This is the post I was trying to write earlier this month before I realized that what was supposed to be a parenthetical explanation of my Mix-o-matic playlist was spiraling out of control and needed to be quarantined in its own post. If it wasn't already clear from that post, I spend an absolutely inordinate amount of time thinking about the star ratings I assign songs in iTunes. That's partly because I rely on ratings to fuel things like the Mix-o-matic, and partly because I am just kind of obsessive that way. So obsessive, in fact, that I feel obliged to expose my thought process here.

In my last post, I said I could boil the meaning of all my star ratings down to one or two words. To recap:

• ★: (Barely) Tolerable
• ★★: Indifferent
• ★★★: Good
• ★★★★: Very Good
• ★★★★★: Perfect

As I said, I like the simplicity of these definitions, but I also feel like they could bear with some additional elaboration. (As a side note: a couple years ago, I tweeted a somewhat different, but similarly terse, set of star definitions. That one's actually an abbreviated version of the rating system I espoused for my books on LibraryThing, and while it has much in common with my song ratings, there are enough differences between how I think about individual songs vs. entire books that I am going to stick to the above terminology here.)

★: Tolerable

One-star songs are songs I basically don't ever want to listen to (the first rule in almost any new smart playlist I create is: Rating is greater than ★), but which don't merit permanent deletion from my library. Actually, many of these probably do merit deletion, but because I am a pack-rat at heart and increases in computer storage capacity have finally outpaced my ability to add new music, there's not much that I cannot find some excuse for keeping around. The most common reasons I give myself for keeping one-stars include:

• Historical interest: I have, for example, ripped a few compilations of recorded poetry, and there are a lot of tracks from those that I find pretty unbearable but keep on hand just in case I am suddenly press-ganged into guest teaching some literature class. Granted, the likelihood of this has become vanishingly small, but it doesn't hurt to be prepared.
• Interest to others: A fair number of these songs are ones that I am, in theory at least, ‘holding’ on someone else's behalf. A few were puzzling favorites of my ex-wife (Partridge Family Christmas Album, I am looking at you); others are insipid children's songs that even my kids have outgrown. I guess I am really only keeping these out of nostalgia.
• Completism: Although I spend virtually all of my listening time in ‘shuffle’ mode, I remain, at heart, a believer in the album as a musical form. I still mostly buy complete albums (albeit digitally), and when I do, I feel compelled to keep all the tracks from the album, so that I could (again, in theory) listen to it in its canonical form. While some of these are truly bad, a larger segment of this category are actually non-songs: the intros, outros, interstitial skits, and other filler tracks that some artists seem to find irresistible.

In short, the one-star category is basically the junk-drawer of my iTunes library, a place to shove things I almost certainly do not need, but can't quite take the step of actually throwing away. It's of some consolation to me that it's a very small group (fewer than 500 tracks in a library of over 17,000).

★★: Indifferent

As I discussed at length in my previous post, two stars is my default rating for everything new that enters my library. Consequently, it has ended up being a rather large (4,490 tracks as I write this), eclectic, and ever-shifting category. For the new arrivals, the rating doesn't really mean anything except that I haven't listened to a song much. Of the songs that have remained twos for a while, some are pretty decent songs that for one reason or another haven't managed to catch my attention (it is to catch these that I've engineered the Mix-o-matic to continually cycle through the two-star category). Some are basically-OK songs that have that one little thing (a little too jokey, a little too screamy, a little too sappy, an unfortunate string arrangement, an ill-advised children's chorus) that finally gets on my nerves. Many are truly indifferent: I wouldn't go out of my way to listen to them, but I don't mind if they pop up once in a while. And some are teetering on the edge of being consigned to one-star oblivion.

★★★: Good

Three stars is where songs start getting interesting. A three is my baseline ‘I like this’ rating. Either I am fairly good at predicting whether I'll like an album before I buy it, or I am pretty generous with this rating, because it is, by far, the biggest category (currently 8,058 tracks, almost half my library). As in the two-star category, there is some upward and downward mobility here. If I like a new song right off the bat, I won't hesitate to immediately bump it up to a three rating. Sometimes it doesn't stand up to repeated listening and winds up getting demoted back into the twos. On the other hand, it's not at all uncommon for a song to keep growing in my estimation and ascend to a higher rating. Obviously, though, a great many songs stay threes and that's a perfectly respectable grade in my book.

In the context of albums, I generally consider anything that is not mostly threes (or above) to be a disappointment. A couple ‘indifferent’ tracks are acceptable, particularly if they are balanced by some really great songs, but if an album is approaching a 1:1 ratio of twos and threes, I am going to somewhat regret the purchase. More twos than threes is the kind of performance that will have me thinking long and hard before buying anything else by that artist.

★★★★: Very Good

Four-star songs are those that are more than just likable. These are the songs I remember by name, the ones I would put on a mix tape for someone, or select for a hypothetical Greatest Hits compilation for an artist (four stars is also, incidentally, the threshold I use for loving a song on Last.fm). The paths to four-star status are varied. It's almost never granted on first or second listen, although a strong first impression always helps. It can take weeks, months, even years for a song to really grab hold of me. Many of these songs are what you might expect as ‘hits’—crowd-pleasing anthems, tear-jerking ballads, technical tours de force—but sometimes it will be a subtler touch that nudges a song over the edge from merely good to memorable. One of the oddest examples I can actually put my finger on is a single line in The Mountain Goats, Slow West Vultures (from We Shall All Be Healed, 2004): Get in the god-damn car. Out of context, it doesn't seem like much, but at that precise moment in the song, delivered in the precise way it's delivered, I get a jolt from it every single time. It's that jolt, or something like it, that I am looking for. One commonality among four-star songs is that once the rating has been earned, it is very rarely lost. Even if a song doesn't always have quite the same impact on me as the day I gave it its fourth star, I tend to respect the original moment. For all this implied exclusivity, this category has grown quite populous: 2,688 songs, about 16% of my library.

To bring this back to albums, I generally want to find at least one memorable song on any album I buy. A ‘pure’ three-star album is OK, but not something I'm going to be talking about or recommending to others. It seems to take at least a couple four-star tracks to earn a mention in my year-end album wrap up and several to be in the running for my Album of the Year.

★★★★★: Perfect

And then there are the fives, the Perfect Songs, an exclusive club currently standing at 166 members. These are the songs that I know by heart, that have been objects of obsession, that I've listened to on repeat until they were burned into my psyche. Usually, a song will have to put in some time at four stars before I'll start considering a fifth, and even then, it's not an impulsive decision (about the only time a song can be fast-tracked to five-star status is when it's a beloved song from my vinyl past that's haunted me until I finally found it in digital format, which pretty much speaks for itself). Some of these are pretty recognizable ‘classics’ (e.g., Johnny Cash's Ring of Fire or John Coltrane's My Favorite Things), but others are there for extremely personal, idiosyncratic reasons (Derrick & Patsy's Housewives' Choice, which was ‘our song’ when I still belonged to an ‘us’). I haven't done any comprehensive genre analysis, but my feeling is that this category is more diverse than my overall library. It's still dominated by indie and punk rock, but there are entries from all over the musical map: blues, country, folk, funk, jazz, pop, rap, reggae, soul, even opera.

Since five-star ratings are so rare, an album containing even one of these is always notable, and multiple five-star tracks merit discussion as one of the All Time Greats (indeed, there are only a handful of artists whose careers [at least as represented in my library] have produced multiple perfect songs). There is one extraordinary outlier I can't help but mention: the 1976 folk album Have Moicy!, credited to Michael Hurley, The Unholy Modal Rounders, and Jeffrey Fredericks & The Clamtones, with an unprecedented ten (of thirteen) tracks on this list. (It is a record that defies explanation, although Robert Christgau gives it a try in his A+ rating. All I can say is that I listened the hell out of it my freshman year in college, and it was so strange and wonderful that it utterly changed my conception of music. Sadly, no one I've played it for since has been nearly as impressed with it.) Have Moicy! is not, however, the highest-rated album in my collection. That distinction would have to go to the Firesign Theatre's magnum opus, Don't Crush That Dwarf, Hand Me the Pliers (1970), which, consisting of only two tracks (This Side and The Other Side), both rated at five stars, is technically the only ‘Perfect Album’ I own (and, perhaps, ever made).

So why should I care?

Good question, and one I don't really have an answer to. I care about this, I think, because I am fascinated by the process by which art commands our attention and insinuates itself into our lives. Music has been a big part of my life for almost as long as I can remember, and with iTunes I have a way of observing and recording my experience of it more closely than with other media. I don't really think aesthetics can or should be quantified, but this obsession with star ratings forces me to think a little harder about how I am responding to music and why. In observing myself handing out stars over the years, I have noticed a few patterns. Brand new music (bands I am listening to for the first time) usually needs to work a little harder to catch my attention. There are a lot of ways to do this—good riffs, clever lyrics, a distinctive vocal style, etc.—but without something, the songs will just fade into background sound. At the same time, I generally want to find new things to like, so if a newcomer's song does have that hook, I may be quicker to give it a good rating than if it were by some more established band. Conversely, artists who have a proven track record with me get something of the opposite treatment. I am going to pay attention to them anyway and will generally give them the benefit of the doubt, but I also have higher expectations for them, and if I feel they are not living up to those, I may well be less generous than if they were some unknown band. Finally, in genres that are outside my core interest of indie, punk, and classic rock, I feel less competent to make finer distinctions in quality and tend to gravitate toward a ‘lazy three’ unless it provokes a particularly strong positive or negative reaction. I'm not especially proud of that last point, but there you go.

Still, there is much about how music affects me that remains utterly mysterious to me. It certainly doesn't help that my understanding of music theory is limited to what I gleaned incidentally from eight years of high school and college marching band, but even with a better technical grounding, I am not sure I would be any closer to being able to explain what the Rolling Stones' Sympathy for the Devil and Mozart's La ci darem la mano have in common that makes them both Perfect Songs. I find myself falling back on a to a conclusion I've stated before, ‘art is a fucking miracle’. That may be something of a cop-out, but I suspect that part of what ultimately makes great art—be it music, literature, painting, or what have you—compelling is its ability to defy rational explanation.