ACCUSE LEONARD Bernstein of anything; you have to admit he is a great performer. He is one of the best conductors of our time, a popular composer, and surely the best known musician. A generation of amateur music lovers got their start with twelve years of nationally televised Young People's Concerts, and Bernstein is still just about the only conductor who can get prime network time for classical music. When he delivered the 1973 Norton Lectures, Harvard didn't have a theater large enough for the crowds of musicians and non-musicians who wanted to attend.

When Bernstein finally walked out onto the stage of the Harvard Square Theater, graduates of the Young People's Concerts must have felt right at home. There was the familiar piano, and there was the Teleprompter with the script for the evening. Only the orchestra was missing, and that turned up later in the form of filmed musical examples starring the Boston Symphony and the Vienna Philharmonic, conducted by the Maestro himself. Later in the week, each lecture was redone in front of the television cameras and broadcast over local public television.

And now Columbia Masterworks has released the entire six-lecture series as a set of 17 records. They do so "with special pride," according to a squib on the back of the box. Columbia's promotion of the lectures has played heavily on Harvard's prestige, and on the prestige of Bernstein's predecessors in the Norton Poetry chair, who include Robert Frost, T.S. Eliot, and e.e. cummings, as well as such composers as Stravinsky, Sessions, and Hindemith.

The comparison is unfortunate. Bernstein's lectures are a deeply personal statement, infused with the conductor's considerable charm and delivered with humor and sincerity. But it is hard to avoid feeling deceived by Bernstein's performance. He is careless with words and with ideas, and in the end he is always willing to sacrifice accuracy for the sake of the grand gesture.

BERNSTEIN ENTITLED his lecture series "The Unanswered Question" after Charles Ives' short chamber work of that name. Ives meant his piece to ask "the perennial question of existence," and for Bernstein, that is "Whither Music?" Before we can deal with "whither," though, we need to know "whence," and Bernstein's first five lectures are devoted to tracing the origins of what he considers a twentieth century crisis in musical development.

Bernstein claims to have something new to offer in his view of musical history, a new tool for aesthetic theory: linguistics. He suggests that scientific analysis should replace the "purple prose" of most musical discussion with a whole new discipline, "musicolinguistics," and that is where his troubles begin.

Bernstein starts at the parlor game level, trying to find analogies between language and music. First he tries matching up sounds with tones, words with musical phrases, and so on. "I believe it's no accident that the German word satz means both sentence and symphonic movement." This seems a bit simplistic, though, so he tries again with parts of speech, equating nouns with motifs and adjectives with their harmonic underpinnings--Wagner's Fate motif played over a diminished chord could mean something like "cruel fate." Verbs naturally correspond to rhythm, so Bernstein adds some triple meter at the piano and comes up with a complete sentence, "cruel fate waltzes."

This sort of thing is pointless, but it's innocent enough, and Bernstein admits that these analogies are only "quasi-scientific." In later lectures, though, he seems to take them quite seriously--at one point he even calls a bar of Mozart that involves both rhythm and harmony a "participle," as if you could mix parts of speech the way you mix blue and yellow to get green.

This kind of recklessness with language characterizes the whole series. Bernstein has a curious difficulty in deciding what audience to address. He makes a careful effort to avoid bringing in too much musical esoterica--he occasionally calls time out to explain such things as diminished seventh chords and the harmonic series, and then apologizes, saying "But you knew all that." But when he talks about linguistics the Young People's Concert atmosphere disappears and the jargon rolls in thick enough to cut.

Often Bernstein seems to be using linguistic terminology simply because he likes the sound of it. The word "syntactical" appears in his discussions of music at apparently arbitrary intervals, and usually seems to mean something like "important." And when the available jargon is not enough, Bernstein makes up his own, including such unlikely hybrids as "morphosemantics." The result is that much of what he says about linguistics is not so much wrong as it is just empty.

ANALOGIES BETWEEN music and language are intriguing, and when they are treated as metaphors they can even be useful. But Bernstein wants to prove a point, and he pushes his analogies too far. He believes that music is a universal language, not because, as third grade teachers explain, tempo markings are in Italian, but because all music shares certain structures in spite of the obvious variety in the music of different cultures.

In his quest for universality in music, Bernstein begins with monogenesis, the idea that all language evolved from a common origin. As a metaphor, monogenesis lies behind the Biblical Tower of Babel myth, and as a general principle it lies behind a century of serious philology, but it is not an idea with any scientific foundation--linguists believe the dozen or so major language families to be unrelated. Still, it reminds Bernstein of a discovery he made when he was an undergraduate: the first four notes of Aaron Copland's Piano Variations rearranged and transposed in various ways turn up in all kinds of music, both Western and non-Western. A four note tune may seem to be a weak and superficial thread with which to bind music of different cultures, but it's the kind of thread Bernstein likes. A considerable amount of musical analysis in subsequent lectures is based on just this sort of tune-hunting.

BERNSTEIN CONTINUES his search for universality with a discussion of transformational grammar, a field which has become the dominant area of study of American linguistics in the ten years since it sprang fully-clothed from the brow of Noam Chomsky. Contemporary linguistics, focusing on syntax, aims at uncovering the structures underlying language. And this is the source of the universality that Bernstein finds so attractive--beneath their surface differences, Chomsky believes, languages are organized on a few simple and universal principles.

Traditional phrase structure grammar analyzed sentences independent of their relation to other sentences: for example, a passive sentence like "Max was crushed by the safe" would be parsed into subject-verb-prepositional phrase. Chomsky's contribution was to recognize that the same, easily described relationship between that sentence and its active counterpart, "The safe crushed Max," exists between countless other pairs of sentences. He conceived of "transformations" as simple devices to describe the relations between simple sentences like "The safe crushed Max" and complex ones like "Max was crushed by the safe," "What the safe did was crush Max," and so on. Chomsky also imagined that those transformations were part of the internal grammar of every speaker, actually used in producing and understanding sentences, but most linguists today ignore that aspect of his theory. Transformational grammar is generally treated as an abstract model of a language's syntax.

Chomsky also believed that features of transformational theory that were found in every language--linguistics universals--would necessarily be innate, biological structures. Again, in the context of contemporary linguistic theory this is an eccentric view, but it is the view Bernstein has grabbed onto--he sees transformational grammar as a "subconscious process," innate and not learned.