stimulus-response peripheralism, and provided early empirical validation for the concept of cognitive maps—an insight that continues to shape contemporary neuroscience, robotics, and cognitive psychology.

1. Introduction to D.A. Macfarlane and Motor Equivalence

1.1 Defining Motor Equivalence in Experimental Psychology

In experimental psychology, cognitive science, and motor physiology, the concept of motor equivalence designates the capacity of an organism to achieve a unified, invariant behavioral outcome or task goal across varying contexts by deploying entirely different muscular effectors, kinematic trajectories, and mechanical degrees of freedom. First articulated with conceptual clarity by early physiological and behavioral theorists such as Karl Lashley and later formalized in motor control theory by Nikolai Bernstein, motor equivalence stands as a direct challenge to any reductionist model that attempts to equate learned behavior with fixed patterns of peripheral motor activation. In typical biological actions, the goal of an action cannot be reduced to the specific physiological machinery utilized to execute it. A human being, for instance, can write their signature on a piece of paper using the fine intrinsic muscles of the fingers, write it across a blackboard using large movements of the shoulder and elbow joints, or even grasp a stylus between their toes or teeth to trace their name in the sand. Despite the vastly different kinetic profiles, degrees of freedom, and muscular systems employed, the structural topology and distinct stylistic properties of the signature remain recognizably identical.

This fundamental distinction between kinesthetic muscle-chain execution and teleological outcome navigation exposes the profound gulf separating mechanical reflexes from purposive, goal-directed behavior. When an organism navigates an environment, the teleological outcome—attaining a source of nourishment, escaping a threat, or returning to a home base—remains constant, whereas the physical execution can vary continuously depending on obstacles, changes in terrain, or internal biomechanical state. If learning were nothing more than the physical conditioning of a fixed chain of neuromuscular units, any alteration in the physical state of the effectors would result in complete functional breakdown. The existence of motor equivalence implies that the central nervous system must possess an effector-independent representation of the goal and the spatial pathway toward it. The organism plans action at an abstract, programmatic level before translating that plan into the dynamic motor commands appropriate for whatever peripheral muscle groups are available. Consequently, motor equivalence serves as a foundational pillar for cognitive models of action, providing undeniable empirical proof that behavioral organization is governed from the top down by representational architectures rather than assembled from the bottom up by unmediated peripheral mechanics.

1.2 D.A. Macfarlane: Academic Background and Research Aims

Donald A. Macfarlane conducted his landmark research within the intellectually vibrant Department of Psychology at the University of California, Berkeley, during the late 1920s. Under the charismatic and intellectually pioneering mentorship of Edward Chace Tolman, the Berkeley laboratory had emerged as the epicenter of a rigorous theoretical revolt against the reigning behavioral orthodoxies of the era. While mainstream American psychology was dominated by radical reductionism that viewed all animal behavior as mechanical stimulus-response (S-R) reflexes, Tolman’s group argued for a purposive behaviorism that recognized agency, expectation, spatial cognition, and goal-directedness as legitimate, measurable dimensions of scientific inquiry. Macfarlane’s work was conceived not as an isolated empirical curiosity, but as a pointed methodological strike aimed at the heart of the prevailing mechanistic dogmas.

In his monumental 1930 monograph, The Role of Kinesthesis in Maze Learning, published in the University of California Publications in Psychology, Macfarlane established his primary research objective: to isolate, interrogate, and ultimately evaluate the causal role of muscular feedback in the acquisition and execution of learned spatial navigation. At the time, the dominant academic consensus, championed by behaviorists who followed John B. Watson, asserted that spatial learning was entirely reducible to kinesthesis—the internal sensory perception of movement arising from receptors in the muscles, tendons, and joints. The peripheralist camp insisted that when a rat mastered a complex alley maze, it had simply linked a series of kinesthetic sensations into a continuous chain of mechanical reflexes. Macfarlane set out to design an experiment that could definitively decouple the spatial geometry of the learning environment from the specific neuromuscular sequences previously used to navigate it, thereby determining whether navigation was driven by kinesthetic memory or by an abstract representation of spatial relationships.

1.3 Significance of the 1930 Swimming Rats Paradigm

The paradigm Macfarlane designed to execute this conceptual decoupling was as brilliant in its structural simplicity as it was devastating in its theoretical implications: the flooded maze paradigm. By constructing an intricate labyrinth that could function interchangeably as an ordinary dry-floor running maze or be converted into a deep-water swimming maze, Macfarlane developed a powerful tool for experimental manipulation. In this apparatus, an animal could acquire absolute mastery over a specific navigational route on solid ground, running on all fours through a sequence of turns, and then be immediately confronted with the identical spatial labyrinth filled with water. In the aquatic condition, the animal could no longer rely on terrestrial stepping patterns, footfall impacts, or standard running kinetics; instead, it was forced to keep itself afloat and propel its body forward through bilateral paddling and whole-body aquatic mechanics.

This operational transition from terrestrial running to aquatic swimming shattered the explanatory power of the classical reflex-arc model. Under the terms of stimulus-response kinesthetic chaining, the running rat had reinforced a precise series of leg contractions. In deep water, those precise contractions were functionally impossible; executing the dry-land muscular sequence while submerged would result in drowning rather than navigation. If the peripheralist dogma were correct, placing a trained runner into the flooded maze should have obliterated its navigational capacity, resetting the animal to random, chance-level exploration. Conversely, if the animal successfully traversed the submerged maze without making excessive errors, it would prove that the acquired spatial knowledge was functionally independent of the specific effectors used during the acquisition phase. The 1930 swimming rats experiment provided a historic empirical benchmark, demonstrating that animal navigation is governed by higher-order mental representations capable of operating across radically different bodily media.

2. Historical Context: Early 20th-Century Behaviorism and S-R Chaining

2.1 The Rise of Watsonian Behaviorism and Muscle-Twitch Psychology

To fully appreciate the conceptual weight of Macfarlane’s experiment, one must understand the intellectual dominance of early twentieth-century American behaviorism. Following the publication of John B. Watson’s 1913 manifesto, “Psychology as the Behaviorist Views It,” the discipline underwent a conscious, aggressive purge of all introspectionist methodologies and mentalistic concepts. Concepts such as consciousness, imagery, thought, and internal representations were cast aside as unscientific remnants of philosophical dualism. Watson proclaimed that psychology must restrict itself exclusively to the objective observation of overt, measurable behavior, seeking to predict and control the actions of organisms through environmental manipulation. Watson’s radical reductionism framed the living organism as a biological automaton whose operations could be explained entirely in terms of physical stimuli and overt physiological responses.

This epistemological stance crystallized into what critics and historians frequently characterize as “muscle-twitch psychology.” Within this framework, all complex mammalian actions were conceptualized as concatenated chains of peripheral reflexes. Watson argued that even human thinking was nothing more than covert, sub-vocal motor activity—minute contractions of the larynx and vocal cords. When applied to spatial learning, this perspective yielded the kinesthetic chaining hypothesis: an animal learning to navigate an intricate maze was believed to be stringing together a chain of elemental motor acts. The sensory feedback generated by the contraction of muscles during turn $N$ served as the direct physical stimulus ($S$) that triggered the motor response ($R$) for turn $N+1$. In Watson’s view, the internal life of the organism was irrelevant; behavior was driven by peripheral mechanics operating through simple stimulus-response arcs.

2.2 The Mechanistic Assumptions of Early Learning Theorists

The peripheralist ethos was reinforced by the theoretical models of prominent learning theorists such as Edward L. Thorndike and, later, Clark L. Hull. Thorndike’s foundational Law of Effect posited that behaviors followed by a satisfying state of affairs were mechanically stamped into the neural substrate, strengthening the direct synaptic connection between the immediate situational stimulus and the specific physical response. Thorndike conceived this process as entirely automatic, occurring without any conscious awareness or cognitive understanding of causal or spatial relations. The animal did not learn *where* the food was located; it simply formed an unthinking habit of executing a specific physical behavior whenever it was placed within that specific physical environment.

This formulation was subsequently integrated into Clark Hull’s quantitative behavior system, which conceptualized learning through the mathematical accumulation of habit strength ($sHr$). For Hull, habit strength represented an enduring physiological modification that bound a stimulus trace directly to a specific effector contraction, driven by primary drive reduction. Within these theoretical frameworks, maze navigation was seen as the execution of kinesthetic traces bound to physical limb movements. The running animal was viewed as an anatomical clockwork mechanism: its paws struck the floorboards at fixed intervals, its torso flexed at precise angles, and its internal proprioceptors registered the tension of each muscular exertion. Because this peripheralist architecture treated spatial learning as a physical chain of localized motor responses, it harbored an inherent theoretical vulnerability: if the physical execution of any link in the chain were disrupted, the entire behavioral cascade should inevitably collapse.

2.3 Early Critiques and Anomalies in Animal Navigation

Despite the dominance of radical peripheralism, empirical anomalies began to surface within laboratory research, casting doubt on the muscle-twitch paradigm. The most notable early dissent emerged from the work of physiological psychologist Karl Lashley. In a series of pioneering surgical experiments, Lashley severed the dorsal sensory spinal roots of rats—effectively performing surgical deafferentation that eliminated all kinesthetic sensory feedback from the limbs to the central nervous system. According to the peripheral chaining model, these deafferented animals should have been entirely incapable of running a previously mastered maze, as they could no longer receive the internal proprioceptive stimuli necessary to trigger subsequent turns. However, when placed in the maze, Lashley’s deafferented animals continued to navigate the labyrinth successfully, stumbling, dragging their impaired limbs, and using abnormal motor patterns, yet consistently executing the correct sequence of spatial turns to reach the food box.

Concurrently, sensory-deprivation experiments conducted by researchers like Walter S. Hunter underscored the profound resilience of the navigational system. Experimenters systematically eliminated individual sensory modalities—blinding rats through enucleation, deafening them through tympanic destruction, severing their olfactory bulbs to abolish smell, and clipping their vibrissae to eliminate tactile feedback. Remarkably, animals deprived of any single sensory system, and even those deprived of multiple modalities simultaneously, retained a surprising capacity to master complex mazes. These anomalies suggested that animal navigation could not be reduced to any solitary sensory stream or isolated kinesthetic feedback loop. Nevertheless, dogmatic behaviorists maintained that an intact animal, under normal conditions, relied primarily on the continuous, unbroken chain of kinesthetic feedback. It was this fundamental contention that Macfarlane sought to put to an unambiguous, definitive experimental test.

3. Theoretical Framework: Peripheralist Theories vs. Central Spatial Representations

3.1 The Peripheralist Hypothesis of Kinesthetic Feedback

The core assertion of the peripheralist hypothesis was that internal kinesthesis served as the primary, indispensable organizing scaffold for animal navigation. Proprioceptive feedback from muscle spindles, Golgi tendon organs, and joint receptors was assumed to function as a closed-loop mechanical control system. Consider an animal negotiating an alley that extends forward two feet before turning sharply to the right. Under the peripheralist model, the physical action of taking four successive strides of a specific length and cadence generates a cumulative kinesthetic sensation. Upon reaching an internal proprioceptive threshold, this sensory accumulation triggers the execution of a right-hand torso turn and a corresponding lateral shift in limb movement. The turn itself produces a distinct new set of sensory signals, which then initiates the next sequence of linear strides.

This closed-loop model made clear and testable predictions: any disruption to the animal’s physical movement patterns must inevitably cause navigational failure. If an animal’s stride length were shortened, if its turning angle were altered, or if the posture of its body were significantly perturbed, the expected sequence of kinesthetic sensations would be broken. Without the precise internal stimulus to trigger the next motor response, the animal should become disoriented, running into walls or wandering into blind alleys. The peripheral model was fundamentally dependent on the metric consistency of physical movement; it could not account for an organism navigating successfully if its physical locomotion was altered, because it possessed no theoretical mechanism for separating the physical act of moving from the knowledge of where one was going.

3.2 Tolman’s Purposive Behaviorism and Cognitive Mapping

In direct opposition to this peripheral reductionism stood Edward C. Tolman’s paradigm of purposive behaviorism. Tolman introduced a fundamental conceptual distinction between “molecular” behavior and “molar” behavior. Molecular behavior referred to the underlying physiological mechanics of an act: the firing of motor neurons, the release of acetylcholine at the neuromuscular junction, the contraction of specific muscle fibers, and the raw physical movements of joints. Molar behavior, by contrast, referred to the integrated, goal-directed action of the whole organism—an action characterized by its teleological orientation toward a goal, its structural responsiveness to environmental opportunities, and its flexible adaptation to changing physical conditions. Tolman maintained that psychology must concern itself with molar behavior, as the molecular components could vary infinitely without altering the essential nature of the behavioral act itself.

Within this molar framework, Tolman developed his foundational concept of the cognitive map, an idea that would culminate in his landmark 1948 paper, “Cognitive Maps in Rats and Men.” Tolman posited that as an organism explores an environment, it does not stamp in empty stimulus-response habits; rather, it acquires information about the layout of the physical world. The animal learns “what leads to what”—forming internal representations that encode spatial relationships, environmental geometry, and the location of goals. Tolman described these mental structures using terms like *sign-gestalts* and *field maps*, arguing that spatial knowledge is stored as an internal model of the external terrain. A critical prediction of this theory was that such internal representations must be amodal and effector-independent: an animal that knows the layout of an environment should be able to navigate that layout using whatever physical effectors are available, whether it walks, runs, hops, or swims.

3.3 Hypothesis Formulation: Effector Independence in Rodents

Macfarlane recognized that this theoretical confrontation between peripheralist mechanics and central cognitive maps could be translated into an elegant, definitive empirical hypothesis. By utilizing an apparatus that allowed for the sudden flooding of a familiar maze, Macfarlane could experimentally manipulate the organism’s locomotor medium without altering the underlying spatial topology of the environment. If the peripheralist model was correct, an animal trained to run through a dry maze would have acquired a chain of terrestrial kinesthetic habits that would be entirely useless in water. Deprived of solid ground, forced to deploy bilateral paddling movements, and experiencing radically altered proprioceptive feedback, the rat’s navigational ability should break down completely, dropping its performance back to chance levels.

Macfarlane’s formal alternative hypothesis asserted the existence of complete effector independence in rodent navigation. If the learning process resulted in a central, cognitive representation of the maze geometry, the animal’s navigational competence would survive the transition between locomotor media. The rat would recognize its position within the represented spatial field and deploy whatever novel motor actions were required to achieve its goal within the new medium. Should an experienced runner navigate the flooded maze with accuracy comparable to its dry-land performance, the stimulus-response kinesthetic chaining model would be thoroughly disconfirmed. Such a result would demonstrate that the physical movements of running were merely the temporary, interchangeable tools used to execute an abstract spatial plan, rather than the mental architecture of the learning itself.

4. Experimental Design of Macfarlane’s 1930 Study

4.1 Subject Selection and Pre-Experimental Acclimation

To ensure high experimental rigor and eliminate confounding variables, Macfarlane employed rigorous subject selection and acclimation protocols. The experimental subjects were healthy, adult albino rats (*Rattus norvegicus*), a standard and well-characterized model organism in early twentieth-century comparative psychology. Macfarlane chose animals of consistent age, genetic heritage, and developmental history to minimize individual variance in baseline learning capacities and physiological endurance. The animals were housed under standardized laboratory conditions with regulated ambient temperatures and controlled day-night light cycles, ensuring that extraneous environmental fluctuations would not introduce systematic errors into the experimental data.

Because the experiment required testing animals in both food-motivated and aquatic contexts, nutritional regulation and water habituation protocols were established with great care. Macfarlane instituted a standardized food-deprivation schedule, maintaining the rats at a specific percentage of their free-feeding body weights to establish a consistent, stable motivational drive for food rewards at the end of the maze. Crucially, Macfarlane recognized that suddenly introducing a rodent into deep water can trigger profound panic responses—characterized by frantic, uncoordinated thrashing, hyperventilation, and freezing—which could obscure the animal’s underlying cognitive competence. To prevent these emotional and physiological artifacts from distorting his results, all experimental animals were given structured pre-experimental acclimation trials in water. They were introduced to shallow aquatic tanks and allowed to grow accustomed to immersion, ensuring that when they were later tested in the flooded maze, their behavior would reflect spatial navigation rather than acute panic.

4.2 Cohort Organization and Experimental Groups

Macfarlane arranged his experimental architecture around balanced cohorts designed to isolate the direction and symmetry of locomotor transfer. The primary experimental cohort was the Running-to-Swimming group: these animals received comprehensive, daily training in the dry-floor maze until they achieved asymptotic navigational mastery, at which point the maze was flooded, and their navigational accuracy was assessed while swimming. To ensure that the observed effects were not an asymmetric artifact of transitioning from dry land to water, Macfarlane established a reciprocal Swimming-to-Running group. These rats acquired initial mastery of the maze while swimming through deep water and were then tested on the dry floor, where they were forced to transition from aquatic propulsion to quadrupedal running.

Alongside these critical transfer groups, Macfarlane ran dedicated control cohorts that maintained a single locomotor mode throughout the entire experimental timeline. A dry-control group ran the dry maze across all phases, establishing the baseline learning curve, latency norms, and error decay rates of uninterrupted terrestrial training. Similarly, an aquatic-control group swam through the flooded maze throughout all sessions, providing a continuous benchmark for the acquisition and stabilization of swimming-based navigation. By cross-comparing the performance of the switch groups with these continuous-mode control groups, Macfarlane could apply statistical controls to account for natural learning progressions, simple task habituation, and baseline latency differences, thereby isolating the precise behavioral impact of the locomotor transition.

4.3 Architectural Design of the Experimental Maze

The maze constructed for Macfarlane’s experiment was an intricate, multiple-unit alley labyrinth designed to present the subjects with a challenging spatial problem. The apparatus consisted of a series of interconnected T-units and complex alley corridors, creating a branching network where the animal had to make an unambiguous, binary choice at each critical junction: turning either toward the correct path leading to the goal, or into a blind alley that resulted in a dead end. The maze was deliberately engineered to be sufficiently complex that an animal could not solve it through trivial, simple perceptual strategies—such as directly sensing the goal box from afar or relying on continuous right- or left-turn habits.

The maze’s construction required high engineering precision, as it had to function seamlessly under both dry and aquatic conditions. The walls of the maze were built to be completely watertight and rose high above the reach of the animals, preventing them from climbing over the partitions or obtaining orienting cues by peering over the tops of the alleys. The geometry of the true paths and the blind alleys was mathematically balanced, ensuring that the total linear distance, the number of turns, and the mechanical complexity were distributed evenly throughout the structure. Most importantly, the physical architecture of the maze walls, turns, and decision points remained perfectly identical whether the floor was dry or deeply flooded, guaranteeing that the objective spatial topology of the environment remained entirely unchanged across both experimental conditions.

5. Methodological Innovations: The Flooded Maze Apparatus

5.1 Engineering the Water-Submerged Maze Floor

The defining technical achievement of Macfarlane’s methodology was the engineering of the flooded maze apparatus. To allow for rapid conversion between terrestrial and aquatic testing environments without altering the physical alignment of the alley walls, Macfarlane designed an integrated false-floor and drainage mechanism. Beneath the primary navigational labyrinth lay a deep tank equipped with precision water-inlet valves and rapid-drain release gates. In the terrestrial condition, a rigid, elevated false floor made of treated wood and mesh was locked into place, presenting the rats with a stable, dry running surface indistinguishable from a standard laboratory alley maze.

To convert the maze into an aquatic environment, the false floor could be lowered or the water level raised, filling the alleys to a carefully calibrated depth of several inches. Macfarlane regulated this water depth with meticulous care: the water had to be deep enough to ensure that the rats could not touch the bottom with their paws or push off the floor with their hind legs, thereby forcing them to engage in continuous, unassisted swimming. At the same time, the water level had to remain low enough relative to the alley walls that the animals could not reach the upper rims to pull themselves out. Furthermore, Macfarlane paid close attention to the thermal consistency of the water, utilizing heating elements to maintain the bath at a steady, lukewarm temperature (approximately 22°C to 24°C). This prevented hypothermic shock, muscular cramping, or lethargy, ensuring that the animals remained physiologically robust and motivated throughout the aquatic trials.

5.2 Kinematic Dissimilarity Between Running and Swimming

The methodological validity of Macfarlane’s study rested entirely on the objective, biomechanical dissimilarity between terrestrial running and aquatic swimming in the laboratory rat. If the two forms of locomotion shared significant kinematic or muscular overlap, a behaviorist could argue that transfer occurred through shared peripheral reflexes. However, physiological analysis reveals that quadrupedal running and swimming in Rattus norvegicus represent entirely distinct motor patterns. During terrestrial running, the rat employs an alternating, diagonal gait pattern: diagonally opposite limbs (e.g., right front paw and left hind paw) strike the ground simultaneously, absorbing the impact of the body weight against a hard surface, pushing off the substrate, and propelling the center of mass forward through a continuous sequence of stance and swing phases.

When submerged in water, the rat’s biomechanical profile undergoes a complete functional transformation. The buoyancy of the fluid counteracts gravity, relieving the skeleton and musculoskeletal system of the need to support bodily weight. In response, the rat shifts to an aquatic propulsion mode: the front paws are often held tucked close beneath the chin or used primarily for delicate directional steering, while the primary propulsive force is generated by rapid, high-frequency, alternating or near-synchronous strokes of the powerful hind legs, accompanied by undulating lateral flexions of the spine and tail. The biomechanical differences between the two modes can be summarized as follows:

• Gravitational Support: Terrestrial running requires continuous, anti-gravity skeletal loading; aquatic swimming relies entirely on buoyant fluid suspension.
• Substrate Interaction: Running involves frictional paw impacts against a solid surface; swimming depends on dynamic fluid resistance and hydrodynamic drag.
• Effector Deployment: Running utilizes all four limbs in a coordinated diagonal quadrupedal gait; swimming predominantly uses the hindlimbs for propulsion, with the forelimbs relegated to balance and steering.
• Proprioceptive Feedback: Running produces high-impact kinesthetic feedback through deep joint compression and tendon tension; swimming generates low-impact, continuous fluid resistance across the fur and skin.
• Sensory and Visual Perspective: In running, the head is held elevated and stable, providing a wide visual field; in swimming, the head is tilted back to keep the snout clear of the water, shifting the animal’s visual angle and vestibular orientation.

These pronounced physical differences create an insurmountable problem for the peripheralist model. If maze navigation were governed by a chained sequence of specific terrestrial muscle twitches and footfalls, that mechanical sequence would be completely useless in the water. The running motor program simply cannot be mapped onto the mechanics of swimming strokes. For a rat to navigate successfully while swimming, its central nervous system must bypass the specific motor mechanics of running and access an abstract, effector-independent plan capable of steering whatever movement machinery is currently engaged.

5.3 Elimination of Extraneous Cues

To confirm that successful navigation was driven by a true internal spatial representation rather than incidental guidance cues, Macfarlane instituted rigorous experimental controls to eliminate extraneous stimuli from the maze environment. A major confounding factor in rodent maze research is the presence of olfactory trails: animals naturally leave behind scent marks, sebum, and urine traces that subsequent animals—or the same animal on subsequent runs—can easily follow like a physical thread. Macfarlane neutralized this confound by exploiting the unique properties of his flooded apparatus. The water was continuously circulated, drained, and replenished, thoroughly washing away and diluting any chemical traces left by the animals. In the dry condition, the floor was systematically scrubbed and rotated to prevent the buildup of localized scent trails.

Macfarlane was equally vigilant in controlling for distal visual landmarks and directional acoustic cues. The entire maze apparatus was enclosed within high, uniform, featureless walls, and in many phases, surrounded by neutral, floor-to-ceiling cloth curtains that shielded the interior alleys from outside laboratory landmarks, such as ceiling fixtures, windows, shelves, or the experimenter’s desk. The illumination was kept diffuse and uniform, eliminating directional shadows that might indicate a heading toward the goal. Furthermore, the auditory environment was carefully controlled to eliminate localized noise gradients that could serve as acoustic beacons, ensuring that the rats could not navigate simply by following the hum of a motor, the sound of an exhaust fan, or the shuffling of the experimenter. Stripped of olfactory trails, distal visual beacons, and localized auditory cues, the animals were forced to rely exclusively on their internal spatial understanding of the maze’s topological layout.

6. Experimental Procedure and Control Measures

6.1 Phase One: Acquisition of the Maze Under Baseline Conditions

The experimental procedure began with Phase One: the baseline acquisition phase, in which the cohorts of rats were trained in their assigned initial locomotor modality. The terrestrial groups were placed individually at the maze’s designated starting threshold on the dry, elevated floor and were allowed to explore the branching corridors until they successfully located the food box, which contained a standardized food reward. The aquatic groups were gently lowered into the water at the identical starting location, swimming through the flooded labyrinth until they reached an accessible, dry escape ramp positioned in the goal box, where they also received their nutritional reinforcement. Each animal completed a rigorously standardized schedule of daily trials, with consistent inter-trial intervals to prevent excessive physical fatigue.

During every single run, Macfarlane recorded two primary dependent variables: error frequency and total transit latency. Errors were classified with meticulous behavioral precision, distinguishing between forward errors (entering a blind alley while moving along the primary path) and retracing errors (turning around and traversing backward through an alley segment that had already been successfully cleared). Over successive days, the baseline learning curves followed a classic acquisition trajectory: the animals began with high error rates and prolonged latencies characterized by extensive exploration, sniffing, and hesitation at decision points, which steadily declined over trials. The acquisition phase continued until each cohort reached a strict criterion of asymptotic performance—defined as multiple consecutive, errorless runs through the maze. Reaching this behavioral plateau ensured that the animals had mastered the maze’s spatial geometry before being subjected to the critical experimental manipulation.

6.2 Phase Two: The Critical Modality Shift

Once asymptotic mastery was firmly established, Macfarlane initiated Phase Two: the critical modality shift. This phase represented the decisive empirical test of the two competing theoretical models. For the running-to-swimming experimental cohort, the transformation was abrupt and unannounced: having achieved near-flawless performance on the dry floor, the rats were brought to the maze room the following day to find the apparatus fully submerged in deep water. The animals received no transitional training, no incremental deepening of the water, and no opportunity to practice swimming in the maze corridors beforehand. They were lifted from their home cages, placed directly into the deep water at the starting point, and their immediate, spontaneous behavioral response was recorded.

The protocol for the swimming-to-running group was executed with identical rigor: after achieving errorless performance in the flooded maze, the water was drained, the dry false floor was locked into position, and the rats were placed at the dry starting line without any intermediate habituation. Because these transitions were sudden and unpracticed, the immediate test trials were of paramount scientific importance. If the animals’ spatial competence depended on previously conditioned kinesthetic reflex chains, the sudden change in medium should produce catastrophic behavioral disruption, manifesting as an immediate spike in errors back toward chance-level exploration. If, on the other hand, the spatial knowledge was stored centrally as an amodal cognitive map, the animals should execute the correct navigational path on their very first trial in the new medium.

6.3 Rigorous Control Conditions and Error Definitions

To eliminate any ambiguity in the behavioral data, Macfarlane established clear operational definitions for every recorded action. An error was scored the precise moment the tip of the rat’s snout crossed the architectural threshold separating the main alley from a blind alley. Partial entries, indecisive hesitations, and brief pauses at the choice points were recorded separately as hesitation metrics rather than classified as full spatial errors. This distinction was critical: it prevented general, hesitation-inducing emotional responses—such as initial surprise at encountering deep water—from being conflated with a true breakdown of spatial knowledge. The observation protocols were conducted from concealed vantage points to eliminate experimenter expectancy bias, ensuring that the researcher’s physical presence or movements could not inadvertently cue the animals toward the correct path.

Furthermore, Macfarlane instituted statistical controls to separate cognitive decision-making from the physical dynamics of swimming. Swimming through water is hydrodynamically slower and requires far more physical work per unit of distance than running across a solid floor. Consequently, a simple increase in transit latency could not be interpreted as cognitive confusion or memory loss. By separating error frequency (a pure measure of spatial choice accuracy) from transit latency (which reflects both decision time and the physical mechanics of locomotion), Macfarlane ensured that the evaluation of motor equivalence was based on true navigational fidelity rather than the physical speed of the animal through the medium.

7. Empirical Findings and Quantitative Results

7.1 Error Rates Following the Locomotor Transition

The quantitative results of Macfarlane’s 1930 experiment delivered a clear, decisive verdict that ran directly counter to the predictions of Watsonian peripheralism. When the experienced running rats were placed into the flooded maze, their navigational performance did not collapse to chance levels. Instead, the rats swam through the submerged labyrinth with remarkable, near-flawless precision. Rather than entering the blind alleys that branch off from the main corridors, the animals navigated the flooded channels, made the correct turns at the major decision points, and swam directly to the goal box containing the escape platform and food reward.

A statistical analysis of the error curves demonstrated that the error frequencies recorded during the initial transfer trials were nearly identical to the asymptotic baseline scores established on the dry floor. The animals committed virtually zero forward errors at the primary choice points. While a very minor, transient increase in retracing errors was noted in a small subset of the subjects, close behavioral observation revealed that this was driven by initial water-startle reactions rather than spatial disorientation; the animals quickly corrected their heading and completed the maze without entering blind alleys. The switch subjects performed with navigational accuracy that was statistically indistinguishable from the control animals that had swum the maze throughout the entire training period. The experimental data revealed that the animals had retained their knowledge of the maze’s spatial layout, demonstrating that spatial learning could be transferred seamlessly across different locomotor modalities.

7.2 Latency Profiles and Swimming Dynamics

The latency metrics gathered across Phase Two provided further insight into the cognitive mechanisms at work. As predicted by hydrodynamic principles, the absolute time required to complete the maze increased when the animals were shifted from running to swimming. It simply takes a rat longer to paddle its body weight through water than it does to run across a dry floor. However, an analysis of the latency curves across subsequent trials revealed a rapid, steep decline in transit time, stabilizing into a consistent, efficient plateau within just two to three runs:

Most importantly, the animals showed no systematic hesitation at the maze’s critical choice points. Under the peripheralist model, one would expect long, confused pauses at every junction, as the animal would be waiting for a familiar terrestrial kinesthetic signal that never arrived. Instead, Macfarlane observed that the swimming rats navigated the decision points with smooth, continuous momentum. They approached the T-junctions, made their turns into the true path without pausing to touch the side walls, and swam steadily forward. This absence of choice-point hesitation proved that their spatial decision-making was fluid and confident, operating without any reliance on terrestrial proprioceptive triggers.

7.3 Symmetry of Transfer: Swimming to Running

To confirm that this cognitive transfer was not an asymmetric artifact of transitioning from dry land to water, Macfarlane analyzed the performance of the reciprocal group: the animals that were trained in the water and then switched to the dry floor. The data revealed complete, symmetrical transfer. Rats that had acquired their spatial understanding of the maze entirely through swimming—using aquatic paddling movements, buoyant body suspension, and low-angle visual cues—demonstrated near-perfect navigational accuracy the very first time they were placed on the dry floor.

Upon being introduced to the dry starting platform, these animals did not stumble about aimlessly or explore the blind alleys. Instead, they immediately launched into an efficient quadrupedal run, executing the correct sequence of turns and racing directly to the goal box with error scores that matched their aquatic baseline. This demonstrated that the cognitive transfer was fully reciprocal. Whether the spatial representation was acquired through the mechanics of swimming strokes or the cadence of running strides, the underlying spatial knowledge was stored in a centralized, amodal format that was accessible and executable by any compatible motor system the animal possessed.

8. Analysis of Motor Equivalence in Locomotor Transition

8.1 The Disconfirmation of Kinesthetic Chaining

The empirical findings of Macfarlane’s study dealt a fatal blow to the kinesthetic chaining hypothesis that had anchored early behaviorist theory. The assertion that complex spatial navigation is nothing more than a concatenated sequence of peripheral, muscle-bound reflexes could not survive the reality of the swimming rats. Under the strict Watsonian model, the peripheral motor response $R_1$ (a sequence of dry-land strides) generates the internal kinesthetic stimulus $S_2$, which in turn triggers the motor response $R_2$ (a lateral torso turn and turn-specific limb contraction). In the flooded maze, the very first link in this physical chain, $R_1$, was rendered impossible by the aquatic environment. A rat cannot execute a dry-land running stride in deep water; doing so would fail to provide propulsion and lead to drowning.

Because the physical response $R_1$ never occurred, the internal kinesthetic stimulus $S_2$ was never generated. Under the terms of strict peripheral chaining, the entire behavioral chain should have collapsed immediately at the starting line. The fact that the animals navigated the maze successfully on their very first aquatic exposure proved that the reflex-chaining mechanism cannot be the foundation of spatial learning. Macfarlane’s work exposed the core flaw of radical peripheralism: it had confused the physical movements used during performance with the internal information acquired during learning. The specific muscle twitches observed during running were merely one temporary way of expressing the underlying spatial knowledge, not the structure of the knowledge itself.

8.2 Effector Independence in Biological Systems

Macfarlane’s empirical demonstration of motor equivalence in the laboratory rat revealed a universal organizing principle of biological motor control: effector independence. Effector independence means that the central nervous system separates the planning of a goal-directed action from the physical execution of that action by specific peripheral muscles. The brain constructs a high-level representation of the desired outcome—an intentional vector, an environmental trajectory, or an abstract spatial path—and this top-down plan is then flexibly translated by lower-level motor circuits into the physical movements required for the current bodily context.

This principle is directly analogous to the phenomenon of human motor equivalence, best illustrated by the classic handwriting experiments discussed by Karl Lashley. When an adult writes their signature, they typically use the fine motor circuits of the hand and fingers to control a pen. However, that same individual can trace an identical signature on a large blackboard using broad movements of the shoulder, write it in the snow by pacing it out with their whole body, or even hold a chalk stick between their teeth or toes. The absolute size, velocity, and muscle groups utilized across these tasks differ completely, yet the signature’s core topological form, letter proportions, and individual stylistic flourishes remain consistent. Macfarlane proved that this sophisticated capacity for effector independence is not an advanced cognitive luxury exclusive to humans; it is a fundamental property of mammalian nervous systems, functioning just as reliably in a swimming rodent as it does in a human writing their name.

8.3 The Nature of Acquired Spatial Information

Macfarlane’s results provided crucial insights into the nature of the information that animals acquire while learning a maze. The fact that an animal can switch effortlessly between running and swimming proves that it is not encoding isolated, egocentric muscle commands (such as “contract the left biceps femoris, then flex the right tibialis”). Instead, the animal is encoding allocentric or environmental trajectories: an understanding of paths, directions, geometric boundaries, and spatial goals within the physical space of the maze.

This transformation of sensory inputs into an amodal spatial representation means that the information stored in the animal’s nervous system is divorced from the specific sensory and motor channels through which it was acquired. The animal does not remember what the maze “felt like” to its running leg muscles; it remembers the topological structure of the maze itself. It learns that this alley leads forward, that a right turn at the second junction leads toward the goal, and that a left turn leads to a dead end. This abstract representation of space exists at a higher cognitive level than any peripheral sensory stream, allowing the animal to navigate the environment regardless of whether it is running, swimming, or overcoming physical obstacles along the way.

9. Impact on Tolman’s Purposive Behaviorism and Cognitive Maps

9.1 Empirical Vindication of Cognitive Mapping

Donald Macfarlane’s 1930 study served as a primary empirical pillar supporting Edward Tolman’s system of purposive behaviorism. When Tolman published his seminal 1948 paper, “Cognitive Maps in Rats and Men,” Macfarlane’s flooded-maze experiment occupied a place of honor as definitive evidence for central spatial representations. For Tolman, the swimming rats provided unassailable proof that the brain is not a passive switchboard mechanically connecting incoming sensory stimuli to outgoing motor responses. Rather, Tolman argued, the central nervous system acts as an active, flexible information processor that builds a comprehensive mental map of the environment.

Macfarlane’s findings firmly established Tolman’s critical theoretical distinction between learning and performance. Traditional stimulus-response behaviorism assumed that learning and performance were inseparable: an animal’s performance was simply the direct reflection of its accumulated habit strength, stamped into its physical reflexes. Tolman, drawing on Macfarlane’s data, pointed out that learning—the acquisition of environmental information—occurs independently of the specific physical motor patterns used during performance. The running rats had acquired a detailed understanding of the maze’s spatial layout that was completely preserved even when the motor performance was radically transformed into swimming strokes. The cognitive map had been formed quietly in the nervous system, ready to be translated into whatever physical action was required by the situation at hand.

9.2 The Demise of Radical Peripheralism

The publication of Macfarlane’s monograph, alongside complementary findings from the Berkeley laboratory, marked the beginning of the end for radical, unmediated peripheralism within academic psychology. While the mechanistic behaviorist tradition continued to exert significant influence through the sophisticated mathematical models of Clark Hull and the operant conditioning frameworks of B.F. Skinner, the dogma that behavior could be explained entirely through peripheral muscle twitches and localized kinesthetic chains was permanently undermined. Theoretical debates shifted away from whether central processing existed, focusing instead on how to mathematically define, model, and constrain these internal variables without abandoning scientific rigor.

Faced with Macfarlane’s swimming rats, neo-behaviorists were forced to make significant theoretical concessions. They could no longer claim that spatial learning was simply a physical chain of leg-muscle contractions. Instead, they were driven to develop increasingly convoluted theoretical models, such as Hull’s “pure stimulus acts” ($rg-sg$ mechanisms) and fractional anticipatory goal responses, in an effort to explain how spatial coordination could be maintained through internal, central processes while still clinging to S-R terminology. Through this prolonged theoretical struggle, the Berkeley laboratory established itself as the intellectual rival to the mechanistic Yale school, fundamentally reshaping learning theory and paving the way for the cognitive revolution that would transform psychology in the late twentieth century.

9.3 Redefining the Animal as an Active Information Processor

In broader historical perspective, Macfarlane’s experiment played an essential role in redefining the animal within psychological science: shifting the paradigm from a reactive automaton to an active, flexible information processor. The classical behaviorist model had reduced the animal to a purely reactive entity—a passive bundle of reflexes pushed and pulled by external stimuli and internal physiological drives, blindly repeating whatever physical movements had previously been reinforced. Macfarlane’s work exposed this portrait as fundamentally inadequate to describe mammalian behavior.

By showing that a rat could immediately adapt its spatial knowledge to a completely novel physical medium, Macfarlane demonstrated that animals act as proactive problem-solvers. The rodent in the flooded maze is not mechanically executing a fixed motor script; it is assessing its current spatial position, keeping its behavioral goal firmly in mind, and creatively deploying its physical effectors to achieve that goal in the face of unexpected environmental challenges. This early insight anticipated modern cognitive concepts such as motor schemas, internal dynamic models, and cybernetic control systems, prefiguring our contemporary understanding of the brain as an adaptive, predictive engine that uses internal models of the world to guide purposeful action.

10. Critiques, Replications, and Methodological Scrutiny

10.1 Contemporary Critiques from Neo-Behaviorists

Despite its conceptual elegance, Macfarlane’s 1930 study was met with intense theoretical resistance from committed stimulus-response behaviorists determined to defend their mechanistic paradigm. The primary critique raised against Macfarlane’s conclusions was that the animals were not relying on an amodal cognitive map, but were instead using remaining distal visual cues that extended beyond the maze walls. Critics argued that even if the rat’s leg-muscle kinesthesis had been abolished by the water, the animal could still orient itself using subtle visual markers on the ceiling, lighting patterns, or the tops of the alley walls, transforming the task into simple visual beacon following rather than abstract cognitive mapping.

Other neo-behaviorists attempted to rescue the peripheral chaining model by proposing the existence of generalized “orienting reflexes.” They argued that the fundamental links in the associative chain were not specific leg contractions, but rather broad, full-body vestibular orientations—such as rotating the head or aligning the torso 90 degrees to the right. Under this view, these whole-body orienting reflexes were preserved in the water, allowing the animal to string together a sequence of directional body headings without needing an internal cognitive map. Furthermore, some critics pointed to possible vestigial tactile cues, suggesting that the swimming rats might have brushed their whiskers along the alley walls to physically trace their way through the labyrinth.

10.2 Subsequent Replication Attempts and Variations

To resolve these theoretical disputes, subsequent researchers designed controlled replication studies and experimental variations throughout the 1930s, 1940s, and 1950s. To address the visual-cue counterargument, follow-up experiments placed blindfolded or surgically enucleated rats into the flooded maze paradigm. Even when completely deprived of visual input, rats that had mastered the dry maze retained their navigational competence when the corridors were flooded, navigating the submerged channels through sheer spatial memory and proving that distal visual beacons were not necessary for successful transfer.

Other researchers introduced geometric modifications to the maze during the transition, changing the width of the alleys, adding turbulent cross-currents to disrupt vestibular orientation, or altering the starting positions within the labyrinth. Across these diverse variations, the core phenomenon of motor equivalence held remarkably firm. The capacity for cross-modality transfer was subsequently replicated in a wide variety of mammalian species, demonstrating that the ability to decouple an environmental goal from the specific effectors used to reach it is an evolutionary adaptation shared across the mammalian lineage, rather than an idiosyncratic quirk of the laboratory rat.

10.3 Methodological Limitations of the 1930 Experiment

While Macfarlane’s conceptual framework was groundbreaking, a modern scientific critique must acknowledge the methodological and technological limitations of his 1930 study. From a contemporary perspective, the sample sizes employed across his experimental cohorts were relatively modest, reflecting the standard laboratory practices of the early twentieth century. Although the behavioral differences between the groups were clear and consistent, modern neuroscience demands larger statistical power and more formal mathematical modeling to definitively rule out subtle interaction effects.

Furthermore, Macfarlane was limited by the observation technology of his era. He relied on real-time human observation, manual mechanical stopwatches, and paper-and-pencil error tracking. He lacked the high-speed digital videography, automated motion-capture tracking, and computer-assisted kinematic analysis that modern biomechanics laboratories use to examine the micro-adjustments of an animal’s gait and stroke dynamics. Most significantly, early twentieth-century psychologists had no access to direct neurophysiological recording tools; they could observe behavioral inputs and motor outputs with exceptional precision, but the internal neural mechanisms that generated cognitive maps and enabled motor equivalence remained an inaccessible biological black box.

11. Neurobiological and Computational Perspectives on Motor Equivalence

11.1 Hippocampal Place Cells and Cognitive Mapping

The theoretical concepts championed by Macfarlane and Tolman received dramatic biological confirmation decades later with the discovery of the brain’s internal navigation system. In 1971, John O’Keefe and Jonathan Dostrovsky identified place cells within the CA1 and CA3 regions of the rodent hippocampus. These specialized neurons fire robustly whenever an animal enters a specific, localized region of its environment—the cell’s “place field”—regardless of the animal’s physical posture, velocity, or the specific motor actions it is performing. This discovery, elaborated in O’Keefe and Lynn Nadel’s landmark 1978 book, The Hippocampus as a Cognitive Map, provided the long-sought physiological substrate for Tolman’s purposive behaviorism.

Crucially, modern neurophysiological recordings have directly validated Macfarlane’s behavioral findings by monitoring hippocampal place cells as animals transition between different locomotor modalities. Experiments using wireless telemetry and miniaturized calcium-imaging cameras have shown that when a rat switches from running on a solid track to swimming through a water channel, its hippocampal place cells maintain their stable, location-specific firing fields:

This neurobiological invariance provides direct physiological confirmation of effector independence: the hippocampal network maintains a steady, amodal representation of where the animal is located in the environment, broadcasting this spatial coordinate to downstream motor structures regardless of whether the body is running or swimming.

11.2 Motor Cortex and Striatal Separation of Plan vs. Execution

The neurological implementation of motor equivalence depends on a hierarchical separation of planning and execution distributed across the mammalian brain. Modern functional imaging and tract-tracing studies demonstrate that the premotor cortex, supplementary motor area (SMA), and posterior parietal cortex form an executive planning network that represents actions at an abstract, goal-oriented level. These higher-order cortical regions encode intentional vectors—such as “move toward the north-west corridor”—independent of the specific peripheral muscles required to execute the movement. This abstract plan is then routed through the motor loops of the basal ganglia, specifically the striatum, which plays a central role in selecting behavioral goals and action policies based on reinforcement history and current motivational state.

Only at the level of the primary motor cortex (M1), the brainstem motor nuclei, and the spinal cord’s central pattern generators (CPGs) is this abstract spatial trajectory translated into concrete, muscle-specific motor commands. Spinal central pattern generators possess distinct, autonomous networks dedicated to running and swimming: the spinal circuits that coordinate the alternating diagonal gait of quadrupedal terrestrial locomotion are physically distinct from the networks that drive the bilateral paddling and spinal flexions of swimming. Because the brain’s executive planning network is functionally separated from these lower-level spinal pattern generators, the central nervous system can seamlessly re-route an amodal spatial plan to entirely different spinal motor circuits in real time, making motor equivalence possible.

11.3 Robotics and Computational Models of Motor Equivalence

The principles uncovered by Macfarlane’s swimming rats have become central to the development of modern computational motor control and autonomous robotics. In robotics, the challenge of motor equivalence is formalized through the concept of inverse kinematics and inverse dynamics. A robot navigating an unknown environment must solve the degrees-of-freedom problem originally framed by Nikolai Bernstein: calculating how to coordinate multiple motors, joints, and actuators to follow an intended spatial trajectory. Modern autonomous control systems solve this by decoupling the spatial navigation engine (the Simultaneous Localization and Mapping, or SLAM, algorithm) from the physical actuation module, allowing the machine to follow a designated path whether it uses wheels, mechanical tracks, or articulated walking legs.

Similarly, contemporary reinforcement learning algorithms mirror this division of labor through hierarchical actor-critic architectures. In these computational models, a high-level policy network determines the spatial sub-goals and environmental trajectories within a latent, amodal representational space. A lower-level execution network then takes these abstract spatial sub-goals and computes the specific, high-frequency torque commands required for the robot’s physical actuators. By demonstrating that biological systems separate environmental navigation from low-level motor execution, Macfarlane provided an early blueprint for the design of adaptable, robust artificial intelligence capable of operating across unpredictable, shifting physical environments.

12. Enduring Legacy and Contemporary Relevance in Cognitive Neuroscience

12.1 From Animal Mazes to Modern Cognitive Paradigms

The conceptual framework established by Donald Macfarlane’s 1930 study laid the empirical groundwork for some of the most widely used behavioral paradigms in contemporary behavioral neuroscience. Most notably, Macfarlane’s use of water as an experimental medium directly influenced the development of the Morris Water Maze, designed by Richard Morris in 1981. The Morris Water Maze has become the global gold standard for assessing spatial learning, working memory, and hippocampal function in rodent models of human neurodegenerative diseases, such as Alzheimer’s. Just as in Macfarlane’s flooded alleys, the Morris paradigm uses the water medium to eliminate confounding olfactory trails and ensure continuous locomotion, requiring the animal to navigate using an internal spatial map to locate a submerged, invisible escape platform.

Furthermore, the principle of motor equivalence has shaped human spatial navigation research using advanced neuroimaging technologies. In modern cognitive neuroscience laboratories, human spatial memory is routinely evaluated using desktop computer monitors, immersive virtual reality (VR) headsets, and functional Magnetic Resonance Imaging (fMRI) scanners. In these virtual testing environments, human participants navigate complex simulated labyrinths using simple keystrokes, handheld joysticks, or thumb trackballs. The fact that humans can master an intricate spatial environment using minuscule thumb movements on a controller, and then navigate that same environment on foot in real life, is a direct manifestation of the motor equivalence demonstrated by Macfarlane’s rats over nine decades ago.

12.2 Broader Implications for Motor Control and Rehabilitation

The enduring legacy of motor equivalence extends far beyond spatial navigation, playing a vital role in modern neurological rehabilitation and motor recovery. When an individual suffers a stroke or traumatic brain injury affecting the primary motor cortex or corticospinal tract, the localized neural pathways that once controlled specific muscle effectors are often severely damaged or destroyed. Modern neurorehabilitation protocols—such as Constraint-Induced Movement Therapy (CIMT) and functional neuroplasticity training—leverage the brain’s capacity for motor equivalence to reorganize these motor pathways. Because the high-level representation of an action’s goal remains intact within associative and premotor cortical networks, therapeutic interventions can train adjacent, undamaged neural circuits or alternative muscle groups to achieve the same behavioral objective.

Nowhere is the principle of motor equivalence more transformative than in the engineering of modern Brain-Computer Interfaces (BCIs) and advanced neural prosthetics. Contemporary neuroprosthetic systems decode movement intentions directly from the motor cortex, posterior parietal cortex, or premotor areas of paralyzed patients. Because these higher-order cortical regions encode abstract, goal-directed plans rather than isolated muscle twitches, the BCI system can translate a patient’s mental intention (such as “reach toward the cup”) into the mechanical movement of an external robotic arm, a computer cursor, or a powered exoskeleton. In these advanced neuro-engineering systems, the patient’s intended goal is executed through artificial, synthetic effectors that bear zero physical resemblance to biological human limbs—the ultimate technological realization of the effector independence first demonstrated by Macfarlane’s swimming rats.

12.3 Concluding Synthesis: D.A. Macfarlane’s Place in Psychological Science

Donald A. Macfarlane’s 1930 experiment with swimming rats remains a transformative landmark in the history of psychology and cognitive neuroscience. Conducted at a time when mechanistic, peripheral behaviorism seemed poised to reduce the study of the mind to a rigid catalog of mechanical reflexes and muscle twitches, Macfarlane’s research demonstrated the inadequacy of such radical reductionism. By showing that an animal could master a complex spatial maze through terrestrial running and then immediately navigate that same maze while swimming through deep water, he delivered unambiguous proof that spatial learning is an abstract, central, and effector-independent cognitive process.

Macfarlane’s swimming rats helped rescue the concept of internal mental representation, providing foundational empirical support for Edward Tolman’s purposive behaviorism and pointing toward the later discovery of hippocampal cognitive maps. His work proved that even in non-human animals, behavior is organized from the top down by goal-directed plans rather than assembled from the bottom up by unthinking muscle chains. Nearly a century later, the insights derived from this elegant study continue to inform our understanding of how living brains navigate the world, how damaged neural systems can be rehabilitated, and how artificial minds can be built to act with purpose and flexibility.

Conclusion

The journey from John B. Watson’s mechanistic “muscle-twitch psychology” to contemporary computational cognitive neuroscience reflects psychology’s enduring effort to understand the relationship between the body’s physical movements and the mind’s internal representations. Donald A. Macfarlane’s 1930 flooded-maze experiment stands as a watershed moment along this intellectual journey. By forcing laboratory rats to trade their terrestrial running strides for aquatic swimming strokes within an identical spatial environment, Macfarlane established the reality of motor equivalence, proving that goal-directed navigation is governed by higher-order cognitive architectures that operate independently of peripheral muscular mechanics.

This critical finding demonstrated that organisms do not navigate the world through blind chains of reflexes, but through internal spatial models that map out their environments. Macfarlane’s methodology dealt an early, fatal blow to radical stimulus-response peripheralism, providing crucial empirical validation for Edward Tolman’s concept of the cognitive map and foreshadowing the discovery of hippocampal place cells and modern neuroprosthetic control systems. In an era when behavioral science was in danger of reducing the organism to an unthinking automaton, Macfarlane’s swimming rats proved that life possesses a remarkable capacity for flexible, purposive, and adaptive action—an insight that continues to shape our understanding of biological and artificial intelligence today.

References

• Bernstein, N. A. (1967). The co-ordination and regulation of movements. Pergamon Press.
• Hull, C. L. (1943). Principles of behavior: An introduction to behavior theory. Appleton-Century-Crofts.
• Hunter, W. S. (1929). The sensory control of the normal rat in the maze. The Journal of Genetic Psychology, 36(2), 181–237. https://doi.org/10.1080/08856559.1929.10532194
• Lashley, K. S. (1929). Brain mechanisms and intelligence: A quantitative study of injuries to the brain. University of Chicago Press.
• Lashley, K. S. (1951). The problem of serial order in behavior. In L. A. Jeffress (Ed.), Cerebral mechanisms in behavior: The Hixon symposium (pp. 112–136). John Wiley & Sons.
• Macfarlane, D. A. (1930). The role of kinesthesis in maze learning. University of California Publications in Psychology, 4(18), 277–305.
• Morris, R. G. M. (1981). Spatial localization does not require the presence of local cues. Learning and Motivation, 12(2), 239–260. https://doi.org/10.1016/0023-9690(81)90020-5
• O’Keefe, J., & Dostrovsky, J. (1971). The hippocampus as a spatial map: Preliminary evidence from unit activity in the freely-moving rat. Brain Research, 34(1), 171–175. https://doi.org/10.1016/0006-8993(71)90358-1
• O’Keefe, J., & Nadel, L. (1978). The hippocampus as a cognitive map. Oxford University Press.
• Thorndike, E. L. (1911). Animal intelligence: Experimental studies. Macmillan.
• Tolman, E. C. (1932). Purposive behavior in animals and men. Century Company.
• Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological Review, 55(4), 189–208. https://doi.org/10.1037/h0061626
• Watson, J. B. (1913). Psychology as the behaviorist views it. Psychological Review, 20(2), 158–177. https://doi.org/10.1037/h0074428
• Watson, J. B. (1914). Behavior: An introduction to comparative psychology. Henry Holt and Company.