For a zealous but basically uneducated classical music fan like me, there has always been something miraculous about the sound of the mid-century symphony orchestra — that supermassive lushness, all smooth textures and slow tempos and vibrato so wide you could steer a charging white stallion through it. I love everything about that sound. I love listening to it; I love imagining it unspooling out of tinny 1950s radios like a magician's endless silk. I am also someone who never found Tchaikovsky embarrassing and who "got pretty into the Brahms string sextets" after seeing Data play one on Star Trek. So consider the source, I guess.

Even for me, though, it's easy to see how the orchestra's transformation into a colossal, slow-moving force devoted to all-out maximum capital-everything beauty started to feel like a dead end in the decade after World War II. You can only play Beethoven like you're giving birth to the cosmos so many times before you start to wonder what's supposed to happen to the cosmos after it's born. Sometimes, here in the cosmos, it is 4 p.m. on a Thursday, and it's raining, and you're trying to decide whether to pour yourself a drink, and you're not sure whether you should call that particular boy or girl, the one who gave you what you think was a meaningful look, and anyway shouldn't classical music suit the texture of daily life, too? Isn't that one of the things it was written to do? For a generation of young conductors who came of age in the 1950s and '60s, who were no longer facing the existential trauma of the war but rather the paranoia and affluence and uncertainty of its aftermath, there was an intense desire to figure out how to make this music fit differently into life and culture. How to make it sound different.

It was a tricky prospect. You can't just swoop in and start gratuitously fucking with a style of music whose admirers see it as timeless, not if you want to land work. You need an argument. You need a theory. And this is one reason why I am grateful for Nikolaus Harnoncourt, who died on Saturday at the age of 86. Harnoncourt was one of the first, best, and most influential conductors to recognize that the way to re-historicize music for a resistant audience was to historicize the bejesus out of it, to recognize that you could slip music back into the flow of time by returning it to its original state. That is, he realized that since Bach as performed in the 1950s sounded nothing like Bach as performed in the 1740s — it was slower, richer, more romanticized; syrupier, if you want to be like that — the way to make Bach sound fresh for the 20th century was to do some detective work and figure out how Bach sounded in his own time.

Nikolaus Harnoncourt was a child of Austrian nobility; his full name and title were Count Johann Nikolaus de la Fontaine und d'Harnoncourt-Unverzagt. He became a cellist, and in 1953, together with his wife, Alice, he founded the Concentus Musicus Wien, an ensemble dedicated to period-instrument performance. (Alice Harnoncourt, FWIW, is way more interesting and important than I can do justice to here; she served as the principal violinist for the CMW for more than three decades and was leading a revolution in instrumental practice at a time when not all that many women had the opportunity to play the regular violin professionally, much less the pardessus de viole.) Over the next 20-odd years in Vienna, the CMW helped establish what became a hugely influential idiom of period performance: older instruments, faster tempos, leaner ensembles, rougher textures, drastically scaled-back vibrato.

Now, scaling back vibrato may not sound like an incredibly provocative or outrageous thing to do. "This brave hero who had the courage to scale back vibrato" is a weird thing to think, much less type. But in the context of 1950s classical music, historically informed practice landed like a pink wig and a Sex Pistols t-shirt. It sounded breathless and scraping, almost totally alien to concert audiences raised on Karl Böhm and Otto Klemperer. No kidding — this stuff is still controversial today.

Want a quick sense of this? Compare the opening of Bach's first orchestral suite as recorded by the great midcentury Bach interpreter Karl Richter:

And by Harnoncourt and the CMW:

Richter's recording is sumptuous, arresting, reverent, and harmonious, the result of a profound engagement with centuries-old music. But Harnoncourt's recording makes the music sound like it was written three days ago.

I think that what finally made Harnoncourt a great conductor, though, is not just that he was innovative but that he understood that his innovation was a paradox. A lot of early period-instrument conductors settled, as they aged, into a kind of hyper-scholarly suburb of artistic feeling. You get the sense of wizened men emerging from obscure libraries waving new evidence about the baroque inverted mordent, then issuing dutiful 11-CD boxed sets to illustrate what they've found.

But music lives in the moment when it's played. It has to speak to that moment. If it can't, then your fidelity to a long-dead composer's intentions is just a very literate form of cosplay.

Harnoncourt recognized that "make it new by making it old" and "prove that it changes by restoring it to its essence" were, for all their brilliance as theoretical turns, also basically nonsense, and that their real utility for an artist lay in opening up new realms of interpretation, possibilities that couldn't be glimpsed under the old way of doing things. No one is ever going to know what Mozart sounded like to Mozart. The reason to start thinking along those lines is that doing so can help you discover ways of feeling and hearing and understanding that you wouldn't otherwise have known to look for, much less found. Harnoncourt was rarely afraid to follow his insights where they went, even if where they went wasn't what other conductor-scholars thought of as historically accurate. He wasn't a dry purist; he was an artist who made period practice a form of liberation, not servitude.

This means that his recorded output is uneven, and what you like or don't like may depend on whether you prefer more mainstream interpretations (his Mozart, a lot of his Haydn) or more experimental ones (his Bach, which is almost uniformly weird and great). I like his Brahms symphonies a lot. Later in life he became an opera conductor, and his operas are a delightfully mixed bag. You never knew exactly what he was going to give you — which is something to treasure anywhere, but especially in classical music, where establishing a replicable formula is how most conductors make their careers. I'll remember him for helping to jolt classical music forward at a moment when it threatened to become fossilized in its own gorgeousness. He moved it forward by moving it backward, which was his great trick. He moved it inward, too.