I’ve been feeling awfully lazy and unproductive lately, but I suppose it’s always a productive day when you can predate the Oxford English Dictionary. The word in question is atonal, which the OED dates in English from 1922:
1922 A. E. HULL in Musical Opinion Oct. 48/1, I have been working for two years at a system of non-tonal harmony, which I had long been unable to christen. Now, after visiting no less than seven foreign countries I not only find that the thing is widely known as Atonality, but [etc.]. Ibid. 48/3 Keyboard chord-writing as well as linear, tonal as well as Atonal.
Well, if the thing is already widely known, then maybe we’re not quite getting on that bandwagon early, are we? And, in fact, a little investigation finds the French-English critic M. D. Calvocoressi using the term a full decade earlier than that. From “The Origin of To-Day’s Musical Idiom” in the December 11, 1911 issue of The Musical Times and Singing-Class Circular (writing about Mussorgsky):
A score and more of such examples could be quoted. Not only these soft ‘atonal’ harmonies, but also the harsher whole-tome scales and aggregates, much used by Debussy and other contemporaries, appear in several parts of ‘Boris Godounov'[.]
Calvocoressi was a member of Les Apaches, the group of French artistic young-men-in-a-hurry that also included Ravel and de Falla, making a musical splash by defending Pelleas et Melisande from its critics. So it’s interesting to find atonal soon being taken up by the composer-pedagogue Vincent d’Indy, one of those critics, as a bit of a cudgel. D’Indy’s 1912 article “Le Bon sens” isn’t online, but you can find the American composer Daniel Gregory Mason quoting it:
“In the nineteenth century,” [d’Indy] says, “some Russian composers, in the interest of certain special effects, employed the scale of Whole tones, which one may name atonal because it suppresses all possibility of modulation. In the twentieth century Claude Debussy and Maurice Ravel elaborated these methods, making often very ingenious applications of them; but they made the mistake (one must dare to speak the truth of those one esteems) of erecting processes into principles, or at least of letting them be so erected by their muftis, so that the formula now established by fashion is: ‘Outside of harmonic sensation and the titillation of orchestral timbres there is no salvation.'[“]
Because it suppresses all possibility of modulation. You start to understand why, even given late-Romantic levels of dissonance, atonality so bothered the d’Indys of the world—dissonance was OK as long as the movement from key center to key center remained purposeful and perceptible, but lose that modulation, and things start to seem random. Debussy doesn’t seem atonal to us, but his penchant for using familiar chords (dominant 7ths, for example) for color rather than modulatory function must have exacerbated that modulatory uncertainty—in a Heisenberg-like way, in fact. D’Indy’s ideal Conservatoire ear could never correlate where Debussy’s music was with where it was going.