"Twentieth-century opera" -- words to chill the hearts of all but the most dedicated lovers of discord. The images conjured up in the mind of the average season-ticket holder by the simple phrase are generally violent and/or tortured: Berg's Wozzeck stabbing poor, sluttish Marie; Schoenberg's Moses and Aaron discoursing endlessly, atonally on the gap between idea and expression; Stravinsky's Oedipus agonizing in Latin before he puts his eyes out. Most of us prefer traditional fare: the hopeless optimism of Tosca , the Marschallin's lingering farewell to Eros.

As it happens, of course, "Madame Butterfly," "Tosca" and "Rosenkavalier" are all operas of this century, as are "Gianni Schicchi," "The Consul," "Porgy and Bess," "Threepenny Opera," "The Turn of the Screw" and "The Ballad of Baby Doe." Throw them into the mix (and a good many similar titles could be substituted) and the subject of 20th-century opera begins to look more interesting.

In fact, a careful examination of the 78 synopses that are the core of this volume supports the belief that the present century has contributed more to opera than any other -- even with 20 percent of its time left to go, and even if you stack the cards as Martin does by deciding that the century begins "stylistically" around 1910.

The century's contribution is obvious in terms of sheer quantity. In his 1941 survey of the basic opera repertoire, "The Opera Companion," Martin included only 47 operas compared to the 78 in this volume, and seven of those operas date from 1900 or later.

Anyone who goes to an unfamiliar opera (even one in English) without having first done some homework is seriously cutting down on his possible enjoyment of the experience. That is, of course, the reason for the existence of this volume and many others dedicated basically to telling people what happens in these curious, hybrid works of art except when it deals with the handful of 20th-century operas that have entered the international basic repertoire, this volume enjoys a near monopoly of its subject matter.

The best of the standard opera guides is probably "Kobbe's Complete Opera Book,' edited and revised by the Earl of Harewood, who has valiantly tried to continue updating the 71-year-old basic volume but admits in his preface that there are "gaps, most of all probably in the sphere of contemporary opera." Kobbe does include quite a bit of 20th-century material, but many of the composers included are ignored by Martin -- Delius, Weinberger, Falla and Kodaly among them -- and in some cases (Virgil Thomson, for example), Martin chooses to synopsize one opera while Kobbe chooses another.

The only justification I can see for Martin's omission of "El Amor Brujo" or 'Hary Janos" is that he must assume that hard-core opera lovers will already have Kobbe (as they should) and he might as well complement rather than duplicate in borderline cases. Where both writers do discuss the same opera ["Porgy and Bess," for example, or "Love for Three Oranges" or "Albert Herring"], Martin's approach seems a sensible alternative or supplement to Kobbe's. Martin varies his treatment more from one opera to another, and he also offers commentary, while Kobbe sticks closely to straight plot summary.

Some of the entries are brief and simple; Stravinsky's "L'Histoire du Soldat" and "Mavra," for example, receive only three paragraphs each. Elsewhere when it seems necessary or useful, Martin's discussion can become elaborate -- as in his 7-page analysis of the intricate musical structure of "Turn of the Screw." Some composers are treated exhaustively -- Richard Strauss, for example, has several minor works represented in the book -- while others are relatively or totally ignored. The decision in each case is based on the quality and complexity of the material, and on a calculation of how likely an adventurous opera-goer is to encounter a particular work. Such decisions are partly arbitrary, but his opinions on what to include and how to treat it are generally sound.

One handy practice (already familiar from "The Opera Companion") is Martin's inclusion of key phrases from each opera in a foreign language, with pronunciation and translation: "nessun dorma . . . nes. soon dorma . . .nobody sleeps." For those unfamiliar with an opera's language, a quick survey before a performance or at intermissions can be most helpful.

The synopsses are preceded by an interesting and informative but rather miscellaneous series of articles on various topics related to 20th-century opera, and followed by a 50-page statistical survey of 23 of the world's major opera houses, showing what operas have been in their repertoire and how often each has been performed.

Neither of these sections is essential to the book's basic purpose, which is simply to prepare the opera-lover for an encounter with a particular work, but both are welcome. For those who go to the opera frequently, and whose taste extends beyond the traditional basic repertoire, "The Opera Companion to Twentieth-Century Opera" is an essential reference work, considerably more useful than the standard opera guides that limit themselves to already familiar material.