A couple years ago, a friend of mine said "But if you have to know how to hear a type of music..." (He said it in response to what I said: that most people don't grow up hearing jazz, and thus don't know how to listen to it.) His opinion is pretty much standard: how to hear music is not something you learn -- it's deep, intrinsic. And, indeed, it certainly feels intrinsic: hear a minor chord and it's like someone's just fired a burst of electricity into the Sad region of your brain. What music makes us feel seems like a fundamental objective truth, as self-evident as the shape of a circle.
There's certainly no objective resemblance between a chord and the emotion it evokes (at least, none visible with our current understanding of neurology). Music is the ultimate in abstract art, pure untethered symbolism -- just like langauage: there is a similar lack of resemblence between the word cat and the image it invokes (namely, a cat). So perhaps, if we stretch our definitions a bit far, music can be called a language.
But why would it be a universal language? Certainly no other language is universal. Though the ability to match concepts to sounds (and some of the mechanics underlying that matching) seems to be universal, which particular sounds map to which concepts is a product of environment -- infants have the harwired ability to pick up language, but they pick it up from the people around them.
But enough rambling rationality. That stuff didn't get Plato very far (incidentally, Plato thought all music but patriotic hymns should be banned). Let's look for some concrete evidence.
To start, the Greeks. To them, the mixolydian mode (our major, but with a lowered seventh) was the saddest. The dorian (effectively our minor) they considered peaceful. (The lydian (common in neighboring Lydia) was thought to cause moral decay.)
More recently, in the 15th century, a common practice in local churches was for sacred music to be sung in complex arrangements with the main part (the one with the Godly words) on the very bottom, pitch-wise. Obviously, nobody could hear it there, so eventually the catholic church forbade the format (called a motet) -- at least, that's the way it was explained to my high school humanities class by the choir director, who seemed to have fallen out of 1957. Since at least the '60s (earlier in jazz and barbershop quartets), plenty of songs have put their melodies on the bottom or in the middle (some Beatles exapmles: Nowhere Man, In my Life, Because); more likely the church, steeped in a tradition of gregorian chant, simply didn't understand, shared with the choir director an inability to hear melodies that weren't on top.
Later (a couple hundred years ago), an emperor of Japan, in an effort to confer respectability and openness to the world, invited a world-class European orchestra to play Mozart for his court. After sitting politely through the concert, he and his aides discussed how unbearable it had been to withstand an hour and a half of the most horrendous dissonance they'd ever heard (a similar reaction (minus the politeness) greeted Stravinsky's Rite of Spring in reviews of its premier).
(Around the same time, Brahms was considered Mozart and Beethoven's equal, and Bach was mostly unknown.)
Jewish Kletzmer music uses a middle-eastern scale that sounds anguished and depressing to most american ears (there was an SNL sketch where a family stands in glaze-eyed horror before the Chanukah candles singing "burning burning burning..."). The deep, instinctual reaction of most people to that scale is not the same as the deep, instinctual reaction of kletzmer fans (I'm not much into kletzmer, but I've played it at the JCC, where people requested it again and again and danced exuberently around with smiles on their faces).
In 1945, following the lift of a WWII recording ban, bebop burst into the open. Most people find it jittery and dissonant; to its fans, it's not. Quoting myself:
...(not having been played much jazz during my formative years) when I listened to it I heard discordant noise...It wasn't until my trumpet teacher actually bought me a book of bop solos and made me learn to play some of them (which required that I go over them slowly and find all the hidden leaps that had eluded me at the lightning pace on the records) that I actually realized what it was I had been missing and reconfigured my hearing, that I could finally appreciate that famous Charlie Parker quote, Music is finding the beautiful notes.
Today hip hop is going through a golden age; great MCs are churning out great records (many of which are underground and don't get airplay). But in the wider world, it's mostly unnoticed; a lot of people consider rap, at best, poetry read over a beat, which is missing most of the point -- the MC's style, their inflection, the texture of their voice, is analogous to a musician's tone. More importantly, they're not just saying things; the lines form polyrhythms, a complex interplay of expectation against the background track, with pauses strategically offsetting lines and countervailing subrhythms implied by inflection. I don't pretend to fully understand hip-hop, but there's a hell of a lot there.
One possible objection to all this is that if music is divided into different languages, where are the lines? The Beach Boys are a bit like the Beatles and the Beatles are a bit like the Vines and the Vines are a bit like Nirvana, but the Beach Boys aren't like Nirvana. Ella Fitzgerald is a bit like Cinematic Orchestra is a bit like Daft Punk, but Ella Fitzgerald isn't like Daft Punk.
But actually, this is a phenomenon that occurs in language in exactly the same form (linguists call it a dialect continuum). From my textbook:
This is a situation where, in a large number of contiguous dialects, each dialect is closely related to the next, but the dialects at either end of the continuum (scale) are mutually unintelligible. Thus, dialect A is intelligible to dialect B, which is intelligible to dialect C, which is intelligible to dialect D; but D and A are mutually unintelligible. A situation such as this is found near the border between Holland and Germany, where the dialects on either side of the national border are mutually intelligible.
Taste in music is not just a matter of taste.
In response to bizza's w/u:
First, even if western pop is effectively a universal language, music itself isn't. The wide spread of one particular musical language doesn't make other musical languages intelligible.
That's a bit of a sidestep, though -- I think the real issue is whether (and where) inadvertant misunderstandings are taking place within the western musical framework -- or, to put it another way, how many different musical langauges there are (though I think there's a bit of metaphor shear; dialect continuum and all that) -- do people more often dislike a genre becuase they don't have experience with (or enjoy) the emotions it usually evokes, or becuase the emotions they think it evokes are unpleasant and not the emotions it's intended to evoke?
When you look at the attitudes of people who primarily listen to single genres, there's a symmetry -- a majority (or at least a large minority) find other genres either (A) repetitive/boring or (B) dissonant/painful or (C) both. (There seems to be a spectrum of music, with genres at one end more likely to be misheard as dissonant and genres at the other as repetitive. I think an (rather stretched) analogy can be made to the development of (and desensitization to) swear words -- see Kidman's thesis on the topic at http://www.gusworld.com.au/nrc/thesis/intro.htm ) Examples -- Folk, Punk, Emo and some Electronica (house, hardcore, etc.), are widely seen as boring (by each other!); opera, bebop, and modern classical music are widely seen as painful (hip-hop is often placed in both catagories, sadly).
A personal example: I bought OK Computer on sale in an airport in 9th grade. On first listen, it seemed ultra-dark, an incredibly bleak (though not entirely hopeless) concept album about alienation and oppression; I respected it as powerfully imagined stuff, but didn't like listening to it much. Now, 5 years later, it hardly seems dark at all -- there are all sorts of hidden modulations and what are in effect extended versions of leading tones; it's about alienation, but also about beauty, about sheer britpopy joy and transcendence. It wasn't that I didn't like it because it was dark, it was that it wasn't really dark; it wasn't that I didn't enjoy what it was saying, it was that never heard it. (One reason, for me, that the music-language metaphor seems a bit of a reach is this ease of misunderstanding. I might not understand someone speaking dutch, but neither am I likely to incorrectly think I understand them.)
Similarly, with regard to Kurt Cobain meaning Pain -- well, yes, sometimes he did; I'm under no impression that the Beach Boys and Nirvana attempt to communicate all the same emotions. But Cobain's songwriting was so wildly creative, with chord sequences totally unseen in most rock, that it can easily be misinterpreted as intentionally horrible when it's meant to be something else. (Incidentally, I think rock, jazz, and electronica are fairly equidistant, so to speak.)
tdent has pointed out that the idea of music being a universal language probably originated in centuries-ago Europe, which was divided into dozens of spoken langauges but united by a single, common musical langauge.