Decoding Melodic Intuition
through Cognitive Science
by David Fuentes
Executive Summary
This report examines Melodic Figuration Theory (MFT) and its profound alignment with established principles of cognitive science—particularly in music perception, memory, learning, and intuitive reasoning.
This alignment hinges on MFT’s core innovation: the identification of a finite repertoire of 36 melodic figures—primarily three- to four-note patterns that function as kinetic, semantic building blocks. The recurrence of these figures across eras, styles, and genres reveals a remarkable consistency in musical practice. But because this shared vocabulary was never formally identified or taught, composers must have deployed these patterns intuitively to achieve the strikingly similar expressive effects we observe. MFT, then, is the first theory first theory to discover and catalog melodic figures, thereby formalizing this previously unconscious dimension of musical craft to open an unprecedented window into the mechanics of composerly intuition.
The central argument of this report is that MFT does more than parallel cognitive-science findings: it actively bridges compositional practice and empirical research. Its foundational “melodic figures” operate as natural cognitive chunks, mirroring how the brain organizes musical information. MFT’s “melodic syntax” models chunk concatenation and phrase perception. Its principle of “intuitive acquisition” is grounded in implicit-learning research, which in turn underpins its “responsive logic”—a listener-expectation framework built on priming. MFT’s multi-dimensional approach to “melodic behavior”—encompassing harmonic melody, metric gravity/placement, trajectory, register, and contour—corresponds to the brain’s processing of distinct musical attributes via interconnected mechanisms. Likewise, concepts such as “metric gravity/placement,” “trajectory,” and “melodic gestures” correspond to a cognitive “intuitive-physics engine,” providing a novel framework for understanding melody as kinetic force. Finally, the theory’s clear distinction between “normal” and “modified” behaviors supplies a practical, figure-level toolkit for shaping expectation, directly linking melodic structure to emotion and meaning as described by Leonard Meyer.
It is important to note that MFT did not emerge from formal cognitive-science study but from over thirty years of intuitive investigation by a composition professor into observable melodic behavior. The framework was developed with an eye toward implementation in a compositional practice, ensuring that every inquiry served musical practicality and practice.
Building on these cognitive and practical foundations, MFT also establishes the basis for a new model of human-centered artificial intelligence. The proposed AI Composers’ Assistant, AICA (ī′kŭh), is a knowledge-based system built on MFT. Unlike existing AI platforms that rely on vast datasets and probabilistic next-note prediction, AICA’s strength lies in its multi-faceted understanding of compositional decision-making. By modeling intuitive pathways through MFT’s cognitive framework, AICA becomes a genuinely collaborative tool—one that enhances rather than replaces human creative judgment. Further discussion of AICA and its integration with MFT will conclude this report.
PART ONE: THE METHODOLOGICAL FRAMEWORK OF MELODIC FIGURATION THEORY
Melodic Figuration Theory has four interconnected tiers, each building upon the previous tier to build a holistic understanding of melody. While its highly practical mindset sets it apart from the more prevalent analytical paradigms that prioritize deep structure, the difference in perspective means that there is room for both. In other words, any distinctions MFT draws between itself and other approaches are not to challenge their authority, but to strengthen their reach. The distinctions MFT identifies highlight omissions—not flaws—where MFT helps illuminate what “traditional” tools can leave in shadow, especially regarding melody’s surface: the very space where so much of the intuitive, moment-to-moment craft of composition takes place.
TIER 1: The Vocabulary of Melody
Melodic Figuration Theory holds that all tonal melodies are constructed from a finite and
hared vocabulary of 32 distinct building blocks. These include 24 melodic figures, typically ranging from three to four notes in length; nine melodic dyads, which are two-note combinations; and three types of single tones. Examples of these melodic figures encompass a diverse range of patterns, shown on this table. A central tenet of MFT is that composers do not acquire this extensive vocabulary through deliberate, formal study. Instead, they “absorb it through years of singing, playing, and listening until the patterns become second nature.” The result is a hidden common language: figures that surface again and again, not by scholarly decree but through the collective muscle memory of generations of composers.
TIER 2: Five Dimensions of Melodic Behavior
Building upon the melodic vocabulary, Tier 2 explores how melodic figures behave across five crucial dimensions, providing composers with a practical toolkit of effects and guidance for their application.
1. Harmonic melody: While “traditional approaches to music theory” are hardly monolithic, there are some areas where they uniformly concur. Every approach to reckoning the relationship between melody and harmony follows the same either/or formula: categorizing each melodic note based on how it fits (or doesn’t) with the prevailing harmony—as either chord tone or non-chord tone. While traditional approaches excel at vertical/hierarchical analysis, Melodic Figuration Theory prioritizes emergent harmonic coherence within kinetic figures.
By observing characteristic behaviors of notes within each melodic figure, we can predict which ones are most likely to establish a figure’s harmony, even without knowing precisely which chord is in play. Consequently, we can appreciate each figure’s harmonic elements horizontally, in line with melody’s inherently diachronic flow. Distinguishing harmonic elements within the most basic patterns of melody explains how harmony naturally emerges from the figure itself, even without accompaniment. This phenomenon extends to reharmonization, which can involve reinterpretation as well as role reversal (switching chord tone and non-chord tone functions). MFT also addresses “harmonic divergence,” where figures intentionally create dissonance or tension against the prevailing harmony, yet still sound musically compelling due to their internal coherence.
2. Metric Gravity/Placement: MFT defines meter as “music’s essential pulse: a living, dynamic cycle of strong and weak beats that acts as a gravitational engine.” And depending on how the other musical elements interact with this cyclical engine of strong and weak beats, “a composer can generate sensations ranging from deep stability to playful confidence to intense agitation and more.” This is possible because musical elements “take on the characteristic lift and heft of the metric positions they inhabit.” Going further, MFT identifies various metric placements of melodic figures, including “normal alignment” (starting a figure on the beat, ending on an upbeat), “ligatures” (placing the last note on a beat), “pickups” (leading to an upcoming beat), and “straddled alignment” (positioning a figure’s middle note on a beat). This nuanced understanding of metric placement allows composers to intentionally shape momentum and feel of a melody as it unfolds.
3. Trajectory: MFT not only focuses on a phrase’s overall flight, but also on the distinct effect pathway between beats. Using each figure’s inherent “gait”—its unique kinetic profile—the composer shapes each beat-to-beat pathway with an ear toward controlling a melody’s local and cumulative pacing, flow, momentum, and expressive emphasis.
4. Register: MFT underscores the significant expressive and structural role of register in melody, a dimension often overlooked in traditional approaches. Composers consistently leverage register to create “registral dramaturgy,” establishing a “home” register and then stretching, sidestepping, or abandoning it for expressive effect. MFT also differentiates between “fixed” figures, whose registral span remains constant (between a 2nd and 4th), and “flexible” figures, which can expand or contract. Techniques explored include varying registral slope, isolating distinct registral zones within a phrase, and “overstepping the line”—marking an upper or lower registral boundary and then exceeding it for dramatic impact.
5. Contour: Melodic contour, the aural shape traced by notes, is understood by MFT as a multi-layered process involving macro (phrase and section), mezzo (figure and gesture), and micro (note-to-note) motions. These layers interact to shape the melody’s overall profile. MFT examines how contour is used for contrast between gestures, to maintain continuity through repetition, and through “melodic tailoring.” Tailoring can be “adaptive” (adjusting intervals to blend seamlessly, which often means changing figures) or “expressive” (preserving the contour but making a gesture “bigger” or “smaller” for emotional force). The theory also introduces the presence of “nested figures”—subtle, recurring 3-note figures embedded within 4-note figures that create nuanced resonances during melodic development.
TIER 3: Melodic Syntax
In Melodic Figuration Theory, a melodic figure, all by itself, is not yet music. To come to life, musical speaking, a melodic figure must combine with rhythm to form a “melodic gesture.” As such, the melodic gesture—not the melodic figure—is “the smallest intact unit of melody.” Going further, MFT’s melodic syntax articulates the “responsive logic” inherent in melody, presenting its course as a back-and-forth discourse. An initial melodic “proposition” (a gesture) naturally elicits a “response.” This tier defines the “melodic gesture” as “the smallest complete unit of melody.”
MFT identifies three fundamental ways to respond to any gesture, phrase, or section: repeat it, vary it, or contrast it. Drawing upon its deep understanding of melodic behavior across the five dimensions (Tier 2), MFT has cataloged an extensive repertoire of 75 distinct response strategies—25 ways to repeat, 25 ways to vary, and 25 ways to contrast a melodic gesture. This comprehensive catalog provides composers with a rich array of options (schemas) for developing melodic ideas.
TIER 4: Melodic Schemas
MFT defines a musical schema as “a recurring pattern or framework that composers, improvisers, and performers recognize instinctively,” then draw on to create or interpret melodies. Key characteristics of schemas include recognizable patterns, cognitive efficiency, flexible application, and cross-genre validity.
According to MFT, every element of music can be represented as a schema within this paradigm, such that learning music is, in essence, “learning schemas, by book or by ear.” MFT, through its keen awareness of the properties and behaviors of melodic figures, uses schemas in part to shine a light on “overlooked facets of melodies.” In addition to the 75 syntactical schemas mentioned above, others include:
1. MFT-specific schemas for integrating melodic figures within standard musical schemas. For example, using nested figures to embellish a cadential 6/4 melodic formula.
2. MFT-specific schemas for using each melodic figure. Each of the 24 melodic figures brings its own signature behaviors—abilities it excels at—while also sharing other traits with its peers.
3. MFT-specific schemas for using any of the five dimensions of melodic behavior. A thorough review of existing literature shows that composers already employ schemas to manipulate harmony, meter, trajectory, register, and contour—yet many of these remain undocumented. For example, MFT’s “Anda-2” schema observes how melodies often begin off the down-beat to shift metric weight onto the next strong beat (e.g., “My Funny Valentine”).
4. MFT-specific schemas for building phrases. A unique contribution of MFT in this tier is the concept of building “one-off phrase schemas,” which involves analyzing the distinctive syntactical behaviors of existing melodies and then imprinting these behaviors onto new melodic materials.
5. Schemas for breaking schemas. Resisting the common wisdom that “one learns the rules so one can break them later,” MFT maintains that creative deviation from established “rules” is not random but follows alternate, intentional routes to achieve specific expressive effects. By systematically mapping such deviations, MFT provides composers with structured pathways for innovation.
PART TWO: The Cognitive Foundation of Melodic Figuration Theory
The previous section outlined the four-tiered framework of Melodic Figuration Theory on its own musical terms. This section now provides scientific validation for that framework, demonstrating point-by-point how MFT’s practical, musically-derived insights reflect the brain’s own fundamental processes for perceiving, learning, and creating music. The following analysis reveals that the core tenets of MFT are not arbitrary but are deeply resonant with established principles of cognitive science.
A. Chunking: The Fundamental Units of Melodic Perception
Chunking is a foundational cognitive process that involves organizing smaller, discrete units of information into larger, more meaningful, and manageable fragments (Song & Cohen, 2014). This concept gained prominence from George A. Miller’s influential 1956 paper, which highlighted the brain’s limited working memory capacity—typically holding between five and nine chunks of information—and how chunking effectively extends this capacity (Miller, 1956). In the context of music, chunking is a pervasive cognitive activity, relevant whenever musical elements are conceived in groups of notes, rhythms, or articulations, whether during improvisation, sight-reading from notation, or passive listening (Chenette, 2022).
Musical chunks are perceived as holistic fragments of action and sound, generally lasting between 0.5 and 5 seconds (Godøy et al., 2010). The formation of these chunks can occur in two primary ways: through the recognition of familiar musical units or established schemata, such as a motif or a metric pattern, even in the absence of obvious breaks in the sound (Godøy et al., 2010). Crucially, the brain leverages pattern recognition to process these familiar structures as unified entities rather than isolated elements, significantly enhancing informational efficiency (Sloboda, 1985).
Melodic Figuration Theory’s melodic figures align with this cognitive model, presenting themselves not merely as potential chunks, but as the natural, pre-formed cognitive units that the human brain already uses to perceive, process, and generate melody. They represent the “bite-size” pieces the brain prefers for efficient processing, suggesting that MFT is not imposing an arbitrary classification system but is formalizing the very units of melodic thought that the human brain naturally employs. Further, MFT defines a melodic gesture as the “smallest intact unit of melody,” formed by combining a melodic figure with rhythm (Fuentes, MFT, Tier 3: From Figure to Gesture). This directly parallels the cognitive process of chunking, which involves breaking down long and complex sequences into more manageable fragments that can be “learned and then concatenated together to acquire the long sequence” (Song & Cohen, 2014). MFT’s progression from figures to gestures and then to phrases mirrors this hierarchical chunk assembly in cognitive processing. For a potential AI system, this would mean operating with the same fundamental cognitive units as a human composer, enabling truly collaborative interaction by directly addressing the composer’s “native language.”
1. Chunking as a Tool for Learning and Performance
Chunking plays a critical role in learning and memory. By segmenting complex musical pieces into smaller, manageable parts, the brain can process and encode information with greater efficiency (Dueck, 2023). In practical applications like sight-reading, chunking enables musicians to perceive notes in groups, which significantly reduces the cognitive load associated with processing each individual note, thereby enhancing both accuracy and musicality (Chenette, 2022), as well as motor skills (Pike, 2012).
The chunk types commonly referenced in studies include arpeggios, neighbor tones, and recurring rhythmic motives (Maddocks, 2010). MFT’s catalog of 36 melodic building blocks constitutes a more intricate and detailed system of chunks, though it similarly segregates them into three main types: arpeggio, scale, and neighbor figures. This aligns with research showing that while chunking is an intuitive process, conscious effort and the explicit identification of patterns can significantly amplify its effectiveness. Deliberately identifying chunks during practice has been shown to strengthen this innate capability (Chenette, 2022), just as proactive pattern recognition before playing enhances chunk utilization (Sloboda, 1985). This synergy between intuition and explicit knowledge mirrors MFT’s pedagogical approach. In fact, MFT functions as a pedagogical framework that structures this process, providing an overt, structured way to identify and practice melodic figures, thereby bridging instinctive musical understanding with controlled technical mastery.
2. Chunking in the Perception of Musical Structure
Beyond memorization and performance, chunking fosters deeper musicality by enabling musicians to grasp the “big picture” of a piece—not merely individual notes, but their contextual relationships, directional momentum, and expressive intent (Chenette, 2022). This is because chunking occurs on various hierarchical levels, which helps musicians organize musical material into a logical progression—the foundation for understanding musical form (Neuhaus, 2013).
This process of perceiving larger structures is supported by direct neurological evidence. The cognitive underpinnings of chunking extend to the perception of phrase structure—a crucial organizational mechanism for auditory processing (Knösche et al., 2005). Phrase boundaries are signaled by various structural markers, such as small breaks, the lengthening of notes, changes in pitch contour, or harmonic cadences (Knösche et al., 2005). Electrophysiological studies using EEG/MEG reveal a “closure positive shift” (CPS) in brain activity at these boundaries, mirroring prosodic phrase processing in speech (Knösche et al., 2005). Source localization implicates limbic structures (e.g., the hippocampus) associated with memory consolidation and attentional shifting (Knösche et al., 2005). This suggests the CPS reflects not just boundary detection, but the cognitive “packaging” of completed phrases into unified chunks while reallocating attention to new material (Knösche et al., 2005). This neurological process of “packaging” phrases provides a direct perceptual parallel to MFT’s compositional syntax. The theory’s “responsive logic”—in which an initial gesture is answered by one that repeats, varies, or contrasts it—provides a practical framework for creating the very structural markers and coherent phrases that the brain’s CPS mechanism is designed to detect and process. Cognitively, this relates directly to melodic expectation, where the brain constantly generates predictions about “what will come next” based on past experiences and recognized patterns (Nguyen & Yildirim, 2024). MFT’s catalog of 75 distinct response strategies, therefore, can be understood as a detailed toolkit for managing these inherent cognitive expectations and guiding the listener’s attention through melodic transitions.
3. Chunking in Compositional Practice
The evidence confirms chunking as a fundamental cognitive process governing both music perception and creation, from auditory intake to artistic output (Chenette, 2022). A bidirectional relationship exists between auditory chunking and the physical gestures of performance, indicating that chunking is a holistic mechanism supporting the entire creative cycle (Godøy et al., 2010). This provides strong support for MFT’s emphasis on “observable melodic behavior” as a basis for writing melody. If chunking is a natural cognitive process for producing music, then a theory that codifies these natural chunks offers a direct, cognitively aligned framework for composition.
For composers, this process often works in reverse from that of an analyst. While analysis often involves the top-down dissection of a composition into sections, phrases, and smaller components, MFT frames the creative process as a bottom-up construction: “arranging notes into figures, figures into gestures, and gestures into phrases” (Fuentes, MFT, Tier 3: Melodic Syntax). By providing a detailed vocabulary of these foundational chunks, MFT supports the composer’s craft in a direct way. It provides a framework for translating intuitive, passively absorbed knowledge into explicit and controllable skills for composers, thereby improving their craft. It is this codification of the creative cycle—from perception to production—that provides the robust foundation needed for an AI partner capable of true musical collaboration.
B. Gestalt Principles: The Laws of Perceptual Grouping
The question of how our brains naturally group notes into the precise, recurring patterns we call chunks leads us directly to the foundational insights of Gestalt theory. While most associate Gestalt with visual phenomena, its origins are profoundly rooted in music, directly affirming MFT’s core contention that we hear melody in meaningful, unified groups, not as a mere succession of individual notes. Christian von Ehrenfels introduced the concept of Gestaltqualität (“emergent property of form/shape/pattern”) in 1890, arguing that a melody is undeniably more than the sum of its sensory components. It is this emergent quality that allows a melody to be transposed to a new key, using entirely different notes, while still retaining its undeniable identity (von Ehrenfels, 1890, as discussed in Smith, 1988).
MFT’s understanding of melodic figures and gestures perfectly resonates with this principle. While other music researchers have deftly applied Gestalt principles to broader phenomena like phrase grouping, and Eugene Narmour’s Implication-Realization model provides a powerful note-to-note framework, MFT is the first theory to formalize the mezzo-level of intuitive melodic construction. It identifies a vocabulary of kinetic figures and shows how they are combined into gestures. This essential, “pedestrian” perspective illuminates entirely new ways to understand how music “holds together” and “moves” at the primary level of composerly intuition, opening new avenues for research and application.
The overarching principle in Gestalt theory is the Law of Prägnanz, which states that the mind tends to perceive reality in its most regular, orderly, and simple form (Koffka, 1935). At every level, MFT is consistent with this principle by defining every aspect of melody according to its “normal” (most simple and orderly) and “modified” behaviors.
Good Continuation: This principle describes our innate tendency to perceive separate objects as forming simple, continuous patterns rather than disjointed arrangements (Wertheimer, 1938). MFT’s catalog of melodic figures exemplifies this, especially as it emphasizes behavior at every turn. For example, regarding contour, figures like the Run exhibit a consistent direction while others like the Arc form curves. Regarding trajectory, a figure ending with stepwise motion normally continues moving by step into the next, maintaining a seamless melodic flow. And figures that end with a “gap”—a leap of a third—have a strong proclivity to fill that gap by moving to the next figure by step in the opposite direction to the gap’s motion.
Closure (Reification): This principle explains our predisposition to perceive incomplete forms as complete, mentally synthesizing missing units to create a unified whole (Wertheimer, 1938). In music, this explains how listeners intuit a clear harmonic underpinning when hearing an unaccompanied melody. MFT establishes compelling new criteria for this phenomenon by showing how the inherent shape of each figure makes it clear which notes function as chord tones, allowing the brain to “close” (fill out) what’s missing from the harmonic information.
Figure-Ground: This principle describes our ability to differentiate a dominant figure from its background (Rubin, 1921). One way MFT does this is by precisely identifying figures like the Pendulum and the Pendulum Auxiliary (Fuentes, MFT, Tier 1: Melodic Vocabulary), which inherently contain a fulcrum against which other notes pivot, establishing clear melodic motion against a fixed background. Another is by showing how figures like the Double Neighbor and Double 3rd have apparent commotion “atop” underlying motion.
Proximity: This principle states that elements close to one another in time and space tend to be perceived as a group (Wertheimer, 1938). MFT leverages this fundamental law through its highly developed approach to register. By analyzing the inherent span of individual melodic figures, MFT isolates distinct registral regions within a melody, providing a precise lens to understand how proximity shapes melodic coherence.
Similarity: Elements that share features, especially contour, tend to be perceived as a group (Wertheimer, 1938). In MFT, figures with similar contours stand in correspondence, so listeners readily hear them as related even when they aren’t identical. Every 4-note figure contains a nested 3-note figure, reinforcing these implicit links. Similarity also underpins “tailoring:” substituting one figure for another with a closely matched contour when harmony prevents direct repetition, thereby preserving continuity.
C. The Intuitive Physics Engine: The Perception of Melodic Motion
In his seminal work on musical forces, Steve Larson (2004) proposed a compelling framework for understanding melodic expectation. He argues that we perceive melody through metaphorical forces—gravity, magnetism, and inertia—that are rooted in our intuitive “experience of physical motions.” This frame-work aligns with the concept of an “intuitive physics engine” within cognitive science, which enables humans to make rapid and accurate inferences about the physical world (McCloskey, 1983). MFT’s emphasis on the kinetic and gestural nature of melody finds its scientific explanation in this cognitive capacity, especially through three key concepts:
1. Metric Gravity/Placement: If anything is “structural” in MFT, it’s the beat itself—a point of kinetic contact, like a foot touching the ground mid-stride. The beat isn’t a final destination, but a deliberate point within a cycle that energizes and propels the melody. MFT’s concept of metric gravity posits that the beat acts as a gravitational center. The placement of notes in relation to this metric pulse is a primary source of a melody’s kinetic energy and expressive character. Notes that land squarely on a strong beat feel stable and grounded; notes that arrive early or late create a sense of kinetic energy, like an object in motion that defies simple gravitational pull. This provides a scientific model for why MFT’s focus on normal and modified ways to align melodic figures within the meter is so crucial for shaping melodic character.
2. Trajectory: MFT’s focus on the figure as a kinetic entity provides a precise “mezzo-level” methodology for controlling the melodic surface. Such distinctions describe the very kind of information an intuitive physics engine processes. Its distinctions between “normal” (direct) pathways for reaching a “target” (the upcoming beat) and “modified” routes have cognitive referents in how humans move from point to point in the physical world. Also, it observes that the effect of linking one figure to the next hinges upon the type of motion between the first figures last two notes. For example, stepwise motion sounds a the end of one figure sounds smoothest when it continues moving by step. What’s more, MFT’s observation that the perceived effect of a leap is not merely about its interval size but about how it navigates the musical forces of metric gravity (the beat with its accompanying metric cycle) and inertia (the melodic line’s particular manner of motion and follow-through). For example, a dissonant leap made after landing solidly on a beat will sound elegant and intentional, whereas the same leap executed across the beat can sound clumsy and disruptive. Thus, the conclusion that for melodic leaps, “timing matters more than distance,” is a profound insight that cognitive science can now help explain (Fuentes, MFT, Tier 2: Trajectory).
3. Gestures: MFT’s central concept of the “melodic gesture” is overtly put forth to replace more standard terms like fragment, segment, cell, or motive—precisely because these technical terms inadvertently reduce the small parts of a melody to abstraction, stripping them of their inherent kinetic energy. In sharp contrast, “A gesture harnesses motion to convey meaning with immediacy” (Fuentes, MFT, Tier 3: From Motion to Meaning). This is as true for melodic gestures as it is for hand gestures. MFT codifies these kinetic gestures, providing a systematic approach to understanding and manipulating the very forces that drive melodic motion and listener expectation.
D. Implicit Learning and the Priming of Expectation
MFT explicitly states that composers do not learn its vocabulary through deliberate study, but rather internalize it through years of listening and playing until the patterns become second nature. This statement draws a direct and compelling parallel to the concept of implicit learning, where knowledge is acquired incidentally “without the aid of explicit instruction” or “conscious awareness” (Ettlinger et al., 2011). Implicit learning involves the brain’s sensitivity to “statistical regularities” and “transitional probabilities” within auditory sequences, enabling the acquisition of highly complex information through “mere exposure to its structure” (Ettlinger et al., 2011), (Tillmann & McAdams, 2004). This provides the cognitive mechanism for how composers assimilate MFT’s vocabulary and “absorb” the “consistent ways that composers shape and refine the basic building blocks of melody,” explaining why its figures appear consistently across diverse genres and historical periods.
The brain’s inherent reliance on “pattern recognition” (Sloboda, 1985) for efficient information processing further supports this. The human auditory system possesses a remarkable ability to establish “memory traces for invariant features” in the acoustic environment despite continuous variations (Tervaniemi et al., 2001). This process is facilitated by musical expertise, particularly for musicians who rely heavily on auditory information over visual scores (Tervaniemi et al., 2001). This explains how MFT’s “limited set of 36 common, small-scale patterns” becomes deeply ingrained and forms the basis of “observable melodic behavior” in compositional practice.
If implicit learning allows humans to internalize complex musical grammars and regularities without explicit instruction or conscious awareness (Ettlinger et al., 2011), then MFT, by systematically identifying and codifying these “observable melodic behaviors” and “consistent ways that composers shape and refine” figures, is essentially reverse-engineering this implicitly learned musical grammar, making the unconscious rules of melodic intuition explicit and computable. MFT reverse-engineers observable behaviors of melodic intuition into explicit schemas, making implicit patterns accessible for practice and computation while acknowledging intuition’s complex, non-symbolic nature.
This process of intuitive acquisition also sets the stage for priming, a form of implicit memory in which exposure to one musical stimulus automatically prepares the brain to process a subsequent, related stimulus (Ettlinger et al., 2011). Hearing a melodic gesture activates a network of associated concepts, making the mind more attuned to what might follow. This mode of expectation is not a conscious prediction, but a subconscious process that gives the mind a powerful advantage in comprehension (Ettlinger et al., 2011). This is the cognitive mechanism behind MFT’s “responsive logic” (MFT: Tier 3, From Motion to Meaning). An initial melodic gesture (the “proposition”) primes the listener’s brain with a set of expectations. The subsequent gesture (the “response”) must then manage those expectations in a musically meaningful way by choosing to repeat (confirming the expectation), contrast (subverting it), or vary (confirming it with a twist) (Fuentes, MFT, Tier 3: Not Just One Thing After Another). This transforms the act of hearing from a passive experience into a dynamic, anticipatory one, and it provides a compositional model for creating melodies that feel cohesive and compelling (Ettlinger et al., 2011).
E. Expectation and Meaning: MFT as a Meyerian Framework
The concepts of priming and implicit learning lead directly to a highly influential theory of music cognition: Leonard Meyer’s thesis that musical emotion and meaning are generated through the fulfillment and thwarting of expectation (Zangwill, 2021). For Meyer, a musical event has meaning because it points to and makes us expect another musical event. The satisfaction, surprise, or even frustration we feel in response to the music that follows is the basis of our aesthetic experience.
MFT’s understanding of melodic figures and gestures perfectly resonates with this principle. While other music researchers have deftly applied Gestalt principles to broader musical phenomena (e.g., phrases, phrase grouping, rhythm, and form), MFT is the first theory to leverage these laws specifically to systematize the mezzo-level building blocks of intuitive melodic construction—how its figures form melodic gestures and how those gestures interact. This gesture-level analysis illuminates entirely new ways to understand how music generates coherence and propels motion at the fundamental level of composerly intuition, opening transformative avenues for research and application. As such, MFT’s distinction between “normal” and “modified” behaviors offers a practical toolkit for managing Meyerian expectation:
Normal behaviors represent the most common and predictable patterns. By using them, a composer builds a clear set of expectations in the listener’s mind, creating a sense of stability and logical flow.
Modified behaviors, by contrast, offer proven ways to play with those expectations. By choosing to stretch, bend, or contradict a normal pattern, a composer can create surprise, add tension, and generate the very “meaningful deviations” that Meyer identified as the source of deep musical affect.
This is a crucial distinction. Unlike theories that focus on note-to-note (or “micro-level”) predictions, MFT’s approach is inherently Gestalt-based, operating at the level of the gesture—the smallest recognized perceptual whole. It is here, within the very DNA of the melody’s building blocks, that MFT provides a practical toolkit for manipulating the listener’s expectations, thereby generating emotion and meaning in a way that bridges the gap between the single note and the complete phrase.
This Meyerian framework extends from the intuitive to the conscious. While the manipulation of figures often occurs at an unconscious level, MFT’s schemas (Fuentes, MFT, Tier 4: Melodic Schemas) provide a framework for the deliberate, conscious management of expectation at a higher structural level. The concept of “schemas for breaking schemas,” in particular, formalizes the creative act of “haphazardly” deviating from established patterns. It is Meyerian expectation management writ large, transforming what might otherwise be random “rule-breaking” into a structured, artistic choice. This allows composers to move from the unconscious assimilation of melodic grammar to its deliberate and masterful control, bridging the gap between mere imitation and genuine originality.
PART THREE: Implications for AICA and Future Research
The detailed validation in this report—from the neurological basis of chunking and the perceptual laws of Gestalt, to the intuitive physics of melodic motion and the Meyerian creation of meaning—provides a robust scientific foundation for Melodic Figuration Theory. This foundation validates MFT’s central claim: by discovering and cataloging the finite vocabulary of melodic figures, MFT has formalized a previously unconscious dimension of musical craft, offering an unprecedented window into the mechanics of composerly intuition. It is from this unique vantage point that we can now appreciate the profound implications for Artificial Intelligence, particularly for AICA, which is directly built upon this framework.
AICA is founded on the strategic premise that a system built on MFT’s understanding of “how” melody is created is better equipped to emulate human compositional intuition than one built on the analysis of “why” a finished work is structured. This positions AICA not as a probabilistic “black box,” but as a system with an “explainable, elegant internal logic.” Its ability to “truly understand how melody works” is supported by MFT’s “reverse-engineered” implicit grammar, enabling it to offer “principled, creative possibilities” rather than algorithmic guesswork. For instance, MFT explicitly dissects consistent behaviors—like varying the slope of adjacent gestures or using nested figures during melodic development—that skilled composers use intuitively.
This directly addresses two of the most profound and often overlooked challenges in music creation, both for current AI models and, historically, within traditional theory: “relevance” and “development.” Relevance, as understood in Melodic Figuration Theory, speaks to music’s unique capacity to embody our deepest human experiences—how musical shapes resonate with our sense of “how things are or, even more profoundly, how they ought to be in our world,” allowing listeners to “catch what we’ve captured in sound and resonate with it.” This profound human connection, while acknowledged by composers and performers is a critical dimension rarely mentioned, let alone explored, in traditional theoretical discourse. Aaron Copland, for example, said “I have often asked myself what it is that I am trying to do when I write music. I can only answer that you compose because you want to somehow summarize in some permanent form your most basic feelings about being alive, to set down… some sort of permanent statement about the way it feels to live now, today.” (Copland, 1952) Most importantly, the rise of AI music makes the issue of relevance essential, as only humans can sense what sorts of musical gestures embody human convictions and experience. Indeed, many analyses within the MFT framework—especially those of songs—explicitly detail the connection between specific melodic behaviors and their profound human implications.
Similarly, development requires transforming musical material in ways that “feel intentional, rewarding, and human”—a nuanced process of intuitive judgment that current AI models, relying on statistical probabilities, often require some degree of arbitrary mutations, effectively “gambling, not composing.” MFT, with its deep understanding of observable melodic behavior and its reverse-engineered implicit grammar, spells out a far more extensive range of techniques for melodic development than previously available in traditional texts. This comprehensive knowledge provides AICA with the unique ability to present a composer with options based on genuine compositional intelligence, thus spurring her ability to imagine intriguing continuations and outcomes.
That’s why, for developers, this project represents more than just another music AI; it is an opportunity to pioneer a new class of creative tools. AICA offers the ambitious challenge of building true musical intelligence, moving beyond mere statistical syntax to engage with the semantics and expressive purpose of melody. Developers now have a theory that lets them codify a system that understands compositional intent, building a partner capable of strategic, musical reasoning, not just a pattern-matcher.
Furthermore, AICA offers a significant ethical advantage in an era marked by lawsuits over data scraping and algorithmic plagiarism. It requires no opaque training data and runs no risk of “lifting” copyrighted material, as its knowledge is derived from the foundational principles of melodic behavior rather than a dataset of finished works. This “clean” approach makes AICA a responsible, legally sound, and genuinely original system. Beyond mitigating risk, this foundation unlocks a profound secondary function: interacting with AICA becomes an educational experience. As users observe MFT’s concepts applied to their own ideas, they not only achieve better musical outcomes but also become better composers themselves. This transforms AICA from a simple utility into a lifelong pedagogical partner, opening a vast and untapped market for a tool that can both bolster and cultivate musical creativity.
References
Baker, D. J. “Chapter 3: Individual Differences.” In Modeling Melodic Dictation. Retrieved July 24, 2025. https://davidjohnbaker1.github.io/document/individual-differences.htm
Chenette, T.. 2022. “Formalizing chunks.” In Foundations of Aural Skills. Retrieved July 24, 2025. https://uen.pressbooks.pub/auralskills/chapter/formalizing-chunks/
Chenette, T. 2022. “Sight reading in chunks.” In Foundations of Aural Skills. Retrieved July 24, 2025. https://uen.pressbooks.pub/auralskills/chapter/sight-reading-in-chunks/
Copland, A. (1952). Music and imagination. Harvard University Press. https://archive.org/details/musicimagination0000copla
Dueck, R. W.. 2023. Developing a Progressive Chunking Technique in Music Literacy: An Analysis of the Method Books and Sight‑Reading Materials for Beginner Piano Students. [Master’s thesis, Université d’Ottawa]. / University of Ottawa. hdl.handle.net/10393/45642
Ettlinger, M., Margulis, E. H., & Wong, P.C.M. (2011). Implicit memory in music and language. Frontiers in Psychology, 2, 211. https://doi.org/10.3389/fpsyg.2011.00211
Fuentes, D. (2025). A brief introduction to melodic figuration theory. Figuring Out Melody. Retrieved July 24, 2025. https://figuringoutmelody.com/melodic-figuration-theory.pdf
Fuentes, D. (2020). The building blocks of melody. Figuring Out Melody. Retrieved July 24, 2025. https://figuringoutmelody.com/the-building-blocks-of-melody/
Godøy, R. I., Jensenius, A. R. and Nymoen K. 2010. “Chunking in Music by Coarticulation.” Acta Acustica united with Acustica 96 (4): 690–700. https://www.duo.uio.no/bitstream/handle/10852/36826/Acoustica-Godoy.pdf
Knösche, T. R., Neuhaus, C., Haueisen, J., Alter, K., Maess, B., Witte, O. W., & Friederici, A. D. (2005). Perception of phrase structure in music. Human Brain Mapping, 24(4), 259–273. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6871182/
Maddocks, R. 2010. “Effective Memorization In Music.” Getting There for Musicians (blog), July 2010. Retrieved July 24, 2025. https://intenseproductions.blogspot.com/2010/07/effective-memorization-in-music.html
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63(2), 81–97. https://psychclassics.yorku.ca/Miller/
Neuhaus, C. 2013. “Processing Musical Form – Behavioral and Neurocognitive Approaches.” Musicae Scientiae 17 (1): 109–127. https://journals.sagepub.com/doi/abs/10.1177/1029864912468998
Nguyen, B. K., & Yildirim, I. (2024). An intuitive physics approach to modeling melodic expectation. In Proceedings of the 46th Annual Meeting of the Cognitive Science Society. UC Merced. https://escholarship.org/content/qt8ds971xf/qt8ds971xf.pdf
Ochwani, J. (2025, May 2). Chunking: Learn its psychology and examples. Octet Design Solutions Inc. Updated May 2, 2025. Retrieved July 24, 2025, from https://octet.design/journal/chunking/
Pike, P. D. (2012). Sight-reading strategies for the beginning and intermediate piano student: A fresh look at a familiar topic. American Music Teacher. Retrieved July 24, 2025. https://test.mtna.org/downloads/Informed/AMT/AMT%20AOY/2013AOY.pdf
Sloboda, J. A. (1985). The Musical Mind: The Cognitive Psychology of Music. Oxford University Press. https://archive.org/details/musicalmindcogni00slob
Smith, B. (1988). Gestalt theory: An essay in philosophy. In B. Smith (Ed.), Foundations of Gestalt Theory (pp. 11–81). Philosophia Verlag. http://ontology.buffalo.edu/smith/articles/gestalt.pdf
Song, S., & Cohen, L. (2014). Impact of conscious intent on chunking during motor learning. Learning & Memory, 21(9), 449–451. https://doi.org/10.1101/lm.035824.114
Tillmann, B., & McAdams, S. (2004). Implicit learning of musical timbre sequences: Statistical regularities confronted with acoustical (dis)similarities. Journal of Experimental Psychology: Learning, Memory, and Cognition, 30(5), 1131–1142. https://doi.org/10.1037/0278-7393.30.5.1131
Zangwill, N. (2021). Music and emotion. In E. N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Winter 2021 ed.). Stanford University. https://plato.stanford.edu/archives/win2021/entries/music/
