London's Music Shops Are Disappearing — and So Is the Argument Over What a Score Actually Is

Barbican Chimes, a sheet music and instrument shop that has served London's classical musicians since 1985, is closing. It sat next to the Guildhall School of Music & Drama, one of London's major conservatories, which meant its real customer base wasn't tourists — it was students buying exam repertoire and professionals buying orchestral parts they couldn't get anywhere else. It's part of a chain, Chimes Music, in business since 1951. Its closure looks small on its own. It isn't, once you see what else has gone down with it.

In May, Travis & Emery closed after 65 years at 17 Cecil Court, a narrow alley off Charing Cross Road. It specialized in old and rare music books and scores, and its story is stranger than a shop has any right to be. It was founded in 1960 by Valérie Travis, daughter of Sir Edward Travis, the wartime director of the Government Code and Cypher School at Bletchley Park. She married an organist and Bach scholar, Walter Emery, and gave him a typewriter she'd acquired from a captured U-boat, which he used to type up his musicology notes. When her nephew Giles inherited the shop in the 1990s, it was drowning in debt with three years left on its lease — until he found a page from one of Beethoven's own sketchbooks under an old box of tissues, sold it, and cleared the debt. The shop went on to sell a vocal score of Peter Grimes signed by Benjamin Britten for about £40, and shelved scores that had once belonged to Lord Harewood.

The shop's manager, Charlie — also a working choral conductor — told one interviewer he conducts from paper exclusively: "At this point I feel like I have to." His reasoning wasn't sentiment. A printed score takes a pencil mark, needs no power, and can be annotated, inherited, and sold on with all of that accumulated history intact. That's a small, practical case for paper. There's a much bigger one underneath it, and it's the one actually worth making.

A musical score isn't just a storage format for sound. It's a notation system that encodes reasoning, not just pitch. Take a triad spelled E♯–A–B♯. Play it on a piano and it sounds identical to F–A–C. But no trained musician would call those interchangeable, because the spelling tells you something the sound alone doesn't: which direction that note is resolving, what key context produced it, how the composer is arguing through the harmony rather than just landing on it. That distinction is called enharmonic spelling, and it's been a live part of how literate musicians read intention off a page for roughly three hundred years. It is also a distinction the standard MIDI protocol — the code underneath every piano-roll and loop-based production tool — cannot represent at all. MIDI assigns each pitch a single number; E♯ and F share one. The software has no concept of the difference, because it was never built to encode musical argument, only sound.

This is where two of the field's more serious scholars come in, because the authority of the written score has never been a settled question even among people who take it seriously. John Butt, in Playing with History, spends a chapter unpacking notation and lands on an uncomfortable point for anyone who wants the score to be a simple set of instructions: notation is an imperfect guide to practice. It never tells a performer everything; it demands interpretation, argument, judgment. Richard Taruskin went further and more provocative in Text and Act, arguing that the entire modern obsession with textual literalism — the urtext editions, the "historically informed" performance movement that insists on playing exactly what's on the page — is itself a distinctly modern, ideological stance, not a neutral recovery of the composer's true intent. Taruskin's point wasn't that the text doesn't matter. It was that treating the text as sacred is itself a choice people are actively arguing about, and have been for decades.

That's the argument today's musical culture has quietly stopped having — not because one side won, but because an entire category of people making music professionally never entered the debate to begin with. "Composer" now covers both a person who can tell you why a chord is spelled with a double sharp and a person building tracks entirely inside a DAW's piano roll, where that distinction is structurally invisible. Neither is illegitimate as a way of making music. But collapsing them into one word — treating literacy in a three-centuries-old argument-encoding notation system as equivalent to fluency in a modern production interface that can't even hold the argument — is exactly the kind of flattening that happens when a field stops maintaining the physical, institutional infrastructure of its own literacy. Bookshops like Travis & Emery weren't just selling paper. They were the last retail-level evidence that reading a score was a skill people still argued about, bought rare examples of, and paid to preserve. When that infrastructure goes, the argument doesn't get resolved. It just stops being visible enough to have.

Further reading: John Butt, Playing with History; Richard Taruskin, Text and Act.