The Textual History of Tirso's «La venganza de Tamar»

Alan K. G. Paterson

In Tirso studies, there is a dearth of textual criticism. Apart from G. E. Wade in his examination of La Santa Juana, primera parte, no one, as far as I know, has taken real advantage of the manuscript material in the Biblioteca Nacional. Yet this one collection can yield unexpected results; several manuscripts, for example, offer versions that are substantially different from the texts in the printed Partes. The aim of this article is to piece together four versions, printed and hand-written, of Tirso's fine play, La venganza de Tamar. I hope to demonstrate that Tirso's plays reward textual study, and to suggest tentatively that the author submitted the stage-versions of his plays to a calculated process of revision in order to fit them out for the printed page.

Most of Tirso de Molina's plays were published in four Partes between 1634 and early 1636; the quantity and speed are all the more astonishing when we remember that he had stated his intention of publishing comedias as far back as 1624, in the prologue to Gigarrales de Toledo - a promise meagrely fulfilled in the tired-looking Doze comedias printed in Sevilla in 1627 by Francisco de Lyra¹. There is nothing slovenly about the later Partes. Their elegance and typographical care remind us that volumes of printed comedias, at least after Lope's personal supervision of his own ninth parte de comedias, could reach high standards of presentation². The plays were collected by a nephew, Francisco Lucas de Ávila, whose well-known account of how he stole them from the drawers in his uncle's two writing-desks is detailed enough to cast doubt on the theory that Lucas de Ávila was a decoy invented by Tirso to bluff his detractors. In the prologue to the Parte tercera (Tortosa, 1634), Lucas coyly warns his reader off an issue that I intend to examine here:

Tirso, that is, provided the material (presumably overlooking his nephew's first betrayal), and Lucas de Ávila edited it. In the Quarta parte (Madrid, 1635), we have another indication that the texts were prepared before being sent to the printers; in a piece of prosopopeia, the book confides that

Whether Lucas de Ávila is an apocryphal character or not, we have two clear statements that copies of the plays were made, with Tirso's knowledge and possible help, and that alterations were carried out during transcription. These two passages can be usefully kept in mind, a thread in the labyrinth of La venganza de Tamar.

Three complete texts of La venganza de Tamar have been preserved, two printed and one in manuscript. The text in the Parte tercera has been followed in all recent editions. Sometime during the first three decades of the eighteenth century, the printer Francisco de Leefdael issued the play in suelta form in Sevilla³. In the Biblioteca Nacional, Madrid, a manuscript version dated 1632 is preserved. The mutual relationship between these three texts is close, yet they all retain highly individual characteristics. Indeed, they raise a textual problem of such complexity that its final solution must await the possible discovery of fresh information which has escaped the present writer.

The manuscript (hereafter referred to as M) offers the most convenient point of departure into discussion. Its title-page bears an original inscription: «M.S.P. | fuerza de tamar | Anonimo | 1632 |». Three notes have been added: the name «Godínez» has been appended below «Anonimo», followed by «identica a la venganza de thamar»; further down the page appears «Letra del Lbro | Martinez». To judge from the information contained in Paz y Melia's descriptive catalogue of Biblioteca Nacional play manuscripts, the three later annotations were the work of Agustín Durán. His additional information probably came from Hartzenbusch's edition (Teatro escogido de Fray Gabriel Téllez, 12 vols. (Madrid, 1841) x, 398), where the suelta attributing the play to Godínez is described. All, therefore, that the original description tells us is that M belongs to 1632, that the play had a variant title, and that the author had been accidentally or deliberately forgotten, either in this copy or in the textual tradition that M represents. As a text, M is not perfect; dittography, haplography, omissions, and doubtful readings make up the usual catalogue of scribal sins. Few are serious. There are several hints that M is a theatrical text, whereas the Parte (P) is suitable for the printed page. For example, when Tamar and her maid Dina make their first entrance, the lines in M and P are distributed thus:

M has identified Tamar and her servant to an audience, an unnecessary detail in a printed version where character cues do the job of identifying. Another striking example occurs when Tamar sings the «romancillo» «Pajarito que vas a la fuente»⁴; in the preceding lines in M exeunt Tamar and Dina:

P indicates no exit, nor does the text cover one. M, that is, is geared for performance and permits a graceful «primera dama» with an unsuitable singing voice to retire to the wings, where a talented voice can take over. Another interesting example: just before the climax to Act II, M takes Tamar and Amón off-stage, presumably out of respect for propriety. For a reader of P, this is an unnecessary refinement and is omitted. We can add that M economizes on actors and actresses; Jonadab and Eliazer assume the roles of «criados» in Acts I and III; Michol is entirely suppressed in Act II. Details such as these leave one in little doubt that M was made for the theatre.

We meet with our first surprise when we compare M and P more closely; it soon becomes apparent that the two texts represent two different versions of the play. The total effect may be, to all intents and purposes, unchanged, but individual lines offer variant readings. A glance at the opening lines of Act I illustrates this sufficiently, although it is no exaggeration to say that few lines in the entire play are not subject to alteration between the two texts:

In fact, we have two authentic and basically sound texts with variants which only Tirso's, or Lucas de Ávila's, stylistic criteria can explain. On occasions M introduces material omitted from P, for readily understandable reasons. Eliazer's diatribe against quackery, for example, in Act II is extended by forty lines in M to include a scabrous tale about a frisky quack and an anaemic girl:

M, however, is more ready to pare down the text; only twelve lines of David's triumphal oration appear (Act II, Scene 4); the dialogue between David and his sons when they hear of Tamar's disgrace is shortened (Act III, Scene 4). P lingers over decorative elements; M omits completely the song on the various kinds of fleecing in society and gives only the «estribillo» to the song «Que si estáis triste la Infanta» (Act in, Scene 9). But the most significant difference is that M ceases to follow P in Act in after the lines «todo Israel me ha de ver | en alto por los cabellos» (Blanca, III, 403b), and adds ten lines which conclude the play; i. e. David's lament and the return of Salomón and Adonías are omitted. Although the discrepancy in length between the two texts is unimportant, stylistic differences are enough to make the problem of priority interesting and instructive.

It is common sense to suppose that at the outset Tirso wrote one version of La venganza de Tamar for the stage, which we may call O. An analysis of prosody offers no sure evidence about the affinity of M and P to the original. M is certainly at fault in some passages, but so also is P; e. g. in the following «quintilla»:

For want of better evidence, I suggest that those variants associating M with performance are sufficient to establish its chronological priority; the theatre, after all, preceded the book. We could add to this argument several details which, although insignificant on their own, can best be explained if we think of M's being revised into P. In the opening lines of Act I, quoted above, the «aquestas» of M appears as «estas» in P. According to Corominas, the demonstrative adjective «aqueste» underwent a semantic change in the early seventeenth century: it became archaic. Tirso would hardly have altered a modern form into one already old-fashioned; it is more probable that he retained the «aquestas» of line 1 to provide an archaic note, and then altered the second line, carefully ensuring that the number of syllables remained correct. This guess is strengthened by a similar revision in Act I, Scene 8 (Blanca, III, 372a):

In this example, the reviser nodded and failed to harmonize his correction with the romance prosody. In both cases the variant «aquesta» — «esta» is best explained by the sequence M → P. Secondly, as Eliazer is mimicking a dialogue between a group of doctors in Act II, Scene 1, his enumeration of the luxuries indulged in by charlatans concludes in P with

These «hypócritas» are not, presumably, doctors, but those whom Quevedo described as in «poder de culteros», the affecters of the «cultiparla», for whom «vaquita» would facetiously echo the Latin diminutive «vitellus». In the last line, Tirso must have been poking fun at a «culto» mannerism, a literary game which he played zestfully, it seems, from 1619 onwards. However, M ascribes vaquita to «los estudiantes de ogaño». This is a more «natural» reading, for outdated student parlance is what one would expect from a quack. M presents the original reading, changed at a later stage to disparage a «culto» affectation.

The suelta text (S) interrupts this uncomplicated textual sequence. We have already noted that S attributes La venganza de Tamar to Doctor Felipe Godínez (the only occasion on which his name is mentioned in an early version, since the manuscript annotation is in all probability the work of Durán and derives from S itself). A few biographical details about this forgotten playwright are enough to explain his presence on the S title-page. Godínez, a contemporary of Tirso, appearing in the Viaje del Parnaso (1614) and as late as 1657 in a Madrid certamen, was a priest, preacher, doctor of theology, Buen Retiro poet, and a dramatist of repute in his time⁵. Today, he is remembered for less glorious reasons: he was accused of being a crypto-Jew, reconciled by the Inquisition in Sevilla in 1624, and then allowed to resume his career in the Church⁶. Inevitably thereafter, his name was associated with Judaism; four years after the author's reconciliation, Lope described a Godínez play entitled La Godina as «más judía que de los godos»⁷. The good Doctor's predilection for Old Testament themes in his theatre can scarcely have helped to restore his damaged reputation. All this explains why Godínez's name appears in Leefdael's suelta: at one stage La venganza de Tamar must have lost its true author in the course of exchange between performing companies, and, given its Old Testament background, was attributed to the Old Testament dramatist par excellence.

It has been suggested that S is a refundición; this is hardly so, for S is a simple hash of a brilliant original. It massacres entire scenes and knits amputated passages together with clumsy additions. Two major additions, however, do not bear the crude mark of a literary surgeon's stitches: S terminates Act I with twenty-five lines quite distinct from M and P, modelled on the exemplum of Valerius Maximus, «Si eos, qui nos amant, interficimus, quid his faciemus quibus odio sumus?» (Factorum et dictorum memorabilium libri novem, Book V, cap. 1); at the end of Act III, immediately after the murder of Amón, S parts company with P and M, providing one hundred and sixty unique lines in the course of which Absalón meets his death, is reunited on stage with his brother's corpse, and is lamented over by a distraught Tamar who is restrained by her father from plunging a knife once more into her half-brother's body. Despite its terrible defects, S has peculiarities that lift it far above the status of a lowly hash; behind the ill-treated text, one can detect a third version of La venganza de Tamar.

The most enigmatic feature of S is that it bears no direct affiliation to either M or P. We can only predicate, on the strength of this, that it came from a lost version (let us call it S1) which enjoyed a measure of independence from M and from P, although it rarely offers a reading not to be found in one or the other. Let two examples illustrate this point. The first is Eliazer's complaint about travelling by relays of post-horses («postas»):

Here, as on many occasions throughout the play, M and S share the same text and P offers an alternative. Yet there are as many instances of the following type of variant (also from the opening scene):

In such cases M offers the variant text while P and S remain in agreement. Furthermore, the interweaving of the texts is not limited to one-line units; there are entire passages found in P and S, or in S and M, but missing from M and P respectively. Isolated examples tend to diminish the size of the textual problem; if we remember that throughout the entire play the three texts constantly change their allegiances, then the enigma becomes colossal and mystifying.

Verse analysis does nothing to help relate M, P, and S. For one thing, S is an unreliable text thanks to Leefdael's careless printing and the haphazard cutting. Moreover, M too has its deficiencies, not to mention P. Hence, to find a passage common to P and S but omitted in M is no indication of filiation, and so on through the various permutations. One observation can be maintained: since M, P, and S are interrelated, yet mutually distinct, there must have been one stage in the play's development when one version was established by someone's choosing in an eclectic fashion from the other two. We need not visualize a scribe copying from two texts placed on his working-desk; more probably his copy text was a «borrador» carrying marginalia, interlinear annotations, legible corrections and so forth, some of which he incorporated into his third version. It can be appreciated that the order in which the three texts appeared could have followed one of three courses:

(Since M is assumed to be earlier than P, we cannot entertain any stemma in which M derives from P.) The second line is, I think, the most fruitful to follow: S1 is the earliest text, revised to create an intermediary text, M, which as far as textual transmission is concerned existed only as a series of annotations imposed on S1. A further revision took place based on the copy S1 + M, and P was the product. If we incorporate revision and integration into a genealogical diagram (indicating revision by a broken line and integration by an arc) this is the result:

Are there reasons for considering this order of transmission preferable to any other?

First, we must return to Lucas de Ávila's phrase «poniendo éste nueuas añadiduras». This surely means that the texts from which the late Parte versions emerged were «borradores», similar to the Santa Juana autograph, which bore an original text and subsequent corrections or annotations. Lucas de Ávila had to play the role of a modern editor, choosing between the original wording and recent alterations, possibly with his uncle's assistance. It is difficult to imagine the three texts of La venganza de Tamar coming from anywhere other than Tirso's study; their intimate disparity is best explained by such circumstances.

Secondly, one particular detail in S seems to justify the second line of transmission. In the scene where Amón rejects the various women suggested to him by his half-brothers and servants, P and M give the same reading for these lines:

The word «ojosa» has no meaning at all in this context. Now see the sense that S makes of the lines:

Elisabet, to hide her pitted complexion, defends herself with lavish make-up; the sense is perfect. On other occasions S carries the manifestly correct reading; e. g. in the opening scene, the obscure lines in P «ni viejo admite la paz, | ni mozo quita del lado» appear as «ni viejo dexo la paz, | ni mozo quito del lado»; despite the fleeting oddity of «dexo» (should it read «dexa»?), S makes better sense than P. Consequently, we can, in all probability, cancel out the first and third stemmae.

Finally, in order to argue that S is the residue of the earliest version of La venganza de Tamar, we are justified in dealing an interpretative stroke to cut a textual knot. This play concerns justice; it explores not only one character's entry into forbidden passion, but the consequences of passion upon the society in which he lives. A nation which has emerged from a war against an outside foe finds itself at war with itself, a war fought at the very centre of power. As the third act gathers momentum, two characters are lifted into prominence, acquiring dramatic shape as they respond to the conflict of motives and ideals which Amón's crime precipitates. David, as a father, pities a wayward son; as a king, he is called to uphold justice. Absalón, ostensibly acting as the instrument of outraged decency and loyalty to principle, skilfully removes an elder brother who blocks the path to the throne. Both men create a situation in which justice cannot discriminate between objective retribution and personal desire. A fresh conflict grows out of Amón's crime; mercy and justice, justice and private motive, enter into tragic confusion. As these threads of argument are drawn together in the final scenes, the three texts desert each other and follow very different paths. S includes the death of Absalón, and a gruesome moment when Tamar tries to plunge a knife for a second time into Amón's corpse⁸. Exegetically speaking, the S ending is more faithful to the traditional interpretation of Tamar's violation: biblical scholars encompassed the murder of Uriah, seduction of Bathsheba, rape of Tamar, murder of Amón, and death of Absalón within one exemplary theme. Artistically, S underscores retribution: the King's two sons perish, Tamar triumphs and sorrows, and the voice of the stricken father is scarcely heard. The Parte version ends on an infinitely more subtle note: David stands unrivalled at the conclusion, broken by his son's death and by the failure of forgiveness. A simple reorientation draws the spectator into a deep emotional and intellectual drama. It is credible that Tirso moved from the «exegetically correct» ending to a thematically more subtle one. P represents an enlightened reworking of the original.

As for M, it is cut short in the final scenes, after following P for about thirty lines. Since one tendency of M is to reduce throughout David's role (presumably to suit the requirements of the performing company), we can assume that an untampered M would have terminated the play in the same manner as P.

At the present stage of the argument, it looks as if S, the latest and worst text of La venganza de Tamar, is in fact a mutilated printing of the earliest version. As we pursue the problem further, this order of transmission satisfies, at least partially, the fresh questions that arise from the second act of Calderón's Los cabellos de Absalón.

Scholars have long been aware that the second act of Calderón's Los cabellos de Absalón was lifted virtually intact from Tirso's La venganza de Tamar. What has not been known is which text Calderón borrowed from the senior dramatist⁹. The question has been raised by Professor Sloman. Here, one may think, is the occasion to provide a definitive answer. Surely one of the three texts which have been discussed was the one employed by Calderón? This is not the case, however. Before demonstrating this, we must organize the textual history of Calderón's own play, best known through the Vera Tassis edition of 1684. The Vera Tassis version is not the earliest; a British Museum suelta is the first in a small family of printings¹⁰, betraying in its errors a close affinity to La venganza. For example, Tamar's speech in La venganza, Act III, from «Mayor ofensa, y injuria» to «y a mí dame más enojas» (Blanca, III, 390a) is reduced in the suelta (let us call it C) to

This disfigured text is repaired in a second group of sueltas, represented by one in the New York Public Library and another in the British Museum¹¹; in both, the truncated line is replaced by «como burlas assi, ingrato»; Vera Tassis spotted the error in prosody, and altered the line to «como burlan tus antojos». As far as we are concerned here, the C text of Calderón's play is of immediate consequence.

Calderón, needless to say, altered isolated passages in Tirso's original so that they harmonized with his own dramatization; such alterations are self-evident. The variants that cause real difficulty are those between C and the three extant versions of La venganza de Tamar, for a little close study shows that C derives from none of them, although it approaches the parte text more than do M or S. For all that, C belongs to a pre-Parte category, e. g.

• P.   (Absalón)   Valgame Dios, que vozes seran estas? (Blanca, III, 392a)
• M   dau.   balga me dios que vozes son aquestas
• S   Dav.   Valgame Dios! que vozes son aquestas?
• C   (Absalón)   valgame Dios, que vozes son aquestas!

There are approximately forty other examples that indicate that C did not come from P, but from some previous version. Yet this previous version was not S1; C advances considerably towards P, following the Parte ending and rejecting in favour of M and P many of the readings preserved in the Leefdael suelta. Nor was Calderón's copy-text M. He must have handled a fourth text, almost identical to P but still unique; let us call this intruder S2, and try to identify it.

The relationship between M and S2 is one that eludes me. On occasions, Los cabellos de Absalón rallies to the parte version of La venganza, while M clings to an S1 reading, e. g.:

• P:   del torpe principe Amon (Blanca, III, 392b);
• M:   del cruel principe Amon;
• S:   del cruel principe Amon;
• C:   del torpe Principe Amon.

On other occasions, the process is reversed, e. g.:

• P:   pegome su contagio (p. 392b);
• M:   pegome su contagio;
• S:   su contagio me pego;
• C:   su contagio me pegó;

My own impression won from the perplexing mass of variants produced by M and C is that the latter edges closer to P. As for forming a final picture of this intricate textual family, we do best to fall back on the idea that Tirso revised annotated «borradores» for his final, printed versions. Let us suppose that S1 was first revised to extract the text contained in M; as far as transmission is concerned, M left traces on the S1 copy, but thereafter qua M it became detached from the parent text. Then, at a later stage, S2 emerged as the product of further revision, leaving a harvest of fresh readings on the original S1 copy; at times, S2 would ignore the «añadiduras» which were bequeathed by M, and return to the original version. At a still later stage, Tirso used the overlaid S1 text to produce the definitive version in the Parte. This version could inherit material from S1, from the first stage of revision, and from the second stage of revision, not to mention creating entirely fresh lines where Tirso or Lucas thought fit. If this reconstruction is correct, two consequences should be expected. First, no reading unique to S1 should find its way into P; readings from S1 transmitted to P had to be accepted either by M or by S2 for the simple reason of legibility; it would be extraordinary if a version unique to S1 made its long way unnoticed into P. A comparison of the four texts fulfils this condition: no reading unique to S1 (i. e. as far as one can tell from S) is found in P. Secondly, it would be surprising if C and M were to agree together exclusively. This point needs elaboration: M, one suspects, must contain readings which left no trace on S1, and which could not, under any condition, find their way into C once M began its separate existence. In fact, there are instances of agreement between C and M to the exclusion of P, but in only one of these instances can one assemble four texts for comparison due to a breakdown in S; the one instance is insignificant. M, therefore, where it becomes an idiosyncratic text, left no impression on C. To conclude the reconstruction, a straightforward line of transmission can be established, interrupted by an arc where revision took place:

Textual studies may have to pursue devious paths to arrive at well-defined conclusions; in the case of La venganza de Tamar, we have had to use involved means to find a stemma which is disarmingly simple. S1 progresses steadily towards P, accumulating new material as it gives birth to each new version. We have the evidence of the Santa Juana manuscript that Tirso wrote «borradores» which bore legible variant texts. This in itself is commonplace; men rarely write identical versions of their own work. Otherwise, what need of editors? Lucas de Ávila was probably the first Tirso enthusiast to fall foul of textual problems as he prepared his uncle's manuscripts for the printers. Is there not a hint of legitimate pride when he associates himself with the final revision of these plays? But the history of La venganza de Tamar tells us something more important. The frequency of alteration between the various texts is too great for us to dismiss revision as haphazard; if we assume that seventeenth-century playwrights avoided soiling their hands in the correction and publication of their texts, the assumption should be reconsidered in the case of Tirso. Working over the various texts of La venganza de Tamar, one gains the impression that the evolution of the play is connected with a desire for improvement; stylistically and thematically, the Parte version reads with greater assurance and subtlety than any other. Keeping in mind that Tirso's major publications took place long after he had ceased writing for the theatre, I suggest that his final versions were directed by a wish to adapt a theatrical mode to the printed book; the new standards set by conscientiously prepared and printed partes de comedias must have exerted an incalculable influence on later dramatists (and here I think of Calderón), who inherited the audiences and aspirations of their immediate predecessors. Spain, unlike contemporary England, never experienced a full-scale «war of the theatres» which ended an alliance between popular and learned taste; that the comedia developed intact may, in some measure, be due to the careful transference of an oral art into a prestigious, printed form. Tirso's efforts, illustrated particularly in La venganza de Tamar, warrant more than passing attention.