While The Trojans was written, as I said, in the form of a 19th century French Grand opera, with all the things that entails -- five acts; spectacular moments like the mass suicide of the Trojan women and the Royal Hunt and Storm; ballets; extensive use of the chorus; a heroic tenor role -- it wasn't actually performed as such in Berlioz's lifetime. The most common approach to the work, in the 19th century up to the late 20th century when it finally started to get into the repertoire, was to split it up and perform it as separate works: the first two acts as a self-contained work about the fall of Troy, and the last three acts as an opera about Dido and Aeneas. I actually see the point of this, and wonder if it might be a good approach to take even today; the work is very long, but more than that, there's a stylistic disconnect between the two halves. The first part, focusing on Cassandra, is almost unrelievedly gloomy and grim; it fits the subject, but it doesn't put the audience in a frame of mind to stay the full five hours or whatever. The Carthage section has more variety; this is where the Shakespearean influences start to become apparent in Berlioz's libretto, resulting in great moments like the duet for two of Aeneas's men, who scoff at all this business about destiny and leaving a nice place like Carthage just to go to Italy. Performing those three acts as a self-contained work makes a certain amount of sense, though obviously there are connections, thematic and musical, between the two halves of the work (the ultimate irony of the piece is that, in fulfilling his destiny to go to Rome and revive the destroyed culture of Troy, Aeneas has helped to destroy the culture of Carthage, converting its people from a civilized, cultured people into bloodthirsty, vengeful warmongers like the Greeks in the Troy sequence). And the whole work is bound together by the strangely "neoclassical" approach that Berlioz takes; the model, dramatic and musical, is 18th century French opera, and in particular the solemn, serious-minded mythological operas of Gluck, like Ipheginie en Tauride.
One notable thing about French Grand Opera is that the three 19th century grand operas that are generally regarded as the greatest, Rossini's William Tell, Verdi's Don Carlos, and Berlioz's The Trojans, were all failures to varying degrees. Tell came the closest to being a hit, but it was considered too big, too serious, too ambitious, and not spectacular enough to be a hit with the audience, as opposed to the critics. (The story goes that the Opera house started performing only the second act of Tell; when told about this, Rossini replied "What, the whole second act?") And even now, none of those three operas can claim to be among the most popular of all time, not compared to non-French operas that are influenced by French Grand Opera without fully being part of that form, operas like Aida and Tannhauser. Among the true Grand Operas that were big hits in their time, Halevy's La Juive and Meyerbeer's Les Huguenots get revived sometimes, and there are still diehard fans of the form, like Tom Kaufman. But if this is the spot for me to say that Grand Opera deserves a re-evaluation, I don't really think I can do it; I love a lot of French opera -- Massenet, Bizet, Offenbach, Gounod -- but the big, blockbuster grand operas tend to leave me cold, even when written by a composer I like (I love Massenet's small-scale works, but his attempts at grand opera, like Herodias and The Cid, just have too many dead spots and bombastic musical gimmicks).
But then, it's unfair to evaluate a lot of Grand Opera by the music alone, or even the music and the text alone. Composers like Meyerbeer and librettists like Eugene Scribe always calculated their work in terms of the effect it would make not only in the theatre in general, but in the context of a specific production at a specific opera house (the Paris Opera). The Trojans was something Berlioz wrote and then tried to get produced; William Tell is a work of pure musical integrity, almost separate from the circumstances of production; but Meyerbeer's The Huguenots or Robert the Devil hardly exist apart from the way they were produced. As Ethan Mordden puts it in The Splendid Art of Opera:
Compared to Rossini's masterpiece [William Tell], Robert does look overblown and nugatory, a two-ton frisbee. Meyerbeer's was an eclectic's touch. The Opera has installed an organ? Let's have a church scene. The technical staff has found another way to deploy gas light to create a mist in the air? We can use that in the ruined cloister scene. There's a tournament in the second act? We can slip another ballet in there, as a kind of appetizer for the big one in the third act -- and there, as chief defrocked nun, let's get the formidable Marie Taglioni. The tone of that act is a little too macabre, though -- let's make the second tenor a comic character, so he can sing a funny duet with the demon; that'll be sure to go over.
But that description touches on the facet of French Grand Opera that is at once its great strength and weakness: the Grand Opera was the true Gesamkunstwerk, a total work of art in which every aspect of the production -- music, text, direction, sets, choreography, lighting -- works toward the same end and contributes to its overall effect. Wagner invented the term Gesamkunstwerk to describe his own operas, and he certainly claimed that he was working toward a total theatre experience. But in practice, he was writing operas without any real reference to what could be done in the theatre; he wasn't able to have Rheinmaidens who could convincingly sing and "swim" at the same time, nor a dragon that looked genuinely scary. Wagner didn't even write the vocal parts with any particular reference to what could be done by real-world performers, with the result that he wrote tenor parts that no tenor in his time (and no one ever, perhaps, except Lauritz Melchior) could sing all the way through without burning out his voice by the third act. Wagner talked aboout total theatre, but really the totality of what he envisioned existed only in his mind, not in the theatre.
It was in Paris that opera was truly a full-fledged theatre experience where all the elements worked together to the same end. It was an end that Wagner scorned -- cheap sensationalism and bourgeois morality wrapped up in a package of hummable tunes -- but, like a Hollywood blockbuster of the present day, a Meyerbeer/Scribe opera used all the best available theatrical devices, including cutting-edge special effects, to make the finished product something more than just music and singing. Like a '30s musical comedy, the result doesn't necessarily look like much once you take away the original production and the original performers, but then, it wasn't meant to. And to give Robert the Devil his due, what Meyerbeer had is what Berlioz never had: a sense of exactly how his music would work in a theatrical context and how to hook an audience so completely that they'll stay with you for five hours. That's why Meyerbeer's operas were hits in a way that The Trojans never really has been -- and in a truly great production with great singers and special effects, I suspect many operagoers would enjoy those operas more than Les Troyens, even while acknowledging that Les Troyens is the greater work of art.