Music and language are related in so many ways that it is necessary to categorize some of those relationships. I will then address each category in turn.
First, there is the seemingly never-ending debate of whether music is itself a language. The belief that music possesses, in some measure, characteristics of language leads people to attempt to apply linguistic theories to the understanding of music. These include semiotic analyses, information theory, theories of generative grammar, and other diverse beliefs or specially invented theories of what is being expressed and how. This category could thus be called "music as language".
A second category is "talking about music". Regardless of whether music actually is a language, our experience of music is evidently so subjective as to cause people not to be satisfied that their perception of it is shared by others. This has led to the practice of attempting to "translate" music into words or to "describe" musical phenomena in words, or to "explain" the causes of musical phenomena. The sheer quantity of language expended about music is enormous, and includes writings and lectures on music history, music "appreciation", music "theory", music criticism, description of musical phenomena (from both scientific and experiential points of view), and systems and methods for creating music. These approaches may include the linguistic theories of the first category, as well as virtually any other aspect of the culture in which the music occurs: literary references; anecdotes about the lives and thoughts of composers, performers, and performances; analogies with science and mathematics; scientific explanations of perception based on psychology and acoustics; poetry or prose "inspired" by hearing music; even ideas of computer programs for simulations or models of music perception and generation.
A third category is composed of a large number of "specialized music languages". These are invented descriptive or explanatory (mostly written) languages, specially designed for the discussion of music, as distinguished from everyday spoken language. The best known and probably most widely acknowledged specialized music language is Western music (five-line staff) notation. Myriad others can be found all over the world, ranging from guitar tablature to computer-readable protocols (e.g., MIDI file format).
Not only is the role of language in the learning and teaching of music important, but the study of the role of language is important, as well. "What we talk about when we talk about music" is a matter that is too often taken for granted and too little investigated.
Language can be secret. Words and phrases are continuously evolving, and after a generation or two, a word can come to be used rather differently than the way our ancestors used that specific term. Many situations influence the evolution of language. Words and descriptions must be found for new inventions and new institutions. English is the native tongue of many different countries, and musical usage has evolved in different ways in each English-speaking country. Many of the new terms and usages from a number of different musical cultures thus become accessible to all users of English musical language. English speakers borrow terms from other languages, and immigrant groups in English-speaking countries insert a healthy amount of new terms into their adopted societies. A peculiar brand of musical English is evolving within the European Union. One musical institution can come to have several designations or one term might come to have several meanings.
Consequently, the English musical language is characterized by frequent use of common words with uncommon meanings, tautologies, translation of musical designations, use of Old English and Middle English words, Italian words not in the general vocabulary, terms of art, Latin terms and many others.
Music, as a cultural product and symbolic form, is part of social life. “As organized sound, it expresses aspects of the experience of individuals in society.” John Blacking (1973).
Yet these very characteristics also present the enigma of music, which seems neither to want nor to be able to say anything. This is the paradox within music’s very expression: without referring to any visible image of the world, music nonetheless reveals something about the world, and, free of any referential ties, its language depends on something other than itself.
What is interesting, the phenomenon of sound, ordered by sound system, does not become music and that every musical creation implies a relationship between the universe and human inner world.
Including for these purposes, music was often accompanied by language. The texts of music surprise with their variety, which are completely satisfies all the requirements for the most accurate displaying, transmission and solution of musical tasks.
In the Late-Medieval and Renaissance periods and especially with regards to sacred music, the text was supposed to be above the music in every way. There were several cycles lasting a generation or two where a new idea about how to bring out the text would pop up, composers ran with it and the music soon became more decadent and the text harder to understand. This was particularly a problem in the church, where the mass texts became so buried in florid lines and massive amounts of counterpoint that it was nigh impossible for a parishioner to follow along. In fact, the popular legend attached to Palestrina's famous Missa Papae Marcelli (for contrast, listen to some music by either of the Gabrielis) is that he wrote it so that the church wouldn't ban music all together! He proved that music could be both understandable, appropriately secondary, and yet still beautiful and singlehandedly changed the Pope/Council of Trent's mind. Things likely didn't go down quite like that, but the gist is accurate.
If we focus on earlier periods in history - starting with the era of Josquin des Prez, towards Monteverdi and his contemporaries and followers and up until and including Handel - text was perceived as more important than the music in all aspects, and definitely in theory, if not sometimes in practice. Music of the "seconda prattica" was almost without meter (some researchers would say definitely without meter), built almost solely around the text and how well it can be "declaimed".
Another iteration of this happened right at the beginning of the Baroque era--the era we associate with over the top embellishment actually began in part to take things closer to the text! This mode of thinking gave way to more lyrical sensibilities in the late Baroque era (such as Rameau, Vivaldi, Purcell) but they still associated a certain musical palette with certain words and wrote their music based on this musical language (so that, if somehow transported to listen to Rossini's operas, they might have been delighted but completely at a loss to understand the relation between the text and the music).
The new style called monody or seconda pratica (second practice vs the prima pratica of the Renaissance) was intended to take things closer to human speech rather than more abstract text-setting. Listen to a few of Montiverdi's later madrigals, with "Cruda Amarilli" being one of the touchstone pieces studied today. Contrast with Gesualdo or Lassus. A really fun contrast is in Montiverdi's opera L'Orfeo, in which Orpheus sings the incredibly beautiful "Possente Spirto" to try get into Hell to go after Eurydice. It's highly ornamented and "high art" style...but doesn't work. A more simple, monodic plea does.
The superiority of the text persisted in public and critical thinking even past the point where text dictated the actual music: in the time of Mozart, the composer was still perceived as a "workman" who had been contracted to fashion out the actual architectural designs of the librettist (to point: DaPonte wrote of writing his operas, not Mozart's).
When the Romantic Movement caused composers to seek out their personal "genius" and create increasingly miniature structures, text did truly become secondary to music (to its arguable disadvantage).
As things went on, I think whether text is secondary or not often depended on genre too. For instance now, if a composer sets a Kyrie Eleison, everyone knows the words, so he can be quite adventurous and allow the text to just be a vehicle for the sounds. However, in a solo song, the music is likely to depend greatly on text for its direction. They really are partners, though, and a trite text can be greatly elevated through good music- e.g. Der Winterreise. The text appeals to the rational side of our brain and the music the emotional.
The opera is essentially a fight over which the more important element was: music or words. At the end, they agree to work together, but music is assigned to all the "serious" music, whereas text gets to have prominence in comedy.
A century and a half later, Richard Strauss wrote his opera Capriccio. Capriccio has the exact same subject as Prima La Musica, except that now at the end, they are true equals--it is deemed impossible to choose between them (though music gets the last word).
In opera performance, even up until the 1940s, it was not uncommon to see opera performances in which people on stage were singing in two different languages, each completely unaware of what the other was saying.
I think it's fair to say that over the course of these 150 years, there was only a very slight shift towards text. Shortly after Strauss's death, however, we start to see signs that text is becoming more important. For one thing, performances all take place in the same language (and not just any language, but it becomes increasingly unacceptable to translate to the vernacular: opera is performed in the original language essentially exclusively). Sets and costumes become more about story-telling and less about providing a context for some beautiful music.
Today, text is at least as important as music in performance. The music is still extremely important, obviously, but communicating drama is the main goal of the modern opera house. As the conventional audience becomes more text-savvy than music savvy (which was not always the case), this drama becomes most-easily expressed in text.
Aside from opera libretto, most classical vocal/choral music uses sacred or secular texts which already exist and can and do continue to function without any music at all. The point of setting them to music is to illustrate/interpret/reflect on/present the text through the prism of the composer's sensibilities. The music is inspired by the text, but the added value in the new composition is the music which now becomes primary.
Natalia Rojcovscaia-Tumaha
from Master Degree Thesis "Structural and Semantic Analyses of
Musical Terms in English and Russian. Russian Translation Strategy"
Institute of International Relations of Moldova