There is no doubt that pitch intervals are crucial for cognition and memory. For instance, Deutsch (1969) claims that it is easier for us to remember the absolute pitch of the each and every component tones, and, in fact, most people traspose tunes so easily that it is extremely difficult for them not to do so. In most melody studies, therefore, the focus has been on pitch organization. However, pitch intervals are not the only material in a melody and even a simplest pattern of a melody is not merely a collection of intervals. That is, although pitch intervals are very important in a melody, there are other factors as well which must be involved in forming a melody, such as rhythm, note-grouping and harmonic background. The basic assumption underlying the present paper is that music is more than a "collection" of independent elements, but is a "fusion" of those, and, therefore, the separation of one parameter from others (e.g. the extraction of pitch material) is only theoretically, not perceptually, possible. In a melody, for instance, it is impossible for listeners to hear only pitch intervals at one time and hear only rhythms at the next time. People perceive a melody as "a continuous single entity" and that is the very notion of "melodic contour." Then, what is the smallest unit in a melodic contour? From the viewpoint that a melody is not a static object, but a dynamic process through time, there cannot be any independent symbol which can be seperable from its context. In a Mozart melody, for instance, a leap can be a big event and provoke "tension" to the listners. The same progression within a different context, say,in Webem's pointilistic works, the same-interval leap would not provoke any tension. One does not listen note by note or interval by interval. The minimal perceptual unit of the first movement of Beethoven's Fifth Symphony would be the initial four-note motif rather than the individual pitches or interval in the motif. In most melodies, there are such patterns of progression that replicate sequentially. In the present paper, the sequential replication of pitch movement (up and down) is called a "pitch pattern," and that of time organization (long and short), "time pattern," These two respective patternings are essential in a melody, but one does not perceive them as separate. Sometimes there can be conflicts between the pitch and time periodicities that makes the melody "ambiguous" in terms of the clear distiction of phrases. These two patternings, as the wrap and woof, weave the melodic fabric, and what we perceive is the weaved fabric (a product per se), not the weaving procedure or the rules of weaving. The current trend of the scholarship focused on melodies (in both music theory and cognitive psychology literature) can be summarized, in a word, as "pitch-centrism." "Melody" is often considered as "pitch succession," melodic interval is commonly regarded as "pitch interval." and melodic contour is generally confused with "pitch contour." The inter-dependency beween pitch and time domains demands the study on the interactive process between the two different variables, without which the perception of a melody is impossible, and only possible is either the perception of pitch or that of time.

In 1949, consolidating their previous efforts, Shannon and Weaver published The Mathematical Theory of Communication (i.e. Information Theory), which is, in short, a theory of measuring how effectively a message is transmitted from sender to receiver in a unit of time. Although the theory itself originates in communication engineering, it is also useful in the realm of music, especially in analysis, in twentieth-century composition, and in recent developments in aesthetics and psychology. The term "information," used in information theory, is to be understood not as used in ordinary usage, but as a technical term indicating the measured quantity of a message. Precisely speaking, the word "information" in communication theory rather refers to "the measured amount of information" or "a mathematical abstraction of information content" The notion of "entropy" was also introduced by Shannon and Weaver as a measurement of potential information contained in a message. By definition, the less predictable the communication system is (i.e. the more even the distribution of possible outcomes is), the higher entropy will be. In pitch class analysis, therefore, when the distribution is rectaguar (when all twelve pitch classes are evenly used), one can expect the maximum entropy (3.585). From 1956, many theorists have investigated musical works of various historical periods. Among them, R. Strauss' Lieder turn out to contain the most information, and nursary tunes yield the least information. In general, however, the results could be divided into several categories. That is, stylistic differences of music-historical periods in terms of the use of pitch-classes readily emerged. Although infromation theory is thus useful and even advantageous in music analysis, several questions can be raised with regard to the reliability of the data, the perceptual and cognitive validity of statistical results, and the contextual aspects of music which are totally ignored from the perspectives of statistics. Many scholars have attempted to overcome the technical and philosophical limatations and problems of information theory; as a result, recent research regarding music as a medium of communication has developed in both scope and quantity.