Terence was praised in antiquity for the excellence of his plot construction. Donatus deemed as praiseworthy the existence of two love affairs (bini amores) in all Terence's plays but the Hecyra, and Evanthius commended the richness of Terence's plots (locupletiora argumenta) constructed of double affairs (ex duplicibus negotiis), likewise observing that all the plays except the Hecyrs feature two young men in love. Terence's plays continued to be read, admired and even sometimes imitated in the Middle Ages, and the Renaissance comedy was modelled principally on his comedies. Herrick, who examined the leading commentaries on Terence from the fourth-century work of Donatus and on, has shown that Terence was considered a master of dramatic structure and his comedy monopolized the sixteenth-century discussion of comic theory: the 'dramatic' rules established for comedy mostly derived from the practice of Terence. On the issue of the double action, Donatus's observation was usually repeated and the double structure commended for the enriching of the action. Even after the interpretation of the single action recommended by Aristotle and Horace became established, the double plot continued to be practised in the Renaissance comedy, which, as is well known, influenced comedies of later periods. It is therefore interesting to note the change of attitude among modern scholars and critics towards Terence's plots.

Legrand, [in his The New Greek Comedy, 1917], for example, required the two love affairs to be of equal importance and tightly connected one with the other, using these two arbitrarily selected qualities of the double plot as a norm of evaluation. Accordingly, the Phormio received poor marks for having the two love affairs 'run parallel and without influence upon one another for too long a time' (i. e., they are not tightly connected), and the Andris was criticized, since 'one of the lovers becomes a matter of indifference to the spectators', and thus fails the requirement of 'equal importance'.

These requirements are based on an Aristotelian conception of dramatic unity and transfer Aristotle's critical method and criteria from tragedy to comedy. The double plot is regarded as artistically satisfying only if it fulfils the requirements of one action which evolves according to the rules of probability and necessity. The greater the extent of the causal interaction of its two issues the nearer will the double plot be to a representation of one action, whereas their equality makes sure that neither may be regarded as a subplot. Such requirements not only demand that plots of comedies should conform with the plots of tragedies, but they also ignore, or criticize, plots with two issues of varying importance, or plots with other, not causal interconnections.

Legrand's normative approach was adopted by Norwood [in his The Art of Terence, 1923], who augmented it with yet another requirement. He was the first to coin the term 'the duality method', defining it as 'the method of employing two problems or complications to solve each other'. Since Norwood believed in evolution and in intellectual progress, he deemed early works as immature and discerned in Terence's comedies a constant ascending line of improvement towards perfection. Thus, he proclaimed the Andria immature, which coincides with Legrand's view, but disagreed with Legrand's negative evaluation of the Phormio, which, being a later play, should reveal traits of artistic perfection. In this he was followed by Duckworth [in his The Nature of Roman Comedy, 1952], who also canonized 'the duality method', which, thanks to him, came to be taken almost for granted by others.

An amusing illustration of such a normative approach is the following criticism of the Andria: 'Charinus and his love for Philumena do not suit the standards for the dual plot as defined by Norwood and as practised by Terence in his later plays, particularly the Phormio'. Surely it is not Terence who has to suit Norwood's, or any other scholar's, standards. Normative evaluation based on arbitrarily selected or invented ad hoc dramatic qualities is unhelpful. What is needed is an understanding of the actual ways in which Terence's plots are constructed, an understanding which can be achieved through an unbiased reading of the plays, without any preconceived attitudes.

For quite a long time scholars and critics of the Roman comedy were chiefly interested in establishing the degree of originality in Terence's use of his Greek models and in their speculative reconstruction, to the relative neglect of other questions. Bent on finding contradictions, inconsistencies and other proofs of Terence's inferior workmanship, scholars used a methodology borrowed from Homeric studies. This contributed a great deal no doubt to the lack of interest in characteristics specific to the dramatic genre. Plays were treated in a way similar to that used in examining epic narrative as texts to be read, not as scripts to be performed. This attitude, more than anything else, has affected the treatment of dramatic plots, either single or double.

In this [essay] I propose to examine the question of 'the duality method' through a reading of one play, the Andria, Terence's first comedy.

A typical summary of any of the Terentian double plot comedies by scholars who subscribe to 'the duality method' tells the story of the two love affairs, stressing the degree of their importance and their interconnections. Duckworth, for example, summarizes the plot of the Andria as follows:

In the Andria, Pamphilus—passionately devoted to his mistress Glycerium—does not wish to marry Chremes' daughter, who is beloved by Charinus, and Pamphilus assures Charinus that he has nothing to fear. When Davus' plans go astray, Charinus' hopes are temporarily shattered but the recognition of Glycerium as Chremes' daughter solves the problem of both young men, and Charinus is free to marry Philumena.

Factually, this is a correct summary, but it is construed of various elements of unequal dramatic standing. It tells the story of the Andria. But while a play has a story to tell, the story of a play does not necessarily coincide with its dramatic stage plot, namely with the words and deeds of the characters who enact before the audience the actual stage events. The dramatic stage plot differs from what the formalists call syuzet, which is the order and presentation of the events in a narrative. It may contain narrative elements which describe past events in a non-chronological order, but these are described in a direct discourse by one of the characters and form a part of a present tense stage event. The sum of all present tense stage events is the dramatic stage plot.

Readers of a dramatic text who wish to construct its dramatic stage plot must think in terms of performance and take into account only what the play's characters are to say and do on stage before an audience. What is narrated by the dramatis personae as being said and done elsewhere are narrative elements incorporated into the present tense stage event, but with a different mode of existence. In order to construct a dramatic stage plot of a play, one has to compose a short story, namely to substitute the play's dialogues with a narrative. But, since plays are not texts conceived in terms of a narrative medium, they usually defy attempts at extrapolating their plots by playing tricks on the summarizing narrator. For it is easy to narrate what is narrated, and it seems almost natural to perform the transformation of the dialogue into narrative by adulterating it with actual narrative elements picked out of the dialogue itself. The result is a story of the play not at all representative of its dramatic stage plot. Since stories are not plots, substituting one for the other leads to erroneous analyses of the plays misrepresented in such a way.

Duckworth's previously quoted summary of the Andria's plot is a fine example of such a story. It not only misrepresents what actually happens on stage, but also distorts the nature of the Andria's dramatic conflict, for it creates the false impression that the two love affairs are equally important and that they are the main interest of the dramatic plot, that, in fact, the Andria is a romantic comedy. Terence's comedies, however, are not romantic dramas, but intrigue plays in which the emphasis is not one the lovers themselves but on the persons who plan to bring them together or set them apart. The two young men are not the main protagonists of the play and the girls do not appear on stage at all. If the importance of a character is commensurate with the length of his role, the first role belongs to the slave Davus, the second to the head of the household, the senex Simo, and only the third to his son Pamphilus. The same conclusion is bound to be reached if the importance of a character is measured in terms of the dramatic conflict he originates or in which he is involved. Even a brief glance at what actually happens in the Andria reveals that the action stems from a conflict between Simo and Davus.

The antecedents to the play's dramatic conflict are described by Simo in the exposition: his wealthy friend Chremes, who, impressed by Pamphilus's allegedly exemplary behaviour, had offered Simo a marriage deal, had withdrawn his generous offer after a rumour reached him of Pamphilus's affair with the meretrix Glycerium. Simo plans to find out the truth and test his son's obedience by pretending that the marriage is to take place as arranged. He also hopes to fool his slave Davus into believing that the marriage is real, so that his schemes against it will be wasted harmlessly. Simo's main concern is that it will be not his son but Davus who will foil his plans. He deduces his son's unwillingness to get married from Davus's fear, and it is Davus whom he orders to see to it that his son will consent to the marriage. When Simo warns Davus not to try to pull any smart tricks, it is clear that it is now Davus's turn to do exactly what he is warned not to, that is to show how clever and inventive he is.

Thus, from the onset of the play, it is obvious that the two protagonists who advance the action are Simo and Davus and not Pamphilus, the adulescens in love. When Pamphilus is brought on stage for the first time (Act I, Scene 5), it is not to offer a course of action but to clarify his position, which up till then has been reported by a third party. In order to appraise correctly the schemes devised on his behalf, the audience has to hear directly from him what is his attitude to Glycerium and the proposed marriage, but it is Davus, not Pamphilus, who acts. His scheme counteracts Simo's, and in spite of all its (prepared) unexpectedness it neatly parallels Simo's stratagem: it is a bluff pitted against a bluff. Davus has found Simo out, correctly guessed his aim, and consequently advised Pamphilus to agree to the pretended marriage in order to avoid disobedience and to embarrass his plotting father. Pamphilus does whatever he is advised to do.

Both Simo and Davus are prompted to action by mutual disbelief, both constantly stand on guard and examine each other's moves. The chief comic ingredient of these situations is that each schemer, confident of his own cleverness, is led to a false assessment of events, disbelieving what is true and vice versa. Simo is, of course, only too happy to believe that his son is willing to marry Philumena. But, when Davus in his eagerness to eliminate all doubts overplays his hand with the explanation that Pamphilus's sadness is caused by Simo's close-fisted preparations for the marriage, Simo's suspicions are alerted (What is the old plotter (veterator) up to? 467), but his assessment of the events is false. Simo believes that the birth of Glycerium's child (his grandson) is staged in order to scare off Chremes. Not only is he led to disbelieve what is true, but, conditioned to explain events according to his expectations, he also jumps to the false conclusion that he has succeeded in fooling Davus exactly as he had planned, namely that his scheme of the pretended marriage actually helped him to uncover and to annul Davus's machinations.

Davus is quick to take advantage of Simo's mistaken conclusions and immediately proceeds to exploit his disbelief: You are right, you have detected the truth. The baby is not Glycerium's baby, she ordered it to be brought to break the marriage which is to her disadvantage but which Pamphilus now wants. Since it falls in line with his reasoning, Simo believes Davus, takes his son's promise as a firm basis for further action, and proceeds as he initially planned. He persuades Chremes to change his mind again and to agree to the marriage (III, 3). Since Davus fooled Simo into believing that he, Davus, is not fooling him at all, Simo feels free to tell him about his initial plan of the pretended marriage. Now Davus is really gloating over his success, which reaches its peak in his ironic comment: Such cleverness! I could never have guessed it (589).

Immediately, however, he learns that the situation has been reversed. Simo, although fooled and outwitted, has actually achieved his goal, whereas Davus, the archschemer, finds out that his cleverness, instead of averting the marriage has helped to bring it about. Thus, for Davus, the first round ends with a setback: he must devise a new manoeuvre. It is surprisingly funny and structurally ingenious that Simo himself is the source of inspiration for this second stratagem. Davus decides to use the baby to scare off Chremes and stages an encounter with the slavegirl Mysis for him to witness (IV, 4). His scheme consists of two cleverly combined and balanced parts. Davus knows that the baby is the baby of Glycerium and Pamphilus, but by pretending that he believes it was brought by a midwife to deter Chremes he hopes to evoke a strong denial from the simpleton Mysis which will convince Chremes to believe the opposite. He also hints that Glycerium is an Athenian citizen with whom a marriage can be contracted. Although this is true, Davus considers it to be false. He adds it nevertheless in order to increase the baby's potential as a deterrent.

Davus's carefully planned and executed scheme—the funniest farcical stage scene of the entire comedy—is a great success. This time, the breaking of the second marriage offer (like the second marriage offer itself) is part of the actual dramatic stage plot and not of its expository antecedent narrative. In this way, the incidents of the dramatic action, by paralleling the narrated events of the exposition, achieve a unified continuity of the entire narrated and acted out sequence of events.

By pressuring Chremes, Davus succeeds in overturning Simo's plans. According to the best tradition of comedy in the battle of wits between the master and his slave, it is the slave who has the upper hand in the end. The final virtual power, however, is in the hands of the vanquished. The victor is at the mercy of the defeated. This part of the plot is neatly rounded off by Simo's carrying out his initial threat. He arranges for a physical punishment of Davus for not taking heed of his warning.

The arrival of Crito from Andros leads to the anagnorisisor recognition scene. His credibility is established by Chremes, whose acquaintance he is, and the entire episode is integrated into the mainstream of the plot by Simo's attitude of disbelief. Simo's doubting of Crito's integrity and his suspecting of yet another ruse iterates his former disposition and the attitude he displayed in all the previous instances which the audience has witnessed.

From the above short analysis it is clear that the opposition with which Simo and Davus meet, impersonated for each in the person of the other (the dramatic opposing forces), causes them constantly to alter their plans up to the turning-point of the play and its resolution. It is an intrigue plot, in which the complicating factor is the outwitting of the antagonist, not a romantic comedy, in which the lover himself takes a part and forwards the action. The chief ingredients of such a comedy are lacking: there is no courtship, no chase, seduction, persuasion, capture of hearts or conquest, in a word there is a lack of the chief ingredient, the battle of the sexes.

It is no wonder that the adherents of 'the duality method', who tend to upgrade the importance of the love stories, entirely misconstrue the Andria's dramatic plot structure. Norwood considers Davus a spurious character, 'a fly in the wheel', and 'a fifth wheel on the coach'. He is, also on this point, blindly followed by Duckworth, who writes: 'Davus … [is] a bungler whose suggestions and schemes confuse everyone but actually accomplish little.'

The secondary plot of the young Charinus and his slave Byrria is linked to the main plot, the Simo—Davus battle of wits, through causal and analogous means. Charinus is in love with Chremes' daughter Philumena, the girl whom Pamphilus, his friend, tries not to marry. This unique plot aspect of the Andria—in all the other double plot comedies of Terence each young man is linked with a separate young lady—is used as a causal connection of the two plot issues throughout the entire play sequence. Any fluctuation in Pamphilus's fate directly affects Charinus and repeatedly justifies his appearance on stage. This causal link also serves as an analogous combining element for the tying together and delineating of the actions and reactions of the two young men and their slaves. For even more important from the point of view of the plot structure is yet another unusual feature of the Charinus—Byrria plot element, the fact that Charinus does not have a family. As the only fatherless young man in the Terentian comedies he is also the only adulescens who does not need to secure his father's consent to a marriage. It precludes his plot issue from developing into a traditional intrigue plot in which a schemer slave outwits a senex for the sake of an adulescens in love. Where there is no senex, there is no room for any plotting with his fooling or persuasion as its goal. Thus, the inertness of Byrria is not only understandable but actually necessary.

Prevented from being a schemer, Byrria is not a mirror-image of Davus but rather his antithesis. He is represented as hindered from action by contemplation and by weighing of consequences. His cleverness, and clever he is, is not externalized in creative machinations but expressed in proverbial formulations of practical wisdom. The relations between Charinus and Byrria, presented as different from the relations between Davus and Pamphilus, provide the dramatic justification for Byrria's inaction. When Byrria proposes a course of action, Charinus refuses to accept it and accuses him of never giving him good advice. When Pamphilus urges Charinus and Byrria to devise a way for marrying Philumena while he will endeavour to avoid the marriage, it is immediately made clear that nothing will be done: Charinus is represented as quite satisfied with Pamphilus's declaration of intents (sat habeo 335). This is dramatized by Charinus's prompt dismissal of Byrria, which highlights the incompatibility of the pair. Consequently the next scene (II, 2) presents the dependence of the two adulescentes on the doings of one scheming slave, and ends with an analogous situation: Davus repeats Pamphilus's advice and urges Charinus to forward his case on his own, for no marriage for Pamphilus need not necessarily mean a marriage for Charinus. The repetition underlines Charinus's future inaction and the continuing lack of initiative. He is expected, as before, to be satisfied with whatever is planned by Davus and done by Pamphilus. It is, therefore, in line with these aroused expectations that his only line of action is to send the passive Byrria to find out Pamphilus's whereabouts (II, 5).

As a result, the causal connection between the two plot issues is highly effective. Charinus's mistaken accusation of Pamphilus as interested in marrying Philumena (IV, 1) has an ironic effect and adds to Pamphilus's accumulating heap of misfortunes exactly at the right point in plot-time, whereas Davus's promise to extricate Pamphilus by a new stratagem from the spot into which he 'succeeded' in putting him bears hope also for Charinus. And again, this scene's end parallels that of the previous one: Davus refuses to act on behalf of Charinus and the latter departs conveying the impression of continuing inaction. Finally, in the closing scene of the play, Charinus is represented as ultimately taking the route of action previously advised by Davus, to plead with the girl's father through his friends. The advice itself is unusual and stems from the unusual feature of Charinus's lack of family. Young men do not negotiate their marriages in comedies, it is done by their fathers. But now, once Pamphilus is found to be Chremes' son-in-law, Charinus can use his good offices for the achieving of his desire (V, 5). The approval of Pamphilus's marriage to Glycerium, with its entailed huge dowry, turns him from a dependent lover into a rich pater familias whose authority is instrumental in bringing about a happy ending of Charinus's plot issue. Consequently, the Charinus—Byrria plot issue is intertwined into the main plot issue also through the thematic dependence of an adulescens on an able slave, albeit not his own, as well as by the contrasting analogy with Byrria, which enhances Davus's scheming ability and rounds off his characterization as an all-knowing, all-providing factotum with a finger in each and every pie.

As befits intrigue plots, the resolution of Charinus—Byrria's plot occupies but a tiny fraction of the play's actual stage time. The resolution of the main plot, although somewhat longer, is still far shorter and less prominent than the intrigues through which it has been achieved. Moreover, both resolutions are foreshadowed and expected. What is new and unknown are the stage situations leading to them. What Styan writes on the Restoration comedy [in his Restoration Comedy in Performance, 1986] is applicable, without much change, also to the Menandrean comedies of Terence: 'the comedies were not constructed like modern well-made plays. The outcome of each play was more or less known for the unexciting thing it was, and there were to be no surprises.'

The chief requirement of Norwood and his followers for the Terentian double plot, that its two problems solve each other, is a theoretical construct of a reader, neat, clean and nice, but hardly stage-oriented. For a spectator, resolutions do not exist throughout the entire stage time devoted to the plot sequences; they merely terminate them. Obviously readers' constructs of dramatic plots have all the characteristics of an end product of a silent reading process. They resemble the constructs of plots of narrative fiction, whose readers need not consider the different essences of the plot's various component elements and the different degree of their stage prominence.

To sum up: from the above analysis of the actual stage happenings of the Andrin it may be concluded that both the plots are plots of intrigue, each of a different quality: one based on action, the other on inaction. The secondary plot highlights by a contrasting analogy the parallel but opposite elements of intrigue in the main plot. The modern tendency to define Terentian double plots in terms of love interests and call them romantic comedies obscures the intrigue elements and shifts the focus away from the stage action as viewed by spectators.

Before Legrand and others voiced their normative requirements, Donatus and other students of Terence described the Terentian double plot structure (see above) as pleasing in its variety. Thus, for example, Gibaldi Cinthio wrote (in 1554):

double structure … has made the plays of Terence succeed wonderfully. I call that plot double which has in its action diverse kinds of persons of the same station in life, as two lovers of different characters, two old men of varied nature, two servants of opposite morals, and other such things as they may be seen in the Andria and in the other plots of the same poet, where it is clear that these like persons of unlike habits make the knot and the solution of the play very pleasing, [in Herrick, Comic Theory in the Sixteenth Century, 1964]

What enriches the dramatic plot is the addition of a second set of characters whose interests and goals multiply the original problems thus adding further angles and possibilities of comparison. Where there are two sets of lovers, fathers and slaves, with two sets of problems, their very presence on stage as part of the actual dramatic action enriches the play. But a rich plot is chiefly a plot which richly activates the spectators' minds. Interest is created and tension is increased not only when the audience's mind is occupied with what is presented on stage but also with what is not. When one plot is acted out on stage the audience does not entirely forget the other, but rather tends to think that actions of the other plot presumably take place at the same time somewhere else. Thus, a double plot is ipso facto richer than a single issue plot. It is no wonder, then, that later comedies, and especially farcical comedies, mirrored the Terentian double plot and even multiplied it.