writing the record

By Robbie on August 19, 2011

This afternoon I have a few free hours on my hands and so should be working on songwriting for the record I'm doing this fall, but since I'm feeling lazy, I'll write about writing instead. Who benefits from all this self-focused sub-Montaigne bloggery is hard to judge; the idea that by herding my vagrant impressions into orderly sentences and paragraphs I will advance artistically is a little too optimistic to swallow. Instead, what I offer takes the place of what I wish I could read this afternoon -- Paul Simon's thoughts on writing songs for his next record, or Bill Callahan's, or Merle Haggard's, or Ron Sexsmith's, or any number of others including I am sure many non-white non-men though none springs to mind.

The short but tenacious tradition in popular music of composer-lyricist-performer-in-one has blighted our landscape in obvious ways, but from the songwriter's point of view there is the very strong advantage that he can lay out crucial performance details as he goes. He can cast the song with instruments and musicians to play them, decide the optimal tempo, lay out subsections to maximize meaning and drama for the listener, and so on; all toward an ideal outcome where the first rendition of the song -- his -- shows off the song in the most flattering way he can conceive and execute. I'm not sure if this has been the experience of other songwriters (and don't really want to discourage people from covering me, a rare enough compliment) but my typical reaction to other people's performances of my songs is acute disappointment. Someone had a different idea of the right tempo, structure, recording values, emotional color, even lyrics. Fine, maybe they improved on my idea; but candidly, I don't usually think so, and in any event I'm sure I'm too wedded to my song as originally pictured to judge. My persnicketiness aside, I think most other writers would confirm that a flawed first version is an awfully heavy shackle for a song to throw off. So when the writing is step one in a procedure that ends with yourself standing and delivering in a padded booth -- and when, needless to say, you are skilled with the relevant tools -- you are really in the driver's seat, with added control and confidence.

In preparing to record, and in the studio as well to the extent I can afford it, the surplus principle has always been dear to me as a quality control tool. If you have enough to throw away, at all points in the writing-to-recording chain (song ideas, song fragments, songs, recordings, mixes), the final product should only improve. Let's see how that's worked for me so far...

Record 1, Country Love Songs. Few unrecorded from the composition pool; five recorded titles unreleased; zero unreleased mixes. Hereafter I'll abbreviate LC for leftover compositions, LR for recorded titles, LM for mixes.

Record 2, South Mouth. LC: lots and lots (most of my country catalog of the few years preceding was fair game). LR: 1. LM: 0.

Record 3, Let's Kill Saturday Night. LC: lots (back-catalog plundering once again). LR: 4 or 5. LM: 4 or 5.

Record 4, Very Best of. Two-thirds of this one is tracks previously unused, so -- unless it's my best record and therefore undermines this entire method -- it doesn't fit into the exercise.

Record 5, 13 Hillbilly Giants. This was a covers record in which I considered -- actually listed on paper -- a couple hundred songs. LR: 1. LM: 0.

Record 6, Couples in Trouble. LC: 0 (this was written as a song cycle). LR: 0. LM: 3.

Record 7, Georgia Hard. LC: 4 or 5. LR: 0. LM: 2.

Record 8, Revenge! LC: 0. LR: maybe 20. LM: 1.

Record 9, 50-vc. Doberman. LC: 1. LR: 1. LM: 3 or 4.

Record 10, Happy. LC: a few (another covers record with many titles considered). LR: 3. LM: 2.

Having gone this far with this possibly inane exercise, let me go on to rank these records based on their leftover stats so I can get an idea whether my method is working for me. A song thrown away isn't equal in value or effect to a recorded title or a mix discarded, since the proverbial clock isn't running when you're at home writing, and since a potential atrocity discarded isn't as consequential to the final product as an actual atrocity discarded. Because of that, I'll weight leftover compositions lower -- 0 is 0; 1 will stand for any number 1 to 5; 2 for any number above 5 (including "lots and lots"). Adding in the other numbers and ranking the albums by their aggregate leftover content, we get:

Revenge! (21)

Let's Kill Saturday Night (12)

two-way tie, 50-vc. Doberman and Happy (6)

Country Love Songs (5)

three-way tie, 13 Hillbilly Giants, South Mouth, Georgia Hard (4)

Couples in Trouble (3)

Very Best of (0)

If you discount the top two results as special cases, distorted effects of corporate waste (Let's Kill) and live-record over-recording (Revenge!), then this is not a crazy ranking, according to my opinion of these records' comparative merits. Personally I'd rank Georgia Hard a bit higher and Happy a bit lower, but Very Best of, my worst record, is right where it belongs, and the broad shape of the list (again, ignoring the top two) I think reinforces the idea that more choice equals better outcome.

At this point, for a record that will probably have 12 songs on it, I've got 15 completed and another half-dozen in various stages. As usual I try to mix up the methods so I don't settle complacently into one. This goes especially for what I think of as the entry points, the windows in -- how the song gets underway. A musical phrase is my commonest portal, and I think my most fruitful, but then you can also poke at verbal phrases, write poetry, screw around on an instrument, or think about a character or theme from a book you liked. Country writers often work from titles, which is efficient, but I'm a little leery of that for just that reason; title-generated songs can feel too shrewdly goal-oriented, so damned fixed on the hook that not enough surprise, or real-life-style randomness, is allowed to filter in.

Writing is the only part of music-making that is just no fun. Even long drives on the interstate have an energizing undercurrent -- a vibratory state of mesmerism that could be broken any second by a headlong plunge into a truck. Sitting alone in a room is four kinds of hard. Trapped-without-distractions hard, swimming-in-your-own-mental-mediocrity hard, making-something-from-nothing hard (hey, that's actually supposed to be physically impossible!), and plain old writing-well-is-hard hard, which is hardest. I guess that's why I'm sitting here blogging instead.

An interesting and mysterious feature of record-making is going into the studio with impeccable blueprints and coming out of it with terrible products. What was misunderstood or unforeseen and why? Alas, foresight and intelligence and preparation are never foolproof. Miscasting, for one, has tripped me up many times. The pianist isn't attuned to the genre; the star soloist, brought in in a rush of excitement and paid twice as much as anyone else, has a star attitude and doesn't give a rat's ass about your session; the engineer is trying for unknown reasons to make you sound like Badfinger. From oboeist to recordist, the songs you write need the right executors, and, better yet, you should be mentally taking the executors' specific talents and techniques into account as you write.

When I talk about integrating an idea of casting into composition -- writing with the individual performer in mind, as Duke Ellington said -- I'm still thinking specifically about the songwriter-performer modus operandi, not Irving Berlin. "White Christmas" can undergo a thousand good maulings, but "Rag Mama Rag" without Richard Manuel's awesome drumming and Rick Danko's awesome untrained fiddling and Levon Helm singing in an unaccustomedly low register and Garth Hudson's parlor-prodigy riffing is nothing but a queer noise in Robbie Robertson's head. When you hear a recording like "Rag Mama Rag" the text, the blueprint, is as good as invisible; just to hear in your head what a one-instrument demo of the song must have sounded like is enough to quench your interest in hearing it. "White Christmas" would be wonderful to have written, no doubt. But this other mysterious kind of writing, where the writing itself ultimately fades thoroughly into the performance, becomes hopelessly lost within it, is for me an irresistible target to try for. I'm not sure exactly how you try for something that serendipitous -- I just try to hear R. Gjersoe and J. Scheinman play as I write, and knock on wood.

I find that my ideas of what makes a good song good, what kind of music is most attractive, and what my capabilities are do not stay quite fixed, at least over a period of seven or eight years. Reconsideration and an evolving sense of beauty keep music from going stale for me and maybe in turn can keep me from going stale for others. "Sense of beauty" sounds lah-de-dah, but when you are sitting working on a song, it has instant and observable consequences. In Jimmy Webb's excellent book Tunesmith he compares a song to a building. The introduction is like an anteroom or foyer, inviting you farther in. The verse is a hallway that opens gloriously onto the living room of the chorus. Though I'm not doing it justice, it's a well-wrought analogy for a kind of song, and you can't write that kind of song without having embraced that sense of what is beautiful and rightly ordered.

By saying "a kind of" I'm giving away how my opinion of the clear dominance of a commercial, Tin Pan Alley-based approach to songcraft has loosened over the last couple years. Paul McCartney's "Yesterday" and George Gershwin's "I Got Rhythm," for example, fit Jimmy's schema. Once you enter them, you know you're in the presence of a master architect. You relax, suspend skepticism, float. But then, songs like "Tutti Frutti," or "Kum Ba Yah," or "Apple Tree Swing," or "Lass of Roch Royal" are something else again. If you sense a designer's hand at all, it's not an architect who has your relaxation foremost in mind. These songs are more like jagged winding forest trails than gleaming symmetry-rich palaces. In a way they're more like life than art: this part's pretty and that part's ugly, this part's fucking senseless, and wait -- who made this and how come it's so captivating and why am I even here?

All these songs, commercial or otherwise, have a lot of basics in common -- logical chord progressions, catchy melodies, words that suit the countours of the melody (or offer a nutritive contrast to them). And I think all songs, whatever their provenance, have a utilitarian raison d'etre, even if the use of a commercial pop song (provoking dance, dignifying mating rituals, stimulating reflection) is, you might say, more decorative than the use of a rural public-domain song (provoking dance, disseminating news, easing the workday). But then there are deep aesthetic differences. Once you try applying some of Jimmy Webb's (or Elton John's, Diane Warren's, Carole King's...) very sensible principles to any song written outside the big tent of postwar American commercial music, you see that we can't pretend to accept a single set of values -- at least not without ignoring or ingeniously redefining (Greil Marcus could do it!) many of the greatest songs ever made. If there's a single system of songwriting values by which "White Christmas" is at the pinnacle, then "Tutti Frutti" is not a lesser song but actually a valueless song. Since "Tutti Frutti" is on its face a really good song, there are evidently more systems than one.

The folk song system, to give the system by which "Lass of Roch Royal" is a masterpiece a crude name, was burbling along in poor folks' communities a long time before Gershwin, and today, despite interconnectedness with and influence from commercial culture, it continues on its parallel course. If you were to immerse yourself in Starbucks samplers for a year and then suddenly "Lass of Roch Royal" or some old folk or blues song hit your ears, it would be like being thrown from the Taj Mahal into Starved Rock State Park. The songwriter is no longer so attentive to your comfort. Why is this verse second and this one third, when they seem interchangeable? How do these two verses connect at all? What motivates this brutal violence? Why is this bird talking? The strangeness of these old tunes is largely accounted for by the strangeness, in relation to ourselves, of their creators, but sometimes the strangeness is also because the erosions and distortions of time and carefree repetition have mixed up verses and phrases, and made weird morphs of regional variants -- really neat. Avant-garde, too.

To try to design the forest pathway from scratch, to aim consciously at recreating these naturally evolved orally-transmitted forms, seems a fool's errand. But again, the target is so inviting, because the music is so bewitching, words and music and rhythm and attitude all together. So who you calling fool? Writers like Gillian Welch and Norman Blake have hit gold by letting pre-copyright music by poor Americans inspire and steer them. Though it's not a big focus of this next record, I'm enjoying a little play with that method as well, because I guess, beneath the fancy talk and justifications, that kind of music is just floating my boat more than "I Got Rhythm," at this moment.