Leonard Cohen and Philosophy (13 page)

It’s this playful engagement with his persona that can make many Cohen songs difficult to cover, as in the case of “Famous Blue Raincoat.” But in this case, there’s still enough space for other singers to make the song their own. Cohen isn’t the only singer with a persona. We could run similar word-association exercises with Marianne Faithfull, Tom Jones, Elton John, Jeff Buckley, Lana Del Rey, or anybody else who’s ever covered a Cohen tune. These associations will help to guide our interpretations of their performances. They help us to “fill in” the information we need to determine who
the “I” is that’s addressing us in the song. In other words, the artist’s persona forms part of the context we need in order to understand what they’re trying to say to us.

Sincerely . . .

One of the reasons that the significance of the golden voice lyric changes depending on who’s singing it is that Cohen has left an opening for other artistic personas to affect our interpretation of it. This is because Cohen doesn’t explicitly tie his persona to the song. Although the song contains a number of references to his persona, they’re only alluded to, hinted at (what philosophers call “implicated” but most people call “implied”) rather than stated directly. A lot of Cohen’s songs have this feature. Even something as personal as “Chelsea Hotel #2” (widely reported to have been inspired by Cohen’s relationship with Janice Joplin) does not reference the Cohen persona directly. It does, however, hint at it through the claim that the singer isn’t a handsome man (another self-deprecating joke that changes character when sung by somebody else). Things, however, aren’t always so simple.

Let’s look at another Cohen song that doesn’t have the same openness to interpretation. “Famous Blue Raincoat” (from 1971’s
Songs of Love and Hate
), written in the form of a letter, recounts a falling out and possible reconciliation between two (male) friends over the affections of a woman, Jane. In the last line of the song, the author of the letter is revealed to be none other than “L. Cohen.” Let’s call this line the “signature.” The presence of the signature gives a particular identity to the person writing the letter—and by using his own name, Cohen explicitly ties the character in the song to his persona.

The presence of the signature poses a special challenge for anybody wishing to perform it. Because the song is tied so neatly to the Cohen persona, it’s difficult for an artist to bring his or her own persona to bear on their interpretation of the song. Put another way, the signature forces artists to
perform the song in character
as
Cohen. As a result, a would-be performer of the song faces a difficult choice: either don the Cohen persona yourself and perform the song as written, or retain your own persona and change the lyrics.

Both options have their strengths and weaknesses. Although the first approach, where you retain the original lyrics, has been by far the most popular choice of artists covering “Famous Blue Raincoat” in the years since it was first released (perhaps nobody wants to be accused of changing Cohen’s words to suit their own performance), how successful this turns out to be will depend on how well one is able to inhabit the persona. Performers who regularly perform in character have an advantage here: chances are that the Tori Amos version of “Raincoat” (on the
Tower of Song
compilation from 1995) will seem to many to be much more successful than other straight covers, since Amos is well-known as someone who sings songs in character—especially when covering other writers’ material (see, for instance, her 2001 cover album
Strange Little Girls
). At any rate, the second approach, where an artist changes the lyrics to weaken the associations with Cohen’s persona, is much more interesting.

Now for a true Cohen fan, it might seem to be nothing short of sacrilege to consider changing the master’s words. After all, if nothing else, Cohen is known for choosing his words carefully. Why shouldn’t we respect his wishes? And yet, Cohen himself has revised his songs over the years (most notably the new lyrics for “Hallelujah” that appear on 1994’s
Cohen Live
), and a few brave artists have made revisions of their own. On the title track of her 1987 album of Cohen covers, longtime Cohen collaborator Jennifer Warnes sings a slightly revised version of “Famous Blue Raincoat.” Although it’s difficult to determine exactly who was responsible for making the changes, it’s worth noting that Cohen worked closely with Warnes on the album, and even performs with her on a track. Taken together with his tendency to revise his own work, this suggests that, in Cohen’s mind at least, the second approach is an acceptable one.

The changes are small, but serve to the make the story
related in the song a great deal more ambiguous. In particular, the letter is now signed “a friend”—and the author is no longer explicitly claimed to have been Jane’s lover. Removing Cohen’s name from the ending helps to pry the letter-writer from his close identification with the Cohen persona. Warnes’s version of the song is less specific—the central conflict may still be a love triangle, then again it could be a parent attempting to come to terms with their child’s new relationship. As a result, Warnes’s lyrical variations serve to multiply the number of possible interpretations of the song.

The Rules of the Game

Up to now, we’ve argued that the ideas and expectations we form about a particular artist will inform our interpretation of his or her performance. So far so good, but we haven’t said anything about how these different interpretations come about. After all, even if we are associating different personas with, for example, the “I” at the beginning of the golden voice lyric, the rest of the line remains the same. How can three people more or less saying the same thing produce three different effects?

Although most of us are certainly aware of the effect context has on how what we say is interpreted by others, there’s a persistent belief that the meaning of what we say is fixed entirely by the meanings of the individual words we utter. After all, words have meanings, and they have those meanings regardless of who’s speaking and what their intentions are. So while the claim made in the golden voice lyric may be (strictly speaking) true when Tom Jones sings it and (strictly speaking) false when Cohen sings it, we need some way of explaining why Jones’s performance counts as a boast and why Cohen’s counts as a joke.

When we talked about the personas that certain artists bring to bear on the material they perform, we talked a great deal about the assumptions that audiences are likely to make about those performers. There are, however, other
assumptions we make that make ordinary conversation possible. The American philosopher Paul Grice attempted to catalogue these assumptions (which he called “conversational maxims”) in an attempt to determine exactly what context contributes to the significance of what we say to each other.

One of the maxims Grice identified is that you shouldn’t say anything you believe to be false (p. 27). Now Grice is not simply saying that you ought to always tell the truth—that would be a claim about morality, not about language. What Grice means is that, in ordinary conversation, we trust that people are being sincere in their dealings with us. This doesn’t mean that you should believe everything someone tells you, just that most of the time we expect that people are telling us what they actually believe to be the case. Because these assumptions are in place, they can be used by savvy speakers to convey information that goes beyond what has actually been said. One way of doing this is to violate these expectations in such an obvious manner that one’s partner in the conversation is able to work out what you’re trying to convey.

To see this in action, let’s return once again to the golden voice lyric from “Tower of Song.” Why does it come across as a joke when Cohen sings it? Well, one reason is that the statement is quite simply false: the most natural interpretation of the lyric is that the signer has a conventionally beautiful singing voice, and Cohen certainly doesn’t. What’s more, Cohen is not deluded—he knows that his voice is in no sense golden. So here we have Cohen saying something that he believes to be false. This is a straightforward violation of the conversational maxim we’ve just discussed. And yet, there’s no attempt to deceive—if this is a lie, it’s a bald-faced one. We’re forced to conclude that Cohen wants the line to be understood ironically. It’s the mismatch between our expectations about Cohen (including our belief that he’s not a conventionally good singer) and the claim contained in the lyric that makes the joke work.

Funny Voices and Angel Song

True story: I was recently at a service at a Roman Catholic Church where Cohen’s “Hallelujah” was performed as part of the liturgy. Although the chorus remained the same, the verses were changed to something a little more unambiguously Christian. Though the song was well performed, I found myself annoyed at hearing the song in that context. First, because the song has become overplayed to the point of becoming an emotional cliché (
see also
: “Grace, Amazing”), and second, because the change in lyrics didn’t accomplish anything that couldn’t have been accomplished by a change in musical approach.

I suspect that I’m not alone in my annoyance with the near-ubiquity of “Hallelujah” in the years since it escaped from near-obscurity through covers by Rufus Wainwright (on, of all things, the soundtrack to
Shrek
) and Alexandra Burke (who released her cover as a single shortly after winning the British version of the
X-Factor
in 2008). The success of the Wainwright and Burke versions of the song (and the tsunami of other versions that have followed in their wake) illustrates another interesting effect of context on interpretation. Most versions of the song that have been released since the boom aren’t that different lyrically from those (such as John Cale’s 1987 version) that preceded it. The difference is largely one of performance. The versions of the song recorded since the boom have tended to perform the song as a hymn or anthem, while the ones that preceded the boom (including Cohen’s own recording) tended to take a more ambivalent approach.

This shift in approach suggests that in thinking about the interpretation of a song, we must pay attention to not only
who
is singing, but
how
they sing it. The way something is performed can affect how we are meant to take it. This is something we are all familiar with. Consider the difference a sarcastic tone can make to the interpretation of a sentence like “That was a
lovely
evening” (or consider how difficult it is to convey sarcasm online). The same idea can be applied
to musical performances: In the late 1990s and early 2000s, there were simultaneous vogues for pop-punk covers of 1970s soft rock ballads (Me First and The Gimme Gimmes covering Carly Simon’s “Nobody Does It Better”) and for lounge covers of contemporary hardcore punk and heavy metal songs (Richard Cheese and Lounge Against the Machine’s version of Disturbed’s “Down with the Sickness”). These performances were clearly intended to be funny—the humor originating not so much in the lyrical content of the songs, but in the mismatch between that content and the performance style. Tom Jones’s performance of the golden voice lyric from “Tower of Song” is another example of performance informing interpretation, since it’s his decision to sing the line with the full volume and power of his conventionally beautiful singing voice that makes it seem like a boast.

On a lyrical level, “Hallelujah” seems utterly ambivalent about religious faith. This ambivalence is reflected in Cohen’s vocal performance, where the verses are almost spoken in a flat, dispassionate voice. Where any emotion does seep in, it’s anything but reverent (consider the extra bite Cohen puts on the “do ya” and “to ya” in the first and third verses of the original version). The chorus, on the other hand, would not sound out of place in a church. The juxtaposition of the plainness of the verse and the elaborateness of the chorus suggests a tension between earthly and heavenly desires.

This tension is conveyed in a very different way in Jeff Buckley’s version from 1994 (which may be the highest profile pre-boom version of the song). Buckley’s performance is spare and ethereal, suggesting at the same time a chilly remoteness and something quite carnal. If Cohen’s version of the song is the sound of us forgetting “to pray for the angels,” Buckley’s version is the sound of the angels forgetting “to pray for us.”

In contrast, Wainwright goes out of his way to bring out the song’s anthemic character, a similar approach to the one k.d. lang took in her version from 2004. Wainwright’s version of the song features a fluid vocal line throughout, and he replaces
the almost-sneering “ya” in the verses with the far less confrontational “you.” This isn’t particularly surprising given the generally operatic character of his own compositions. Burke uses her version of the song to show off the range and power of her voice, perhaps appropriate given that she rose to prominence through a singing competition. Because these approaches use the song to demonstrate technical mastery over lyrical content, they tend to favor singing the verses in the same way as the chorus. As a result, the tension that marked the Cohen and Buckley versions is lost, and the song comes across as a straight hymn.

An Unanswered Question

You may have surmised from the discussion in the last section that I prefer the pre-boom versions of “Hallelujah” to those that came later. You may have also noticed a hint of snarkiness in my description of the Tom Jones version of “Tower of Song.” In making these observations you may have begun to wonder just what makes for a good cover version. This is a difficult (perhaps impossible) question to answer, and I’m afraid that attempting to do so would take us too far off the track we have pursued so far. Having said that, we can at least conclude by remarking that sometimes we ourselves are part of the context of an interpretation, and our own histories and the associations we form with a piece of music will inform both our preferences and our understanding of its content.
¹