I received an interesting message yesterday from SAm, responding to my preview of the forthcoming Elliott Smith album. “It bothers me a little bit to read, here and elsewhere, Elliott’s ‘strummed melodies’ described as ‘simple’,” he writes.

“Elliott was a good lyricist, a very good guitar player, and decent singer, but what makes him totally unique and superior to virtually all of his contemporaries is the melodic songwriting, the ingenuity and originality of the chord progressions. The formal problem that pop songs present is that they must be a)short b) instantly familiar and c) unfamiliar. It’s an exercise in expectations ingeniously subverted and then met.

“Go and play his songs on the guitar, or try to. I guarantee you will find in almost every one a sophisticated and basically amazing sequence of chords within the overall, conventional melodic structure – ie the difference between a good song and a boring one. It takes a certain kind of musical IQ to come up with that, something I’ve never detected in, say, Keith Richards, or the guy who writes the songs in Modest Mouse. OK, rant over.”

I’ve never played a guitar, or any instrument in fact, so I can’t personally test out SAm’s theory. It does strike me, though, as touching on a couple of fairly prickly subjects. One is about why Elliott Smith’s music is so effective and, frequently, affecting. On the surface, he was a fairly conventional singer-songwriter, indebted – like so many others – to The Beatles, with a vague folkiness that comes to the fore on the solo acoustic songs on “New Moon”. A casual listener might skip through a few of his albums and conclude his songs were even fairly samey.

And certainly, Smith worked in a pretty narrow range. If we disregard the emotional impact of his words and voice for a moment, perhaps it is the “sophisticated and basically amazing sequence of chords within the overall, conventional melodic structure” that made his songs so powerful.

Which brings us to the second slightly disturbing thought which SAm’s email provoked. If a song is successful because of a technical melodic complexity which can only be identified by someone who tries to play it, does that mean that so much music journalism is based on hunches? Is Brian May right, and should I try and write my own songs before I’ve earned the right to criticise those of other people?

I think the answer to that one is ‘no’, but of course I would say that. The overall impression that a record makes is more important and more interesting to read about, I reckon, than a musicological exposition. Part of the charm of Elliott Smith’s songs is that they appear simple, that they create an illusion of direct, spontaneous emotional communication even though real craftsmanship underpins it. But maybe occasionally a look at the melodic nuts and bolts can give us an insight into what makes a song work?

I’m going on a bit now. But anyway, we have one more report coming up from South By Southwest to come (tomorrow, hopefully). Other things to look forward to on Wild Mercury Sound soon: Mark E Smith and Mouse On Mars‘ stomping collaboration, Von Sudenfed; the garage glam of The 1990s; the medieval electronica of Colleen; the fantastic Marnie Stern; and, hopefully, previews of Bjork and The White Stripes. Just don’t ask me to analyse their compositional skills. . .