Played-in instruments

Why do players insist a violin sounds better after years of playing than it did new?

Almost every string player will tell you the same thing: a new violin is stiff, tight, a little unwilling, and after a few years of serious playing it opens up. The claim is not rare or eccentric. It is nearly universal among people who have spent their lives with the instrument, which is a strong reason to take it seriously.

But notice the shape of the claim. It says the object changed. And that is a claim about wood and varnish, testable in principle. So the honest question is not whether players are imagining things — something clearly changed — but which part of the system did the changing. Because a violin being played is never a lone object. It is an object, a player, a room, and an ear, all coupled together, and any one of them could have moved.

Start with the part that is easiest to check. Suppose the wood really does alter under vibration. What would we expect? We would expect a controlled experiment to find it: take matched instruments, vibrate some mechanically for hundreds of hours, leave others alone, measure the response. Such experiments have been tried, and the honest summary is that the results are weak and inconsistent — no reliable, reproducible effect of playing-in on an instrument's measured acoustics has been established. That is not proof the effect is absent, but it is a poor showing for a change described as unmistakable.

Now separate two things we tend to run together: age and playing. Wood genuinely changes over centuries. Hemicellulose, one of its structural polymers, slowly degrades, and old wood absorbs less moisture and damps vibration differently than new wood. That is a real, measured chemical process — but it happens to a plank in a loft just as it happens to a violin on a stage. It is a clock, not a workout. It cannot explain why this instrument improved while I played it and my colleague's identical one, in its case, did not.

So look at the other elements. What else changed over those years? The setup, for one. A new instrument's soundpost settles, its bridge is shaved and repositioned, the strings are replaced dozens of times, the fingerboard is dressed. A luthier adjusts an instrument toward its owner's taste in small increments, over years. That is a real physical change to the sounding system, and it is driven by the playing — but by the maintenance the playing occasions, not by the vibration itself.

And then there is the largest term, the one that is hardest to see because we are standing inside it. The player adapted. Learning an instrument's response is a motor-learning problem: where the bow finds each string's sweet spot, how much weight each register wants, which vibrato speed makes this particular box bloom. Over years, a player's technique is quietly reshaped by one instrument. Ask what that predicts. It predicts exactly the reported experience — the instrument seems to be giving more, because you have learned what to ask.

Does the evidence support giving the player that much weight? Consider what happens when we remove the adaptation. In double-blind trials led by Claudia Fritz and colleagues, published in 2012 and again in 2014, experienced violinists — and in a later study, listeners in a hall — compared old Italian instruments, Stradivari among them, against fine new ones under conditions that hid which was which. Players could not reliably identify the old instruments, and on average preferred the new ones. Whatever the players were tracking, it was not something that survived being separated from knowing what they held.

So we have three candidates: the wood, the setup, and the player. Which one carries the weight? The evidence points hardest at the third and second, and thinnest at the first — and this is worth saying plainly rather than softening, because the first is the one the folklore names.

A few refinements, because the point is easy to overstate in the other direction.

First, this is not a debunking. The player's experience is accurate as a description of the coupled system: the instrument-plus-player really does sound better after years. What is unsupported is one particular explanation of why.

Second, none of this says old instruments are not excellent. They plainly are. The blind trials show that their superiority is not audibly reliable, not that they are bad — and a great instrument being merely equalled by a fine modern one is a statement about modern makers, not an insult to Cremona.

Third, expectation is doing real work here and should not be treated as a slur. Knowing an instrument is old and expensive genuinely changes how it sounds to you, in the same way a wine's label changes its taste; that is a fact about perception, not a failure of honesty. It is also, importantly, not the whole story — the setup drift and the motor learning are physical and would happen to a player who knew nothing.

Fourth, the honest position on the wood is not settled rather than disproved. Absence of a demonstrated effect in difficult experiments is weaker evidence than absence of the effect.

• Why steel-strung guitars have a better-evidenced break-in story than bowed instruments do.
• What the blind-trial results imply for the economics of old instruments.
• How luthiers actually judge an adjustment, given that the player they are adjusting for is also changing.

Every term the collection defines is gathered in the glossary.