Filología y Lingüística XXVIII(1): 93-105, 2002
ROA BASTOS’S “NOCHE SIN FIN” AND “LA TIJERA”: 
TWO VERSIONS OF THE SAME STORY
Helene C. Weldt-Basson
RESUMEN
Este artículo estudia el modo en que Roa Bastos transformó el cuento corto “Noche sin fin”,
publicado originalmente en la revista Ficción (1957), en el cuento “La Tijera”, publicado más
tarde en su libro de cuentos El baldío (1966).
ABSTRACT
This study focuses on the manner in which Roa Bastos has transformed the short story
“Noche sin fin”, originally published in the journal Ficción (1957), into his history “La
tijera”, later published in his collection El baldío (1966).
Wilfred Guerin’s A Handbook of Critical Approaches to Literature, originally
published in 1966, defines the study of the genesis of a text as an “approach to a work of
literature... which studies the growth and development... of the work... it helps even the
beginning reader to know that we can more fully understand a work if we have, for example,
manuscript copies showing its stages of development, or if we know the sources from which
the author was drawing in developing his own work” (Guerin 197). This, the most simple and
traditional definition of the study of the genesis of the text, has been refined and embellished
over the years, particularly by a group of French theorists in the late 1970s and early 1980s
who began to focus on different issues and interrelationships pertinent to genetic theory. 
Genetic theory is indeed, in simple terms, the study of manuscripts and sources.
However, there are many different theories about the way in which these manuscripts and
sources should be classified and studied. There is a tremendous overlap between the genetic
analysis of the text and other theoretical approaches to literature, such as intertextuality,
psychoanalysis, and linguistics, just to name a few. In the structuralist era, when the text was
seen as a closed unit that should be analyzed without reference to the extratextual, genetic
theory was relegated to the category of mere source study with no direct bearing on textual
analysis. However, in the 1990s, amid the postmodern awakening, it is clear that genetic
theory is worth a second look, and may well provide the key to unraveling many pertinent
problems of textual analysis in contemporary fiction. 
Augusto Roa Bastos observes the difference between contemporary genetic studies
and the frowned upon, traditional source studies: 
 
REVISTA DE FILOLOGÍA Y LINGÜÍSTICA94
Esta investigación informatizada podrá parecer simplemente un retorno por otros medios a la
clásica filología universitaria positivista de métodos infinitamente más precarios. La
manuscriptología no será entonces sino una variante más sofisticada de la antigua investigación
de las fuentes. Los resultados y la dirección que han tomado estos estudios son sin embargo
completamente diferentes. En primer lugar, estamos lejos aquí del concepto del texto
doctoralmente establecido, que ellos consideran un “castillo tambaleante” Este “nuevo objeto”
literario, esta “escritura viva” renacida de debajo de la lápida de las tachaduras, surge en el polo
opuesto del inmutable texto canónico, revolucionariamente y es revolucionario en la medida en
que lo es todo cambio libertador de las rigideces dogmáticas de las definiciones de escuela (292). 
Ana María Barrenechea adds that thetraditional philological work aimed at
establishing a fixed version of the text supported by the author or editor is not really genetic
criticism but rather textology (Barrenechea 13).
In the National Center for Cientific Research in Paris (CNRS), the Institute of
Modern Texts and Manuscripts (ITEM) has been functioning for close to three decades (Roa
Bastos. “La escritura secreta de las tachaduras”: 289). This group dedicates itself exclusively
to genetic study.
The first question raised by these genetic theorists is that of what materials constitute
their legitimate object of study. There is a definite lack of consensus among critics as to what
should be considered valid material in the analysis of the development of a text in its “final”
published form. One of the leading genetic theorists, Jean Bellemin-Noel, uses the term
“avant-text” (pre-text) to refer to these materials and defines them in a broad sense as
“l’ensemble constitué par les brouillons, les manuscrits, les épreuves, les “variantes” vu sous
l’angle de ce qui précede matériellement un ouvrage quand celui-ci est trait‚ comme un texte”
(Avant-texte, texte, apres-texte 15) (“the ensemble constituted by the drafts, manuscripts,
proof sheets, the “variants” seen from the angle of that which materially preceeds a work when
it is treated as a text”). Roa Bastos, known as Paraguay’s foremost novelist and a constant
reworker of his own fiction (lending ample material for genetic study) ratherthan a genetic
theorist, also adopts the broad view, stating that genetic research 
concierne al entero proceso de elaboración de la obra: desde los esbozos inciales a los
manuscritos definitivos; desde los cuadernos de apuntes, los diarios íntimos, la correspondencia
de los autores, sus declaraciones públicas y privadas, a las sucesivas ediciones de la obra
hechas en vida del autor, así como toda otra suerte de documentos relacionados con la obra en
estudio, por aparentemente insignificantes que sean (289).
However, others, such as Jacques Petit, feel that the field of study should be somewhat
limited. He states that one should distinguish between drafts and outlines because authors
never intend to publish their outlines, which simply serve as a means of preparation and thus
cannot be placed on the same level as rough drafts which may be changed our disgarded. Petit
further suggests that some distinction should be made between written drafts and unpublished
printed drafts of a text, the latter being obviously closer to the text in its so-called “final” form.
Finally, Petit also indicates that the genetic theorist should distinguish between “publishable”
texts and “non-publishable texts.” “Non-publishable texts” include unfinished sentences and
developments. 
Petit’s ideas would essentially eliminate some major genetic studies. For example, in
1983, Ana María Barrenechea published Cuaderno de bitácora de “Rayuela”, which is a
WELDT-BASSON: Roa Bastos’s “Noche sin fin” and “La tijera”:  Two Versions of... 95
genetic analysis of the logbook that Cortázar used to write his most famous novel. Much of the
logbook consists of notes and outlines, which according to Petit, should not be included in
genetic criticism.
Other critics, notably Paul Vasak, rather than setting limits on what is to be studied,
attempt to establish an order in the development of genetic materials. Vasak defines three
stages of textual “communication”: The autocommunicative stage in which the text is born
and destined only to be read by the author, the non-public phase in which the text is finished
but unpublished, destined to be read by editors and publishers, and finally the stage of public
communication, in which the text enters the communicative system of society. Thus, stages
one and two, precede stage three as potential genetic materials.
The second major question posed in the study of genetic theory is that of how the
genetic analysis should be carried out. There are many different approaches, each supporting
itself in a different contemporary theory, and none precluding any of the others. For example,
Raymonde Debray-Genette, another major figure in contemporary genetic theory, emphasizes
the importance of studying not just isolated changes from pre-text to text, but rather patterns
of variations (groups of changes that correspond to certain principles). Debray-Genette also
suggests that instead of studying manuscripts as a process of evolution (from pre-text to
“final” text), it might be more profitable to view the relationship between pre-text and text as
one of difference. This approach grants a more autonomous functioning to the pre-texts which
are thus not necessarily posited as relying on the “finished” text. Finally, Debray-Genette
suggests that changes or differences between pre-text and text may be either paradigmatic (a
selection or substitution of one element for another) or syntagmatic (variation in the
combination of elements from pre-text to text).
Debray-Genette’s genetic model clearly differs from the one suggested by A.
Gresillon and J.L. Lebrave in “Manuscrits, Linguistique et informatique.” Here the authors
propose a homogeneous mode of representation applicable to all types of variants based on a
linguistic model which allows all variants between versions of a text to be explained in terms
of substitution. In other words, an addition to the text would be described as the substitution of
a non-void unit for zero, and the suppression or omission of an element would be described as
the substition of zero for a non-void unit. This model does not differ significantly from
Debray-Genette’s idea of paradigmatic difference, but does not appear to account for what she
terms syntagmatic difference (the difference in the combination of elements).
Both Debray-Genette and Gresillon and Lebrave focus on a microtextual analysis of
pre-texts and texts. Other critics, such as Bernard Brun, address issues concerning a
macrotextual analysis. For example, in “Problemes d’une édition génétique: L’atelier de
Marcel Proust”, Brun speaks of the necessity to describe and distinguish between different
types of textual fragments (units separated by blanks). He defines two types: those which it is
impossible to classify, and those which entertain relationships with other units (such as
subordination to another unit or association with it). These units are, in order of increasing
size, a segment, a piece, a section, or a part of a work.
Bellemin-Noel suggests a parallel between genetic theory and psychoanalysis. He
views pre-texts of a text as a text’s “other” rather than its origin. Bellemin-Noel explains that
Freud defined the desire to be the other as the “subconscious.” Thus, the pre-text becomes the
text’s “subconscious” allowing us to find words that have been surpressed and are nowhere to
 
REVISTA DE FILOLOGÍA Y LINGÜÍSTICA96
be found in the text itself, helping us to find the missing link to understanding the text, just as
Freud spoke of repressed painful memories from a person’s youth. This psyochoanalytic
interpretation is very astutely applied in Bernardo Teuber’s genetic study of Lezama Lima’s
Paradiso, where the author shows how Lezama’s unconscious lapses result in variants in
different editions of his work.
The third question is what is the purpose or goal of genetic studies. Debray-Genette
concludes that the major end sought is to “disprove, discover, confirm” interpretations of a
work (38). This procedure can be observed, for example, in Claudine Quémar’s study of
Proust. Quémar analyzes the pre-texts of a section of Proust’s novel A la recherche du temps
perdu that disproves the hypothesis that the onomastic reverie sequence is based on
synesthesia (“colored hearing” or the association of sounds with colors). Quémar goes on to
show through references to the novel’s pre-texts, that this sequence is based on certain
phonetic associations between the words used to describe the place in question and the name
of the place itself (Quémar 81).
Bellemin-Noel points out that genetic studies give a voice to what is excluded in the
text; “le refusé, le réprimé, le refoul‚ acquierent statut de signes, conférent sens du meme coup
a ce qui les remplace (fut-ce une absence) a ce qui les supplante” (18). In other words,
omissions and changes which do not appear in the “final text”, acquire importance with the
regard to this text in genetic studies. This is a unique feature of this type of literary criticism.
A fourth major issue discussed in genetic criticism is that of the need to clarify and unify
the terminology used. Bellemin-Noel proposes a strict terminology in Le texte et l’avant texte
that distinguishes between rough copies, manuscripts, outlines, sketches, and proof sheets.
According to Bellemin-Noel, “brouillons” (rough drafts) refer to all the writings that lead to the
published work; these may be 1. an “ebauche” (“outline”), which is part or all of a rough draft
that is subsequently transformed, annulled or completing by a following stage; it can never be
treated as a finished work; 2. “esquisse” (“sketch”), defined as a typically incomplete stage
which the writer at one point sees as finished, but not publishable, and which the reader can
appreciate as such; 3. “épreuves” (“proof sheets”), which is the case of a work which once
printed, the writer treats not as a finished work, but as a new rough draft; 4. “manuscrit”
(“manuscript”), the stage upon which the printing of a work is based; 5. “ouvrage” (“work”), a
particular piece of writing published under the name of a writer; 6. “variante”, a difference exists
when comparing the manuscript to the work or between various editions of a work; 7. “oeuvre”
(“works”), the entire production of a particular author. Bellemin-Noel also distinguishes between
“biffure” (literally, “blot”) and “rature” (“erasure”): The first surpresses a segment of the text,
whereas the second is a surpression of an element of the text followed by a new formulation of it
(13-18). 
The most striking common thread among many of these genetic theories is the close
relationship between genesis and intertextuality. Debray-Genette, Henri Mitterand, and
Richard K. Curry all link the concept of genesis to that of intertextuality in one form or
another. Debray-Genette notes in her study of Flaubert that he copies from other texts in his
manuscripts. She states: 
Commence le puzzle de l’intertextualité. Ce n’est pas un phénomene purement stylistique: un
génétique complete doit s’appuyer sur une poétique de l’intertextualité. Comme l’indique
Laurent Jenny ‘si tout texte réfere implicitmenet aux textes, c’est d’abord d’un point de vue
WELDT-BASSON: Roa Bastos’s “Noche sin fin” and “La tijera”:  Two Versions of... 97
génétique que l’oeuvre litteraire a partie liée avec l’intertextualité’ (32). 
(the puzzle of intertextuality begins. It is not a purely stylistic phenomen: a complete genesis
should base itself on a poetics of intertextuality. As Laurent Jenny indicates, ‘if all texts refer
implicitly to texts, it is first of all from a genetic viewpoint that the literary work started out
linked with intertextuality’).
Similarly, Mitterand, in his study of one of Zola’s works, evolves a very interesting
theory of how Zola’s outline for his book l “Assommoir was affected by what Zola had read
in various documents. Mitterand proposes that there is a “competence of contents”
(acquired knowledge, received ideas, stereotypes, presuppositions, etc.) that includes both
the author’s explicit models and intentions but also his implicit ones. Thus, there is
documentary knowledge in Zola’s novel which is intentionally elaborated and other
documentary knowledge which is unconscious. Mitterand goes on to show how Zola’s
notion of “the people” was implicitly taken from certain publications of his era, including
Larousse’s Gran Dictionnaire. Mitterand refers to the influence of these documents as the
“discursive model” of the text and goes on to show that there also exists a “narrative
model” (the technical and rhetorical inheritance of the author). Mitterand concurs with
Debray-Genette that it is important to establish the global models of the genesis of the text,
and not just study a series of details.
Augusto Roa Bastos, born in Paraguay in 1917, is one of his nation’s foremost
writers. Winner of the prestigious Cervantes Prize in 1989, Roa Bastos is the author of
numerous collections of short stories including El trueno entre las hojas (1958), El baldío
(1966), Los pies sobre el agua (1967), Madera quemada (1967), and Moriencia (1969). He
has written six novels: Hijo de hombre (1960), Yo el Supremo (1974), Vigilia del almirante
(1992), El fiscal (1993), Contravida (1994), and Madama Sui (1996). 
As is the case with many contemporary writers, Roa Bastos frequently publishes
fragments of his works in progress in literary journals prior to their publication as books. The
author subsequently reworks these fragments in his “final” version of the text. Sometimes the
“final” version isn’t really a “final” version at all, as in the case of Roa Bastos’s novel Hijo de
hombre. The novel originally published in 1960 was rewritten and republished by Roa Bastos
some 23 years later, in 1983. As we have already noted, these prior publications and previous
versions then become “pre-texts” for subsequent versions, and the differences between the
two, provide “variants” whose signficance requires further study. 
This study focuses on the manner in which Roa Bastos has transformed the short story
“Noche sin fin,” originally published in the journal Ficción (no. 10, Buenos Aires, 1957) into
his story “La tijera,” later published in his collection El baldío (1966). 
All genetic studies involve three types of changes between pre-text and text: “blots”
and erasures (omissions), which Bellemin-Noel terms “biffures” (Le texte et l’avant texte: 13);
changes from one element to another, or an erasure replaced by a new formulation, which
Bellemin-Noel terms “rature,” and simple additions of new material (Le texte et l’avant texte:
14). Lebrave and Greillons would refer to all of these as substitutions of void and non-void
units (“Manuscrits, linguistique et informatique”: 180-1). These paradigmatic changes may
also be accompanied by syntagmatic changes: placement of elements in different spaces or
new combinations. As Debray-Genette notes, these differences between pre-text and text
frequently follow certain patterns (“Génétique et poétique: 34). Thus, categories of changes
 
REVISTA DE FILOLOGÍA Y LINGÜÍSTICA98
can be identified that show how and why the author has effected these changes. In turn, this
analysis of patterns or categories provides insight into the writing process, the author’s intent
and goals, and/or heretofore hidden meanings in the text itself.
“Noche sin fin” and its “final” version, “La tijera” are remarkably similar, yet exhibit
some important differences which show how Roa Bastos’s vision of his characters changed
between the writing of the two versions. “Noche sin fin” is essentially the story of a woman
who has gone mad, whereas “La tijera” is the story of a fight for control between two women,
a young girl and her aunt, and the former’ revenge on the latter. A meticulous analysis of the
two versions of the story reveals five different categories of changes: 1. changes in character
development; 2. alterations in plot line; 3. variations for subtlety; 4. changes that eliminate the
hyperbolic or melodramatic overtones of the story; and 5. alterations that create irony/parody. 
“Noche sin fin”/”La tijera” is the story of three prudish sisters, Elsa, Amanda and Elvira,
who live with their niece Lía, the daughter of a fourth sister who is deceased. Elsa, the most
domineering of the three, tries to keep Lía very sheltered and under her control. However, the
sisters themselves delight in reading the shocking headlines in the newspaper about different
atrocities. One day they read about the murder of a financier involved in the corruption of
minors, presumably stabbed to death by one of the young girls he seduced. They believe Lía
may have overheard their discussion of the crime. Lía begins to act strangely and finally
confesses to Elsa that she is guilty of the murder. Later on, it appears that Lía is pregnant.
Amanda also reveals that she found Lía’s scissors with blood stains on them. Elsa, terrified that
Lía will be linked to the crime, calls the police and confesses to the crime herself. The end of the
story reveals that Lía is not really pregnant, but simply stuffed pieces of a quilt under her
clothing. 
Both versions of the story imply that Lía invented her involvement in the crime,
although, as we shall see, this is more explicitly stated in the second version. However, as a
genetic analysis reveals, the reasons why Lía did this, and hence, her characterization, is quite
different in each version of the story.
There are four main characters in the story: Elsa, Amanda, Elvira (the three sisters)
and Lía (their niece). Lía undergoes the greatest number of transformations from pre-text to
text. In “Noche sin fin” Lía is portrayed as weak, innocent, and mentally unstable, whereas in
“La tijera” the character appears stronger, more calculating, vengeful, and perfectly able to
understand the magnitude of her actions. The character analysis that emerges from comparing
the variants of the two editions emphasizes elements whose importance might otherwise go
unperceived by the reader. For example, the portrayal of Lía in “La tijera” as a strong,
independent woman is highlighted by two “blots”. The first is “La tijera”’s omission of the
following explanation found in “Noche sin fin” after Elsa calls Lía “mi monita”: 
“la palabra “monita” en labios de Elsa, siempre tendía un lazo. Pero esta vez su sagacidad
orillaba un abismo. Repitió el conjuro: ¿Miedo de qué, monita?” (36). 
This passage shows that Lía has a weak point with regard to her aunt; she usually
softens and complies when her aunt calls her “monita.” The second version of the story
chooses to omit this detail because it makes Lía more vulnerable. Similarly, in another scene
where Lía is not feeling well “La tijera” omits that “Lía (que) se desplomó entre sus brazos”
WELDT-BASSON: Roa Bastos’s “Noche sin fin” and “La tijera”:  Two Versions of... 99
(33) (referring to Elsa). Once again, this shows Lía to be more vulnerable and dependent,
which does not fit with the new characterization designed for her in “La tijera.” In another
instance, “La tijera”, when speaking about Lía’s “youthful limbo,” omits the description of this
limbo as “puro y hermético.” The notion of innocence and purity would detract from the
strong, calculating image of Lía that the second version projects. Finally, in yet another
instance, when Elsa is fussing over where Lía has been, the original has Elsa say: “Pero
todavía eres una chicuela, Lía” (36), whereas in the second version this becomes: “Pero yo lo
hago por tu bien, Lía” (83). The first version once again conveys the notion of youth and
innocence, which the second strives to eliminate.
A second group of changes relating to Lía deal with a change in her motivation for
pretending to commit the murder which her aunts read about in the newspaper. This is perhaps
the greatest difference between the two versions of the story. In “Noche” Lía mentions the
influence of some past events upon her actions. She claims that Tío paco Velarde, a boy, and a
sparrow guided her actions. Everything written between “Tío Paco Velarde” (39) and “no se
sabía si había querido consolar al chico” (41) is omitted in “La tijera”:
—Tío Paco Velarde . . . , aquel gorrión del bosque de Palermo . . ., aquel chico rubio cuyo
barquito había naufragado en el lago, guiaron mi mano . . .
(...) 
A los siete años Lía ayudó a don Paco Velarde en su teatrito de títeres del barrio... 
Lía nunca pudo olvidar las funciones de títeres. Tampoco la muerte del viejito titiritero con
quien ella afirmaba que había conversado la noche del velorio. 
(...) 
Después, cuando recuperó la movilidad de la pierna, en un paseo por el bosque de Palermo, fue
aquel gorrión que se lanzó de un árbol como una piedra delante de los dientes de un perro que
se había abalanzado contra el pichón caído del nido. El perro retrocedió con miedo ante el
gorrión erizado y aleteando en el suelo. Lía, con los ojos secos, gritó, azuzando al perro:
—¡Mordélo, mordélo, si te animás! Ese pobre pajarito es más valiente que vos!
(...)
El chico, con los brazos cruzados, de rodillas al borde del agua, miró hundirse su embarcación
sin un gesto, sin una lágrima siquiera. Lía se escapó de la mano de Elsa, se aproximó corriendo
al chico, le tocó el hombro y le dijo cuando, él se dió vuelta para mirarla:
—No te pongas triste. Ha bajado a la ciudad de los muertos, donde algún niño parecido a vos
podrá jugar con él...
En eso vió en el agua la imagen del chico y comprendió que allí estaba el otro, un chico oscuro
y fantasmal que, de rodillas bajo el agua, la miraba irritado. Tendió la mano hacia el lago.
—¿No ves? ¡Allí está! Ahora el barquito es de él.
—¡Idiota!—le gritó el chico vivo, y se fue arrastrando los pies. El otro también había
desaparecido bajo el agua.
Pero, de todos modos, Lía se había quedado contenta y orgullosa. No se sabía si había querido
consolar al chico o burlarse de él (39-41).
It is principally this section which leads us to question Lía’s sanity. She believes that a
dead man spoke to her and acts strangely and cruelly in the episodes of the bird and the boat.
“La tijera,” omits this section because the story creates a Lía who is cold and vengeful, and
who triumphs over her aunt’s controlling nature. Thus, any hint that she might be mad is
eliminated in the second version.
There are numerous other alterations that seek to modify the relationship between Elsa
REVISTA DE FILOLOGÍA Y LINGÜÍSTICA100
and Lía. For example, “La tijera” omits from “Noche” this part of a dialogue between the two
characters which conveys a feeling of tenderness between them: “—Tía Elsa—Sí, querida”
(35) followed later by Lía’s “—Qué buena por haber venido! Te esperaba” (36). This type of
communication is more credible in the first version where we are never sure if Lía really
committed the crime or is insane or not, and thus cannot necessarily be accused of despising her
aunt and seeking revenge for her dominance. In contrast, the second version adds elements to
develop this battle of wills between Lía and Elsa. “La tijera” tells us that Lía “ponía ese secreto
entre ella y su voluntad de dominio” (35). We can see that Lía resents Elsa’s attempts to control her.
“La tijera” also omits, adds, and changes comments to emphasize Amanda’s perverse
nature and the antagonism between her and Elsa. For example, in “Noche” we are told that for
Amanda, the reading of the crime page was a “juego mórbido de especial atractivo” (32). In
“La tijera” this becomes a “juego mórbido de la que sólo ella podía disfrutar” (76). This makes
Amanda seem more uniquely perverse in her enjoyment of the sordid details of the crimes, and
thus more negative than in the first version. 
A plot change that also impinges upon characterization in “La tijera” is when Amanda
finds the scissors that Lía allegedly used in the murder and wipes them clean, whereas in
“Noche” it is Elsa herself who finds the scissors. This is an important change, because
Amanda then uses this revelation to torture Elsa and drive her to ultimately confess to the
crime herself. “La tijera” adds the following:
—Aquí está —sacó del cajón de su máquina, de entre el revoltijo de puntillas y retazos, un
envoltorio de papel metido en una media de mujer. Lo desenvolvió, sacó la tijera y se la tendió.
Los ojos de Elsa se desorbitaron al fijarse en la costra rojiza que empañaba una hoja. —Yo revisé
esa noche... y estaba limpia... balbuceó con una voz desconocida. —Porque yo la cambié‚ un
momento antes de que la vieras vos, para evitar un catástrofe. Después la escondió. Pero ahora la
mancha está en otra parte, y ésa ya no se puede lavar y está creciendo. Hay esos análisis de grupos
sanguíneos, y todo lo demás a menos que la hagamos desaparecer con la ayuda de uno de esos
médicos que se dedican a la... cirugía estética de la natalidad descarriada. —¡Cállate maldita!
The second largest category of changes is that of plot alterations. There are several
important differences in plot development between “Noche sin fin” and “La tijera.” The first,
is “La tijera’s” omission of the history of Lía’s mother, Mabel. In “La tijera” there is only one
mention of Mabel. Amanda states “Estamos peor que Mabel, que por lo menos tuvo lo suyo”
(“La tijera,” 80) and Elvira’s response: —¡Pobre Mabel! Harías bien en no meterla a ella en
estas cosas. No te das cuenta de que Elsa quiere evitarle su suerte a la hija?” (80). “La tijera”
systematically eliminates all other references to Mabel, which include the following: “[el
drama pasional] que quince años atrás se había tragado a Mabel, la madre de Lía” (31) and, 
Pero detrás estaba también aquello que era preferible olvidar. La fuga de Mabel, la hermana
menor, su caída en el torbellino de destinos a la deriva, sus últimos tiempos de bailarina en uno
de los dancings del bajo. Por fin el nacimiento de Lía, que costó la vida de la madre, en la
maternidad del Rawson, y la lucha a brazo partido de Elsa para rescatar esos despojos vivientes
de la hermana, con quien se había reconcilaido al filo mismo de su agonía. “Por amor de Dios,
Elsa, sálvala” Fue un compromiso sellado entre la vida y la muerte, que ataba a Elsa para
siempre al destino de la criatura doblemente huérfana (“Noche”: 39). 
“La tijera” prefers to be more subtle in its plot development; it suggests the mother’s
 
WELDT-BASSON: Roa Bastos’s “Noche sin fin” and “La tijera”:  Two Versions of... 101
background without actually stating it. “Noche sin fin” is more direct and gives us the
specifics of Mabel’s past.
Another change in the two versions is with regard to the role of the Ibáñez family. In
“Noche” they live upstairs, on the fifth floor, in the same apartment building as Lía and her
three aunts, and Lía often visits them. In “La tijera,” Lía also goes over to the Ibáñez’s house,
however, they no longer live upstairs. Moreover, there is an important addition to the
household in “La tijera,” where we are told that a boarder lives in the Ibáñez’s house. This
boarder is a Paraguayan exile who turns out to be Lía’s lover at the end of the story. This ironic
plot twist (the supposedly innocent Lía has a lover) occasions a number of changes and
additions. The first set, as already mentioned, are those surrounding the location of the
Ibáñez’s house.”La tijera” omits that their house is “en el quinto” we are told in “Noche” that
Lía went up to “al departamento de los Ibáñez” (31) whereas in “La tijera” “iba al
departamento”; in another instance we are told that no one heard Lía “bajar” (32) which
becomes “llegar” (76);”Noche” speaks of Lía’s games “en el quinto” (34) whereas “la tijera”
speaks of games “en lo de Ibáñez” (80); when Lía confesses in “Noche” she says “no subía al
quinto” whereas in “La tijera” she states “no fui a casa de los Ibáñez” and finally, “La tijera”
changes “Lía continuaba subiendo al quinto” (“Noche” 35) to: 
Lía continuaba con sus clases particulares, con sus paseos por Palermo, los domingos, del brazo
de tía Elsa. Sólo al departamento de los Ibáñez no volvió a ir; Elsa encontró la forma de
suprimir estas visitas sin herir susceptibilidades, o quizás también en la vivienda de los Ibáñez
habría surgido alguna alteración en la rutina con la presencia del refugiado—los chicos del
matrimonio subían a otro departamento a jugar, por las tardes, o salían a la plaza; de tal modo,
la ausencia de Lía pasó inadvertida (“La tijera”: 81).
Finally, the two versions are completely different with regard to what happens
when Lía first returns from the Ibáñez’s house and speaks with her aunts. In “Noche” she
returns complaining that she doesn’t feel well; whereas in “La tijera” the following section
is added where she reveals to her aunts that she met the Paraguayan boarder who lives with
the Ibáñez family:
...los Ibáñez no estaban. Fueron a un cumpleaños, creo, con todos los chicos. —por qué no
volviste entonces—Me quedé hablando con ese muchacho que vive con ellos.—¡Quién!... la
alarma de Elsa se transformó pero no por eso decreció sino que aumentó. —Ese desterrado
paraguayo que llegó a la semana pasada. —Cómo puede ser... no me han dicho nada... —los
labios de Elsa se afinaron más aún hasta desaparecer en un tajo agrietado en los bordes. —No
tienen por qué contarte todas sus cosas—dijo Amanda—. Como paraguayos radicados aquí, es
natural que alberguen a un compatriota en desgracia. Elsa no pareció oírla. —Debiste venir en
seguida—conminó a Lía sin mirarla a los ojos—Nunca hablo con nadie—se quejó la
muchacha.—No tenía nada de malo que me quedara un rato a charlar con él. Me contó las
cosas terribles que están pasando en su país. Me dio mucha lástima, ¿sabes? Se escapó por
milagro del periódico donde trabajaba cuando lo asaltaron e incendiaron. Tienen que
conocerlo... Lía miró a sus tías... (“La tijera”: 78-9). 
Thus, Roa Bastos adds both a political dimension to the story as well as irony to Elsa’s
attempts to shelter her niece and Lía’s supposed innocence. 
There is at least one change that is designed to prefigure or coincide with Lía’s
REVISTA DE FILOLOGÍA Y LINGÜÍSTICA102
affair with the Paraguayan. Instead of “Elsa hizo venir a la tarde a la profesora particular”
(“Noche” 41), “La tijera” states that Lía “continuaba yendo por las tardes a la casa de la
profesora particular” (“La tijera” 87), perhaps because the idea of L¡a leaving the house,
having freedom to move about without her aunt’s constant vigilance, makes plausible her
association with the Paraguayan, which would otherwise never be possible if she were not
allowed to leave the house.
Finally, the relationship between Lía and the Paraguayan is revealed in “La tijera” in a
newly developed ending. The first version has the following ending, omitted in “La tijera”:
Después se levantó [Elsa] despacio y se acercó al teléfono. Discó lentamente pero con firmeza,
con decisión oscura y desolada, como si entre número y número surgieran ya distancias
irreparables.
—Hola... ¿Departamento de Policía?... Soy la que mató a Manuel Alzogaray... He decidido
entregarme... No... No hay necesidad de que vengan a buscarme... Iré yo misma... Me
presentaré allí esta noche... Sí... Sin falta
...
El tubo golpeó en la horquilla con un sonido blando y opaco. El silencio del cuarto de costura
se fué fundiendo con la oscuridad. La noche había caído por completo. No encendieron la luz.
El sordo rumor de la ciudad batía contra los vidrios y los freía tenuemente en las imperceptibles
ondas de su maelstrom. Elsa, Amanda y Elvira estaban hermanadas por fin en su silencio, en las
tinieblas de esa noche que de pronto se hacía desmesuradamente larga, sin límites, sin piedad
(“Noche” 43).
“La tijera” eliminates the presentation of Elsa’s confession, and simply alludes to her
as having gone insane and insisting she committed the financier’s murder. Instead, “La tijera”
adds a new conclusion, in which Lía and the Paraguayan dialogue in bed:
—¿Qué pensás hacer?
—Buscarme otra casa, conseguirme algún trabajo, y seguir arrimando el hombro a la causa. No
puedo continuar viviendo de las limosnas de las cotizaciones y las rifas de los comités de
ayuda.
—¿Por qué no te venís a casa?
—¿Y las purísimas vírgenes de tus tías? No me van a tragar.
—Ahora que Lobo Feroz no está , todo es más fácil. ¿Sabés que la internaron hoy en el
pabellón de alienadas por orden del juez? La pobre está cada vez peor. Ya no nos reconoce. Sólo
cuando la ve a Amanda se pone furiosa y echa espuma por la boca. Continúa emperrada en
cargar con el asesinato del rentista.
—Hasta ahora no entiendo por qué se te ocurrió la tilinguería ésa de hacerle creer el bolazo del
crimen y hasta de que estabas embarazada.
—No sé... qué sé yo. Cuando me di cuenta ya lo había hecho. Además, me divierte andar con el
acolchado ése. Fíjate que esa noche que manché la tijera con tinta roja me entró un escalofrío y
hasta soñé‚ que había matado realmente al rentista. Amanda me sigue jorobando todo el santo
día para llevarme al partero... 
Se rieron los dos y entrelazaron sus dedos. Con el índice y el mayor de la otra, él atenaceó la
garganta y simuló que se la iba a cortar a tijeretazos, aumentando a cada tajo el borbotón de esa
risa que parecía llegar de un parque con árboles, palomas y niños jugando al atardecer.
The percentage of changes effected in the second version are small (in comparison to
what remains the same), and yet with these changes Roa Bastos has succeeded in creating a
WELDT-BASSON: Roa Bastos’s “Noche sin fin” and “La tijera”:  Two Versions of... 103
totally different story. These changes support the interpretation of Lía’s actions in the story as
those of someone who seeks freedom and revenge, in a calculated fashion, and not someone
who acts on impulse, as Lía claims, doing all of this without really knowing why. The changes
between versions accentuate these elements of perversity and vengefulness in both the
character and the plot.
The third category of changes relates more to style than substance: changes aimed at
creating a more subtle text, with less direct statement of events and less clarification, yet more
indirect allusions or “clues” to the story’s ending. Firstly, “La tijera” omits many of the
narrative indicators that are present in “Noche sin fin,” such as “dijo el diarero” (30) and
“siguió leyendo Elsa” (32). In other instances, “La tijera” omits statements that directly
express feelings, motives, etc., which the author then chooses to have the reader deduce more
indirectly (and thus increases reader participation) in the second version. For example, “La
tijera” omits from “Noche” when Lía is confessing her supposed crime to her aunt: “No era
eso lo que importaba . . . Era esto [the fact that she was involving her aunt] que le producía una
letal satisfacción” (“Noche” 37). In other words, the second version omits the explicit
emphasis on Lía attempt to involve her aunt in her crime. This is a specific allusion to Lía’s
motivation, which this second version prefers to reveal only at the end, thus increasing the
effect of suspense and surprise. Similarly, when Lía finishes her confession, in “Noche” we are
told “su voz era suave, casi pérfida. Pero Elsa no estaba en situación de notarlo” (“Noche” 39).
This is a reference to the false nature of Lía’s confession and her aunt’s inability to grasp it.
“La tijera” does provide two additional clues to the story’s ending, but these are both very
subtle and indirect. In “Noche” we are told that Lía’s description of her crime coincides in
“líneas generales” (39) with the version in the newspaper article; in “La tijera” (86) they
coincide “casi literalmente.” This is not an addition of a clue, but it strengthens the clue by
emphasizing more strongly the relationship between Lía’s confession and the crime as
reported by the media. The second clue is the addition of a paragraph in which Elsa finds some
poems among Lía’s things. This clue clearly prefigures her relationship with the Paraguayan,
but there is no way we can know it at this particular point in the story. We are told:
Entre las páginas de un libro hizo una noche un hallazgo que no la compensó mayormente de
sus afanes: un trozo de papel con un poema amatorio. Le resultó desconocida la letra, y Elsa no
podía saber tampoco que versos enteros habían sido copiados de Neruda, de Lorca o de Miguel
Hernández, versos en los que se aludía a la mujer y al amor con imágenes libres y procaces,
verdaderamente libertinas, como se lo dijo a la profesora, que tampoco sabía nada del asunto.
Pero preocupada por la búsqueda del cuaderno fantasma, Elsa no le concedió mucha
importancia y acabó olvidando el poema clandestino (87). 
“La tijera” aims to make its story less cliché and melodramatic. The second version
systematically eliminates elements that might seem “affected” such as when Lía says she
stabbed “el costado del monstruo” (“Noche” 38) which becomes simply “el costado” in “La
tijera” (85). Also, “La tijera” omits when speaking of the scissors that Lía used to commit her
crime, the fact that she doesn’t remember retiring the scissors from the “cuerpo inmóvil del
monstruo” (just “cuerpo” in “la tijera”) and that Lía says that she doesn’t dare to open her
pocketbook to see if the scissors are there “ni creo que me atreva jamás. La tijera con la que
cortaba los trajes para mi muñeca” (“Noche” 38). “La tijera” also omits this paragraph about
REVISTA DE FILOLOGÍA Y LINGÜÍSTICA104
the neon sign that haunts Lía at night: 
Trataba por todos medios de volver a soldar las rajaduras de la campana neumática en torno a la
sobrina. [Colocó una cortina]... contra las ráfagas del letrero luminoso. Lástima que no podía
hacer lo mismo contra la refluencia de esos otros destellos maléficos que todavía en tanto en
tanto le mordían a ella con dolor latente en la caries del alma. Así pasaron cuatro meses.
(“Noche”: 41). 
Clearly, phrases like “la tijera con la que cortaba los trajes de mi muñeca” and “le
mordían a ella... en la caries del alma” sound like melodramatic clichés that Roa desires to
eliminate from his work. This gives a more credible, measured tone to the second version of
the story. At one point, in an addition to the second version, the narrator speaks of a scene as
being cliché, in an explicit commentary that shows the author’s consciousness of this problem.
When Lía is cuddled in Elsa’s armas, we are told it is “una perfecta escena de película cursi”
(“La tijera”: 83).
The final category is constituted by additions that create irony and parody. There are at
least two additions to “La tijera” that accentuate the story’s ironic dimension. The first is when
Elsa states “Lo que se llama el crimen perfecto no existe—dijo Elsa con suficiencia. —Debe
existir pero no se conoce. Cuántos criminales no andan por ahí vivitos y coleando, riéndose de
la policía y de todo—murmuró Amanda” (“La tijera”: 78). This statement turns out to be
ironic, because if the police believe Elsa’s confession (which is not clear from the story—is
she in the insane asylum for having committed the crime or for thinking she is guilty of a
crime she didn’t commit?) then the person who really murdered the financier has committed
the perfect crime. The second ironic addition is when Lía tells her aunts: “no fui a casa de los
Ibáñez. —Dijiste que te habías quedado charlando con ese muchacho refugiado. Lo alabaste
bastante. —Todo eso fue mentira. La verdad es que a mí ese muchacho me resulta antipático
de sólo verlo. La tarde que sucedió aquello, fui... Lía se detuvo de nuevo” (“La tijera”: 83). Of
course, this is ironic because Lía is really having a love affair with the Paraguayan.
There are two references to parody in “La tijera” which were not present in “Noche
sin fin.” The first is when we are told that the flashing of the neon sign was “en una parodia de
las escenas de película con temas de misterio” (“La tijera”: 82); the second, when we are told
that Elsa was “una figura paródica más en esa decoración que no se podría describir de otra
manera” (“La tijera”: 82). In other words, Roa Bastos emphasizes the similarity between this
scene and the movies. It is another way of referring to the story’s melodramatic element,
without the story actually appearing to be melodramatic itself. 
Upon comparing these two versions of a single story, we can see how Roa Bastos
changed his conception of his protagonist, Lía, from possibly insane to most certainly
vengeful; how he introduced new plot elements, including a political and amorous dimension
in the figure of the Paraguayan exile, and how he perfected his narrative style with a more
subtle, concise, and less melodramatic prose. Roa Bastos has created a “new” story with the
“secret writing of erasures.” He has redirected the plot and characterization in ways that can
only be revealed, as we have just seen, through genetic textual analysis. 
WELDT-BASSON: Roa Bastos’s “Noche sin fin” and “La tijera”:  Two Versions of... 105
Bibliography
Barrenechea, Ana María. 1983. Cuaderno de bitácora de “Rayuela”—Julio Cortázar. Buenos
Aires: Editorial Sudamericana.
Bellemin-Noel, Jean. 1972. Le texte et l’avant-texte: les brouillons d’un poéme de Milosz.
Paris: Librarie Larousse.
1979. “Lecture psychanalytique d’un brouillon de poeme “été” de Valéry,” Essais de
critique génétique. Ed. Louis Hay. Flammarion: 103-49.
Curry, Richard K. 1984. “El contexto genético en literatura.” Texto Crítico 10. 29: 136-47.
Debray-Genette, Raymonde. 1977. “Génétique et poétique: esquisse de méthode”. Littérature. 28: 19-
39.
Grésillon, A. et J-L Lebrave. 1982. “Manuscrits, linguistique et informatique,” En: Hay, Louis et Péter
Nagy: 177-87.
Guerin, Wilfred L. 1966. A Handbook of Critical Approaches to Literature. New York: Harper and Row.
Hay, Louis et Péter Nagy (Eds). 1982. Avant-texte, texte, après texte. Budapest: Akademiai Kiadó.
Jenny, Laurent. 1982. “The Strategy of Form.” En Tzvetan Todorov.  34-63.
Mitterand, Henri. 1979. “Programme et preconstruit génétiques: le dossier de “L’assommoir.” Essais de
critique génétique. Ed. Louis Hay. Flammarion: 193-226.
Petit, Jacques. 1982. “Pour une typologie des manuscrits littéraires.” En Hay, Louis et Péter Nagy.  191-
4. 
Quémar, Claudine. 1979. “Reveries onomastiques proustiennes a la lumiere des avant-textes.” Essais de
critique génétique. Ed. Louis Hay. Flammarion: 69-102.
Roa Bastos, Augusto. 1990. “La escritura secreta de las tachaduras,” Escritura, 15.30 (julio-diciembre):
285-93.
1957. “Noche sin fin,” Ficción 10: 30-43.
1976. “La tijera.” El baldío, 2nd ed. Buenos Aires, Losada, 
Teuber, Bernardo. 1993. “O felix lapsus! Autobiografía, crítica genética y geneología del sujeto en
Paradiso de Lezama Lima.” MLN, 108: 314-30.
Todorov, Tzvetan. 1982.  French Literary Theory Today. Cambridge: Cambridge University Press.
Vasak, Paul. 1982. “Analyse de systeme et textologie”. En Louis Hay et Péter Nagy. 195-204.