Eliot put it, ‘If the audience gets its strip tease it will swallow the poetry’). It is doubly surprising that Shaw would think this, as when he wrote ‘The Quintessence’, he was just about to embark (or in the case of Widowers’ Houses) had already embarked on, the creation of plays in which the political message was integral to the plot.

Shaw was (on occasion) happy to acknowledge Ibsen’s influence on his work; he was less happy to admit the influence of the well-made play. One immediate effect of this influence is the location, look and milieu of the plays: although the settings are intriguingly various (moving from outside to inside, cleverly exploiting different times of day) the dominant milieu is the familiar one of the servanted classes at home. Despite their subjects, we never visit a slum tenement or a brothel in Widowers’ Houses or Mrs Warren’s Profession; we never meet a victim of Sartorius’ or Kitty Warren’s grisly trades. But more fundamentally, the influence of contemporary popular drama gave Shaw a template of emplotment into which he could insert a contrary set of meanings, by the simple device of denying the audience’s expectations of where the plot would lead. In all of the Plays Unpleasant, Shaw sets up a moral dilemma for his central characters, absolutely in the manner of the Scribean well made play, if not in two of the three cases with its usual matter. What he then nearly does in Widowers’ Houses, fails to do in The Philanderer, and triumphantly succeeds in doing in Mrs Warren’s Profession is to defy the audience’s expectations of how the plot will be resolved, without losing plausibility or denying its own terms.

Before seeing how Shaw does this, it’s worth looking at the opening of the plays, to see how skilfully – even at the outset of his career – Shaw establishes his characters, their situation and their dilemmas. Again, his beginnings distinguish Shaw from his mentor: however brilliantly he manages his denouements, Ibsen was usually pretty hamfisted with his exposition (The Wild Duck is by no means the only play in which the first act consists largely of one central character telling another central character what they both already know). The opening of Widowers’ Houses on the other hand tells us within seconds who Cokane and Trench are by the simple expedient of hearing them discuss what sights they wish to visit on the current stage of their improving continental tour (‘There is a very graceful female statue in the private house of a nobleman in Frankfurt. Also a zoo. Next day, Nuremberg! Finest collection of instruments of torture in the world’), in the same way that, in Candida, we learn all we need initially to know about the politics and personality of the Rev. James Morrell by hearing him finalize his upcoming diary engagements. This cunning use of Baedeker as a clue to character is then reiterated, not only to establish the next set of characters, but also to remove two of the subsequent assembly from the stage so that a proposal of marriage can take place under the pressure of their imminent return. And just as this device is in danger of wearing thin, Shaw introduces another, when Sartorius (for reasons about which we are already intrigued) asks that Trench write a letter to his relatives soliciting approval of his engagement to Sartorius’ daughter, a task which falls to Cokane, who then calls upon Sartorius’ assistance to complete it. Thus Shaw can map both the spoken and unspoken assumptions of three of the main characters concerning an as-yet-unrevealed skeleton in one of their closets, by the expedient of having one draft a letter on behalf of another in collaboration with the third.

Shaw does not pose himself nearly as much of an expositional challenge in The Philanderer, though it has to be said that he none the less starts the show with a bang. The initial stage direction reads ‘A lady and gentleman are making love to one another in the drawing room of a flat in Ashley Gardens’, from which Shaw goes on to chart Leonard Charteris’ politics, attitude to marriage and questionably-concluded liaison with a third party, in preparation for the entry of that third party, whose opening line unconsciously expresses the very attitudes that Charteris has been repudiating a few moments before. But most elegant of all is the opening duologue of Mrs Warrren’s Profession, in which a middle-aged man attempts to find common intellectual and cultural ground with what he and we see as ‘an attractive specimen of the sensible, able, highly-educated young middle-class Englishwoman’, who chooses to defy his expectations of such a person, thereby effortlessly establishing not only what sort of person she is, but (via his assumptions) what sort of person he is as well. In addition, Shaw has set up a series of vital trails and teasers for the future (including Vivie Warren’s ignorance of her mother’s occupation and intention to become an actuary).

By the end of the first act of both Widowers’ Houses and Mrs Warren’s Profession, then, Shaw has established his agenda, not by stating it, but by posing what is in fact the same question: how does one of the central characters earn their living and what effect will the answer have on the rest? Because he is more skilful by then, Shaw has inserted a false trail into his third play: for a moment, we are fooled into thinking that the big secret is not Mrs Warren’s profession but Miss Warren’s parenthood (and if we’re really clever we note that ‘profession’ can mean assertion as well as occupation, and thus could well apply to both scenarios). But, in fact, the third play like the first is about a presumably virtuous younger character having to confront their dependence on the wickedness of an older one, and the question Shaw poses at the end of each play is how they and the rest respond to this knowledge in practical terms.

For it is in the endings that Shaw’s meaning is revealed. The whole art of the stage is dedicated to concealing a single dirty little secret: that we know how most plays end before they begin. In tragedy, this is because the audience typically know the story already; in comedy it is because, from V century Athens to the end of the nineteenth century, the vast majority of comedies consisted of two young people overcoming parental or quasi-parental obstacles to their union, and getting married. Even in our century, with its bewildering array of new genres, we find that while we may not know the outcome of the story (who did it) we certainly know the ending of the plot (the murderer will be unmasked). Be the milieu the western, the thriller, the spy story or the romance, we will know from the outset who is the villain, who is the victim, and who is the hero, and thereby pretty much how the thing will turn out.

What Shaw took from Ibsen was the blindingly simple idea that this doesn’t have to be the case. As he put it in the ‘Quintessence’, the new drama ‘arises through a conflict of unsettled ideals’, and the question which makes the play interesting ‘is which is the villain and which the hero’. But by setting his plays within familiar theatrical milieus, Shaw gave himself an additional theatrical weapon. In Ibsen, there are no familiar landmarks to help us decide what kind of territory we’re in. In Shaw, we think we know where we are (six of the seven Plays Pleasant and Plays Unpleasant appear to promise a betrothal) but in fact we find we are somewhere else. In Ibsen we don’t know who the hero or the villain is, so we have to work it out for ourselves; in Shaw, we think we know but we find we’ve been deceived.

No wonder then that the inexperienced Shaw had trouble with his endings. Only in his third play does he bring off the reversal he has been striving for in both the others. In The Philanderer, his first, implausible but dynamic, ending was jettisoned in favour of a kind of evasion.