I've been reading Marvin Carlson's book The Haunted Stage. His central thesis is that while all literary texts bring up recollections of works from the past, performance is inherently even more about memory, since every element of a performance brings up recollections of past performances.
A play like Hamlet relies on the fact that audiences have seen other plays before and are familiar with the basic concept of a revenge tragedy, and probably with the story of Hamlet as well. The actor who is playing Hamlet has played other roles before this one, and audiences have likely seen, or at least heard about, other actors playing Hamlet. The costumes, scenery, lighting design, even the blocking probably resemble elements from other plays. Finally, the physical theatre where the play is performed likely conjures up memories of other performances. Even if an audience member has never been in that particular theatre before, he or she probably has seen other theatres to compare with it.
Carlson is aware of how all texts are "haunted" by the past, but he notes that the theatre is a special place due to its physicality. There is a phenomenological element to performance that necessarily complicates and enriches it to a level far beyond that of a simple text. He calls this "ghosting" and describes it this way:
Unlike the reception operations of genre (also, of course, of major importance in theatre), in which audience members encounter a new but distinctly different example of a type of artistic product they have encountered before, ghosting presents the identical thing they have encountered before, although now in a somewhat different context. Thus, a recognition not of similarity, as in genre, but of identity becomes a part of the reception process, with results that can complicate this process considerably.
The key point here is that it is "the identical thing" itself that is recycled, whether it be the script, the actors, the sets and costumes, or the theatre building. Of course, today we frequently expect sets and costumes to be used for one production and then only brought out for a new play if they have been considerably modified. Carlson points out that this is not always the case in non-Western theatre, and it did not used to be the case in the West, either.
He cites the Romantic era as the locus of this shift in the West. Specifically, he mentions Victor Hugo's preface to his play Cromwell, in which the French author rails against the neoclassical tradition of setting every play in the same neutral antechamber. For Hugo, the historical spot where an incident occurred was important, and the passionate feelings of Romantic artists, Hugo included, led to a demand that each new production make efforts to be historically and geographically accurate.
While we no longer expect every production of Julius Caesar to be set in ancient Rome or each new production of Macbeth to be costumed like medieval Scotland, we do expect new set and costume designs for a new production. Still, they will likely remind us of other designs we have seen in the past, and we cannot help but compare them to the elements used in these past productions.
Going back to the Greeks, playwrights have been concerned not simply with telling stories, but with retelling them. Here again, Romanticism changes things, placing a new emphasis on "innovation, individuality, and uniqueness," as Carlson puts it.
But what is the great Romantic text if not Goethe's Faust? And was not Faust inspired in part by the puppet plays that were performed throughout Germany in Goethe's day? And were not these plays in turn inspired by earlier plays, including perhaps productions of Marlowe's Doctor Faustus that likely toured Germany during the 16th and 17th centuries? Carlson points out that Paul Valery wrote a version of the story titled My Faust, though in a way each new telling is an author's own version of a common tale, My Oedipus, My Hamlet.
In the third chapter, Carlson remarks on an audience's ability to see not just an actor on stage, but the ghosts of all of the previous roles inhabited by that performer. He also notes the rather obvious fact that writers "ghost" actors into the texts that they write:
Goethe and Schiller conceived their productions with the specialties of Weimar actors in mind, Voltaire for the actors of the Comedie Francaise, Moliere for the company in which he was the leading player. Even a playwright like Ibsen, with very tenuous ties to his major producing organization, is revealed through his letters to be quite concerned with the specific actors that would perform his roles and with what associations and physical and emotional characteristics they would bring to the roles, certainly predictable concerns in any dramatist who writes with an eye toward stage realization.
This of course leads to the idea of the "vehicle play," which Carlson defines as a work constructed precisely to feature the already familiar aspects of a particular actor's performance." As examples, he cites Sardou's plays for Sarah Bernhardt and Rostand's plays for Benoit-Constant Coquein.