An excerpt from Chapter 3: The Notation of Melody and Harmony of my forthcoming book Music: Notation and Practice in Past and Present:
Modern Western musical notation has its beginnings in the liturgical music of the Roman Catholic Church. By ca. 800 C.E., a common repertory of plainchant had solidified for use in the vast array of worship services performed throughout Western Europe. A plainchant is an unaccompanied vocal setting of Latin or Greek prayers or hymns, sung at specific dates and times through the rotating calendar of the Catholic Church. Plainchant is assumed to have developed from Greek, Byzantine and Jewish antecedents, just as Christian worship evolved from Jewish Temple and synagogue traditions of the early Common Era. During Late Antiquity and the Early Middle Ages, there were many forms of plainchant, some of which continue in use alongside other forms of worship to this day. With the rise in importance of the Pope (i.e., the Bishop of Rome) in governance in the Western church, however, worship practices became standardized. Pope Gregory (dates) is often wrongly credited with inventing the system of notation that would become our modern system, but his role was actually to instigate a cataloguing and standardization of plainchant throughout the church year. Nonetheless, the term Gregorian chant is often used synonymously (and technically incorrectly) to refer to plainchant.
The key difference between Western musical notation and earlier systems is that the earlier methods generally relied on an alphabetic approach to notation in which every note available was assigned a specific symbol, just as an alphabetic writing system assigns a specific symbol to each phoneme in a language. An example of this type of system can be found in ancient Greece, where by the early Common Era, both a vocal and an instrumental set of symbols had developed on this principle. The system appears to have been fairly well-known, and used in a variety of settings, from the theatre (scraps of papyri give musical notation for lines from plays by Euripedes), to religious settings (the Delphic hymns), to funereal memorials, such as the Epitaph of Seikilos, found carved on a tombstone. In the Greek vocal system, symbols for individual pitches were simply written above or below the corresponding words. The Jewish Psalter, the hymnal for worship at the Temple in Jerusalem, used a simpler set of diacritical marks to indicate musical cues within the Hebrew text of the Psalms.
While music is often referred to as a “language,” especially by writers and speakers who gush over its emotional properties and spiritual meaning, the truth is that it is not. There are similarities between music and spoken language, but there are also great differences. Music alone cannot convey precise meaning. It cannot give directions or specific instructions. It cannot describe historical events in unambiguous detail. It can tell stories, but only in vague ways that often rely on a written or spoken description, or a preexisting understanding of the story being told.
At the same time, language is not music. Our ear for language is somewhat forgiving, as evidenced by our ability to understand each other despite individual variations in pitch, rhythm and articulation. It may be difficult to understand a speaker using our native tongue with a strong accent, but we still comprehend the same words. If a musical composition were changed by the same amount in the same aspects, we might be forced to understand it as a variation on the original piece, if not a completely different piece.
The alphabetical approach to musical notation, then, while a clear borrowing of a wondrous invention, leaves some things to be desired. The Greek and Hebrew systems did not have any way of showing duration, for the most part. Duration, in the form of tempo, rhythm and meter, is crucial to musical expression and to the reproduction of a musical idea, which is the purpose of notation in the first place. When notation using a musical alphabet is added to written language, the result is that a performer must either use both as simple mnemonic devices or attempt to simultaneously read both lines flawlessly.
For written language, the process of fluent reading depends on chunking—the mental grouping of characters into recognizable patterns. Fluent readers chunk on the word and phrase level, often making assumptions about the identity of a word based on its shape and first and last letters rather than perceiving every symbol. In music, this sort of chunking might be useful to a certain degree, especially in styles that abound in melodic cliché, but the ability to simultaneously perform this mental operation in two different alphabets would likely be very difficult to develop. Whether this is a reason that the Greek system never became as ubiquitous as our modern system of notation is debatable. More likely, in a largely illiterate society, the need for musical notation was simply not all that great.
 Over the last 400 years, the Roman Catholic Church has increasingly left behind plainchant, and music derived from it, as the regular mode of worship. Unaccompanied singing was appropriate in times and places where sustenance and survival were uncertain, but as Europe became more prosperous, the music used in worship became more complex. The Second Vatican Council of the 1960s introduced the most sweeping changes in Catholic worship by mandating worship in vernacular languages rather than Latin. With this shift, musical styles in worship also shifted to reflect folk and popular styles. Plainchant continues to be used in some religious communities and on certain occasions in worship, but its interest is now primarily historical for most listeners.
 Another example of chunking in everyday life is the splitting of phone numbers into groups of two, three or four digits, thus making them easy to remember. The American convention, for example, uses the three-digit area code, followed by a three-digit exchange, followed by the four-digit extension. To the switching computers at the telephone company, these groupings are irrelevant, but they are extremely useful for humans who wish to remember a phone number. Four digits seems to be an upward limit, and children often have to remember the last four digits of a phone number two digits at a time.