Every attempt at psychoanalyzing detective fiction seems to lead, ultimately and unavoidably, to the notorious notion of the primal scene in which, like in the drop of water, the whole enterprise of clinical as well as applied psychoanalysis is reflected. The seminal work in this respect is the by now classic paper by G.Pederson-Krag "Detective Stories and the Primal Scene" (1983/1949/). Appearances notwithstanding, all subsequent developments in the sphere where psychoanalysis and poetics/semiotics intersect not only remain true to the basic premises stated by the mentioned author but, in actual fact, unwittingly reinforce them. Our present inquiry derives part of its value precisely from the fact that it makes lucid this fundamental continuity which constantly escapes attention thanks to the innovative phrasing of post-Freudian revisions of his theory - Lacanian, Guattaria or Derridaean alike. Hence our first step will be to trace the connec-tion between the notion of the primal scene and a number af tools of literary criticism whose validity is generally taken for granted in order to question the adequacy of the interpretative model currently governing the humanities to the textual reality. In this respect the instructiveness of the detective fiction is obvious for to the theory of literature it is exactly what the primal scene is to the psychoanalytic theory: nucleus or foetus.

The two basic features defining the current interpretative model - intertextuality and the "implied" position of the reader in the text - are interdependent. However, this interdependency up till now has not been sufficiently appreciated. It is precisely with reference to the primal scene that it strikingly comes to the fore.

"The voyeur is never entirely satisfied with his peeping which has the compulsion endlessly to repeat like the detective story addict who rereads the same basic mystery tale without tedium ..

In participating in the detective story version of the primal scene, the reader's ego need fear no punishment for libidinal or aggressive urges. In an orgy of investigation, the ego, personified by the great detective, can look, remember, and correlate without fear and without reproach in utter contrast to the ego of the terrified infant witnessing the primal scene" (Pederson-Krag 1983/1949/ : 18-19).

Thus the fundamentally restorative that is, libidinal, nature of the compulsion to repeat in which we are accustomed to see the beyond of the pleasure principle becomes quite obvious. The repetition compulsion sets the intertextual machine going and at the same time guarantees the position of the reader, to wit, the very possibility of the interpretative attitude as such. The latter aim is achieved through the reader's identification with the detective 1) - a point stressed with equal force by all contributors to the anthology "The Poetics of Murder" (1983) which represents all the diversity of current techniques of reading - however, with one extremely significant lack. This lack being the lack of deconstructive interpretation of detective fiction should be thematized. If we remember that, according to Ferenczi, identification is a forerunner of symbolism, and, on the other hand, acquiesce in seeing in detective fiction "the peculiarly modern destillation of a general literary experience that makes central the subtle interaction with,and interpretation of,language" (Hutter 1983:234) - the resulting paradox seems to be self-deconstructive for deconstruction, and this because the literariness of literature defined in strict accordance with Derrida proves to be concomitant with symbolism to do away with which is one of the acknowledged goals of deconstruction. It might appear that the detective fiction as a distillation of literariness proves the validity of structuralist stance threatening the deconstructive one. However, on more close examination, we are bound to radically shift the grounds. It is namely my contention that if there is a threat to deconstruction it comes from within - from the fact that deconstruction, as yet, has not lived up to its name. By the same token, far from being in decline, as the saying goes, deconstruction can and must be radicalized.To demonstrate both issues on the evidence of detective fiction means to guarantee objectivity and provocativeness of our inquiry.

The only deconstructive encounter with detective fiction has been an indirect one: Derrida's critique of Lacanian reading of Poe's tale "The Purloined Letter". As B.Johnson has shown (1977), an attempt to divorce "structuralism" and "poststructuralism" has ultimately failed. However, the consequences of this failure are far more profound, than Johnson conceives of them. Significantly, Johnson ignores the main flaw of Derrida's critique which subverts the basic assumption of her analysis "that Derrida repeats the very gestures he is criticizing" (Johnson 1977 :154). According to Derrida, Lacan's misconstruction of Poe's text is due first and foremost to the elimination of the inter- textual dimension (Derrida 1988 ; 198-199), albeit Lacan himself, in stressing that "there are two scenes, the first of which we shall straightway designate the primal scene" (Lacan 1988 b : 30) not simply does optate for intertextuality but sets out to defend it in its most basic form as intratextuality without which there cannot be intertextuality proper. 2) It follows that the two exegetical procedures do not contradict each other but are complementary operating on two different, albeit interdependent, levels. The proof of this complementarity is the place granted by Lacan and Derrida alike to the reader.

The involvement of the reader to which intra-and/or intertextual stance leads - is the point where Derrida has to "subscribe" to Lacanian statements. 3). The psychoanalytical name for this involvement was already mentioned: restoration, reparation. And here Derrida who elaborates the notion of repetition compulsion while denying the very existence of the beyond of the pleasure principle (Derrida 1980) is more candid than Lacan who is bent upon preserving the notion of the death drive although turning it into its direct opposite.But it is this candidness which happens to be self-deconstructive for Derridaean enterprise.

The linking, unifying activity of Eros has its semiological counterpart in the notion of symbol on which the traditional - phallogocentric! - exegesis so obviously hinges. This linking is the main device of intertextuality propounded by Derrida as a deconstruction othe hermeneutical tradition. Hence two main paradoxes of deconstruction. On the one hand, the traditional exegesis comes to be reinforced pracisely at the moment which defines itself as the deconstruction of it, while on the other hand and by the same token, the revision of psychoanalysis along intertextual lines, cannot avoid the chief fallacy for which applied psychoanalysis has become notorious, namely reductiveness. Derrida is quite correct in pinpointing symbolism as the principal target of deconstruction, but the intertextual option is not a valid way if we want to attain this aim. The resulting misrecognition is twofold: the garbling of the textual reality in general and the denigration of the reality of Freud's texts in particular denounced for their alleged phallogocentrism. However, Freudian legacy abounds in intuitions which, if elaborated, can furnish us with that very tex- tual model for which deconstruction vainly seeks.

So long as the detective fiction is concerned the most convenient starting point of this elaboration is the concept of anxiety (Angst) inherent in the primal scene. The Freudian difficulties with this concept, although well-known, have not been sufficiently appreciated. And hardly surprising in view of the post-Freudian vicissitudes of his theory, especially of the applied version of it. (Cf.Birraux 1994; Quinodoz 1994; Le Guen et al. 1991; Lyotard 1974).

To put it bluntly, the main tendency in post-Freudian psychoanalysis was (and remains) the suppression (be it through denial or displacement) of the death drive, that is, the libidinization of the latter.As Lacan has pointed more than once,Freud's aim in introducing this notion was to secure the basic dualism of his theory constantly threatened with collapse. Irony resides in the fact that precisely in Lacanian "return to Freud" the two principal protagonists of Freud - Eros and Thanatos - tend to merge in a far more ticklish way than this has ever happened in texts penned by Freud himself. However, it is exactly the concept of anxiety which precludes the possibility of such mergence.

From the start the problem of anxiety has posed itself to Freud as a threat of subversion of his libidinal economy. The strategy deployed by Freud in order to circumvent this threat has undergone changes in the course of his career. The initial attempt was to(dis)solve the problem economically, that is, to see in anxiety a direct result of the conversion of libido. The conversion was said to take place when the tension has reached a certain degree: " ... the physical tension grows, reaches a threshold when, normally, it must produce a psychic affect but for some obscure reasons ... the energy remains unbound and we have anxiety (Angst)"(Freud 1962 : 83); "Where the sexual tension grows but cannot be processed psychically ... the result is a conversion of sexual tension into anxiety" (85)

This scheme has furnished the framework for all subsequent elaborations on the concept of anxiety - Freudian as well as post-Freudian. However, one profound difference between Freud and his disciples deserves to be stressed. For if the Freudian professional candidness allowed for an insight - the insight which was not obscured despite all efforts to find a place for anxiety in the libidinal economy, - Freud's disciples have said and done all to preclude the possibility of thematization along the lines highlighted by Freud. Freud's correspondence with W.FlieB bears witness for this professional candidness alluded to above. The economical solution, however imposing it might appear to be, has not satisfied Freud. Thus in the letter of 22.06.1894 we read: "Your opinion that with the whole anxiety-stuff (mit der Angstgeschichte) is something wrong tallies with my own suspicions" (86). The cause of this discontent lies obviously in the unavoidable question to which the economical solution leads: "But why the result of the conversion is precisely anxiety and nothing else?" (85)

The only answer of which Freud could think, seems to threaten thE basic postulate of psychoanalysis, to wit, Freud's libidinal option: "Anxiety is the expression of the accumulation of another 4) endogenous stimulus, the stimulus to breath (des Reizes zur Atmung) which by definition cannot be processed psychically" (85)

Thus anxiety neurosis (Angstneurose) - a derivative of the obsessional one - seems to disrupt the whole Neurosenlehre for its cause is not sexual but "another" one. This rupture is even more significant because the main device of the anxiety neurosis - conversion 5) - is the same through which the symptoms of hysteria come into being - the symptoms which, in the last analysis, allow for the semiological extension of Freud's theory. If we remember that all post-Freudian revisions of his theory have been undertaken precisely in view of such an extension the result being the current intertextual, i.e., basically, hysterical interpretative model, then to follow Freud's intuitions re- garding the anxiety neurosis means to highlight a promising way to overcome "postmodernism and its discontents".

To be sure, Freud was sufficiently conscious of the self-deconstructive nature of the propounded economical solution of the problem of anxiety. Hence his reservations regarding O.Rank's theory of "the trauma of birth". Paradoxically, it was precisely in discussing Rank's theory in "Inhibitions, Symptoms and Anxiety" (1926) that the most resolute attempt to libidinize anxiety was made.

I say "paradoxically" because the acknowledged aim of this paper was to refute Freud's earlier conception of direct conversion of libido in anxiety (cf.1926 : 138). However, on close inspection, this refutation proves to be an ingenious manoeuvre to "bound" anxiety, to semiotize it. But, thanks, once again, to Freud's candidness this manoeuvre was doomed to fail.

According to Rank, every experience of anxiety should be referred back to the initial traumatical experience of birth. Freud's reservations, polemically exaggerated, in fact, concern only the issue of secondary importance: namely, he refutes the notion of birth as trauma seeing in it the prototype of all latter situations of danger (Freud 1926 : 194). Thus the semiological status of anxiety seems to be secured. Anxiety has a prototype (Vorbild) which will be reproduced thanks to the intertextual device par excellence - repetition compulsion most obviously at work in hysteria. Which means that the whole theory of anxiety neurosis becomes in this crucial point indistinguishable from that of hysteria (cf.Freud 1926 : 164). Paradoxically, the premise of this mergence precludes the very possibility of it. For once again Freud has to answer the fatal question the subversivness of which led him to discard the economic solution of the problem of anxiety. In effect, in order to conceive of birth as a prototype one must explain what makes of it a pattern of anxiety. And once again Freud can evoke only the "physical experience of suffocation" (1926 : 165) which, as we have already heard, has by definition no psychic counterpart, to wit, precludes the possibility of semiological conversion. Therefore the chain of (intertextual) substitutions (anxiety at birthfear of separation from the breast - fear to lose a loved object - fear of castration etc.) is disrupted right at the outset for the first term differs qualitatively from the latter ones. In order to repair the damage Freud is forced to misrepresent the problem, to seek difficulties there where there are no difficulties at all.

Hence the foregrounding of the question whether anxiety is (and to what extent) expedient or inexpedient which leads to rather a leng- thy discussion (1926 : 164-179) culminating in the famous question: "Whence the durability of this reactions to danger?" (180) which remains unanswered. The impression ingeniously forced upon the reader is that the real puzzle of anxiety lies in its inexpediency albeit such inexpediency is a feature common to all pathogenic formations. To stress inexpediency of anxiety means to make of the latter not simply a symptom but a hysterical symptom for, as Freud has pointed more than once, what defines the productions of hysteria is precisely the moment of intentionality (cf. the case of Dora's amnesia /1905 : 175/), to wit, the element of desire in the formation of symptoms. To substitute desire for anxiety - this is what the Freudian strategy aims at.

This substitution is precisely the point where all semiological extensions of psychoanalysis - first of all those of Lacan and Derrida - converge. To optate for desire means to optate for hysterical mode of interpretation. Hence the silent elimination of the notion of the Id in contemporary theory accomplished by mutual consent and without undue advertisement. According to Freud, in case of hysteria it is more correct to speak not of the Id but of "the second consciousness" (1896 c : 11). According to Lacan, so long as desire is involved "we don't know if it should be located on the side of the unconscious or on that of the conscious" (1988a : 45). At stake is neither the issue of faithfullness to Freud nor the question of what remains of psychoanalysis after the amputation of such a basic concept. In fact, this amputation is the last consequence of the Freudian strategy aimed at libidinization of anxiety.

Reread the 32 nd Lecture of his "New Introductory Lectures on Psychoanalysis" (1933). The familiar strategy of "Inhibitions, Symptoms and Anxiety" attains here its final purity. The theory of the direct conversion of libido in anxiety is referred to only as an old error (1933 : 96) while the old goal can be reached more conveniently.

The solution is simple: it suffices to define anxiety as flexibility par excellence (95), that is, to conceive of it along the same lines which up till then governed Freud's explorations into the nature of libido. 6) Hence Ego whose chief characteristic is precisely flexibility (1926 : 185) happens to be endowed with the function to serve at one at the same time as the homestead of anxiety (1926 : 193) and the main reservoir of libido (1933 : 109). But in the framework set out by Freud himself this twofold endowement means that there remains no place for the compulsion to repeat which is not only provisionally suspended, so long as Ego's flexibility is only a libidinal one (1926 : 185), but from now on simply canceled out thanks to the mentioned enrichment. If we remember that the repetition compulsion is the intertextual device par excellence, to wit, the main mechanism of hysterical semiotics, then the Freudian (dis)solution of the problem of anxiety, as reconstructed above, proves to be the total self-deconstruction of his libidinal economy. Just the same self-deconstructive movement subverts Lacanian and Derridaean transcriptions of this economy along the lines of intertextuality.

However, so long as Freud homself is concerned, a number of intuitions run counter to this self-deconstruction forcing upon us the necessity to radically shift the grounds, to discard the traditinal view of psychoanalysis as first and foremost libidinal economy. If elaborated, these intuitions will prove invaluable for purposes of description of textual reality. It is these intuitions the brief summary of which I would like to give now.

Parallel to the development sketched above the formation of another, alternative, theory of anxiety took place in Freudian texts. The latter can be traced back to the same discontent in the initial economic solution of the problem and the concomitant decision "to separate that which produces libido and that which produces anxiety" (Freud 1962 : 202). However, this separation can have another result than the libidinization of anxiety.

Our main evidence is Freud's paper from "Studies on Hysteria" "On the right to separate from neurasthenia a certain set of symptoms as 'anxiety neurosis'". The semiological sabotage which defines obsessional neurosis in general happens to be even more radical in this version of it. The semiological mechanism of substitution, most obviously at work in the compulsion to repeat, is subverted by anxiety. "... phobia of the anxiety neurosis ... does not stem from the repressed idea, but, as the analysis shows, cannot be reduced (erweist sich bei Analyse nicht weiter reduzierbar) nor contested through our therapy. The mechanism of substitution is not valid in case of anxiety neurosis" (1895 : 322). Not just for reasons of professional candidness we are not going to conceal from the reader the fact that on the very next page Freud seems to take back this admission ("The content of phobia is substituted through another idea" /1895 : 323/). The contradiction is ostensible because the passage beginning with the sentence just cited ends with another one which can furnish a starting point of the real deconstruction of traditional exegesis: "... but the substitution in phobia is always only a belated one (...aber die Substitution immer nur nachtraglich zur Phobie hinzukommt)" (1895 : 323). Thus the mechanism of belatedness, Nachtraglichkeit, which has its semiological counterpart in such notions as structuralist "regressive semantic determination by context" (Mukarovsky) and poststructuralist "frame" (Derrida) both of them propounded in order to secure the intertextual dimension, is not simply dismissed but deconstructed. I think that this deconstruction is the only one which deserves to be dubbed so while in comparison with it that of Derrida proves conspicuosly similar to the Freudian libidinization of anxiety. As was shown above, the latter also presupposes a certain gap between the sources of libido and anxiety but only in order to close it through the notion of flexibility allowing to merge both of them. Analogously, a hermeneutical gap between intra- and intertextual dimensions of a given text is vital for reading practices of Lacan and Derrida alike but in the same movement this gap comes to be closed through the activity of the reader, through his/her involvement for which it allows. Hence the orientation towards transference which defines both interpretative models - the models equally stressing as the chief feature of textual reality the mechanism of production of ambiguous intertextuality. 7) On the other hand, the immediate consequence to be drawn for literary theory from the Freudian quotations is the necessity to define textual reality as the deconstruction of intertextuality in and by the very movement of its production. Intertextuality grounded in the tropological substitution (see the works of Harold Bloom) is a surface phenomenon, comes nachtraglich - only in order to be deconstructed by anxiety which governes the textual reality. The final chord of our theoretical exposition will elucidate this postulate.

By now it is sufficiently clear that if we want to separate unequivocally libido and anxiety it is not enough to discard the theory of direct conversion of the former into the latter. Paradoxically, theb real separation, that is, the deconstruction of libidinal economy not only tallies perfectly well with this theory but even presuposses it.

The intervention of anxiety paralyzes the hysterical semiosis grounded in the mechanism of Nactraglichkeit as an ambiguous movement forwards and backwards precisely because it makes impossible the backward conversion into libido. It is this impossibility which makes of anxiety - anxiety without prototype: Angst ohne Vorbid: "Right from the outset there is in the anxiety neurosis a psychic work aimed at binding anxiety, but this work can neither attain the backward conversion of anxiety into libido nor re-establish the connection with that comlexes from which libido stems" (Freud1909a : 350)

The complexes in question are nothing else than the intertextual prototypes invalidated by the intervention of anxiety The resulting subversion of intertextuality is concomitant with the subversion of the interpretative attitude, with the collapse of the "implied" position of the reader in the text secured thanks to the hysterical transference. To be sure, Lacan will be the first to acknowledge the destructive role of anxiety regarding transference but only in order to neutralize/libidinize/semiotize it through the familiar mechanism of Nachtraglichkeit, of intertextual see-saw ("comings and goings"): "Each time anxiety prevents a definitive identification, the fixation of reality (the imago stage). But this comings and goings will give its framework to that infinitively more complex real which is the human real. After this phase, in the course of which phantasies are symbolized, comes the so-called genital phase, in which reality is fixed" (1988a : 69)

The reparative nature of this Lacanian see-saw, to wit, of the Freudian fort/da game on which the former is modelled, precludes the possibility to see in it the beyond of the pleasure principle - the role assigned to it by Lacan. The game is a traditional prototype of human creativity in general. The reparative character of this latter is another doxa which, as every doxa, is mutely taken for granted without any corroboration. Lacan was among the very few who attempted such a corroboration. Significantly, upon the evidence of the detective fiction which for him is a distillation of literariness: "Our inquiry has led us to the point of recognizing that the repetition automatism (Wiederholungszwang) finds its basis in what we have called the insistence of the signifying chain ... It is that truth, let us note, which makes the very existence of fiction possible" (Lacan 1988b : 28-29)

Elsewhere we have already stigmatized Lacan's slyness -a number of crucial flaws in his interpretation of Poe's tale. Our careful analysis has proved that "The Purloined Letter" is not structured as a succession of two scenes one of them being primal, the other - the reproduction of it, as Lacan knowingly and Derrida unwittingly conceive of it. Now I am going to demostrate the same point otherwise - to further substatiate and elaborate my previous argumentation.

This time our evidence will be Agatha Cristie's novel "Towards Zero" (1972/1944/). Her novels, as detective fiction in general, abound with cross-references 8) which make of them, as one is so easily led to believe, a sterling example of intertextuality. Hence the question to be pursued is whether this intertextuality is what defines the textual reality proper. On the other hand, A.Cristie's novels bear witness for her interest in modern psychology. Hence another question: whether the textual reality in her case is governed by the hysterical semiosis which has its counterpart in Miss Marple's technique similar, as it seems to be, to the transferencially oriented psychoanalytic technique of free associations, - or by the utilization of another Freudian theory - that of anxiety as we have reconstructed it above.

As the title seems to suggest, the novel can be (mis)understood as an (arche)typical one within the framework of the genre, and by the same token as a lucide example proving the validity of the traditional exegesis in the psychoanalytical version of it.

"Detective fiction involves the transformation of a fragmented and incomplete set of events into a more ordered and complete understanding. As such it seems to bridge a private psychic experience, like dreams, and literary experience in general. And like psychoanalysis, the detective story reorders our perception of the past through language" (Hutter 1983 : 231)

It is exactly such reordering to which Cristie's title refers: "'I like a good detective story', he said. 'But, you know, they begin in the wrong place! They begin with the murder. But the murder is the end. The story begins long before that - years before sometimes - with all the causes and events that bring certain people to a certain place at a certain time on a certain day... All converging towards a given spot ... And then, when the time comes - over the top! Zero Hour. Yes, all of them converging towards zero ...'" (9)

Thus we have the initial doubling: the first, "primal" Zero Hour when the crime is commited, and the second - when it is solved thanks to the sleuth's ability to reproduce the primal scene in his mind, to reconstruct the order the things initially were. Derrida is quite correct in criticising Lacan for ignoring the fact that every doubling is doubled in its turn,that the two-ness is actually fourthness (1988:203).

Thus, in our case, Superintendent Battle's ability - all alibis and evidences notwithstanding - to protect the innocent accused of murder stems from the fact that the crime he is investigating is a second edition of his personal experience: "And there was the final damning evidence of Mrs.Strange herself - I don't believe there's one of you (except the one who knows) who can credit her innocence after the way she behaved when we took her into custody. Practically admitted her guilt, didn't she? I mightn't have believed in her being innocent myself if it hadn't been for a private experience of my own ... Struck me right between the eyes it did, when I saw and heard her - because, you see, I'd known another girl who did that very same thing, who admitted guilt when she wasn't guilty - and Audrey Strange was looking at me with that other girl's eyes..." (179)

I have said "Derrida is correct". To be more precise, he would have been correct if the textual reality was really governed by the laws of libidinal economy one of them being the doubling of doubles or the Derridaean "frame" which produces intertextuality along with securing the position of the reader. However, in our case the mechanism is disrupted through the intervention of anxiety, so that the textual practice perfectly tallies with textual theory as outlined above.

The girl "who admitted guilt when she wasn't guilty" - by matter of fact Superintendent's daughter - did so because of inexpedient anxiety. The same holds for Mrs.Strange: "'I was always afraid', said Audrey ... 'And I went on getting more and more afraid - the kind of unreasoning fear, you know, that makes you sick! I told myself I was going mad - but I couldn't help it ... But you don't know what it does to you being so afraid for so long. It paralyses you - you can't think - you can't plan - you just wait for something awful to happen. And then, when it does happen ... you'd be surprised at the relief! No more waiting and fearing - it's come. You'll think I'm quite demented,I suppose, if I tell you that when you came to arrest me for murder I didn't mind at all" (182-185)

It might be objected that what is said on the level of content changes nothing (or little) in what is done on the level of form. Although my contention is that in the wake of structuralism we have unlearned to read, that content and form are far more intricately intertwined than structuralists (and for all that, post-structuralists) would have it, in our case this objection is not up to the point at all. Significantly the mentioned interdependence is most obvious in case detective fiction celebrated by structuralism as the most "formal" of literary genres.

From the strictly formal point of view the crucial point is the solution of the problem/crime, that is, the technique deployed by the detective: "The ingenuity of the author is manifested in the preparation of such a situation and its simple and surprising resolution ... He first presents an event as inadmissible, and then accounts for it easily, elegantly, without forcing anything or using elaborate contrivances" (Caillois 1983 : 3)

Thus the textual theory stresses the simplicity of the solution achieved without use of "elaborate contrivances". If we grant A.Cristie the title of the Queen of the genre, then the textual practice in its exemplarity strikingly contradicts this theoretical assumption and does so thanks to the psychoanalytical connection, to wit, through the formal use of the concept of anxiety.

In the novel under discussion anxiety is precisely which precludes the possibility of a simple solution, thus subverting the interpretative attitude - first of the detective, then of the reader who identifies himself with the investigator along the lines of hysterical transference. However, sure Cristie's Superintendent might be that he has reconstructed the crime correctly, the author gives him no chance to prove this "simply" or "fairly".

"We police officers have to act on evidence - not on what we feel and think. But I can tell you that at that minute I prayed for a miracle because I didn't see that anything but a miracle was going to help that poor lady. Well, I got my miracle. Got it rightaway" (179)

The miracle is an unexpected witness who by sheer chance saw the criminal on the night of the crime. Although a miracle, strictly speaking, adheres to the order of "elaborate contrivances", taken for itself it does not seriously damage the libidinal economy or hysterical intertextual semiosis. From the post-structuralist point of view the miracle has its place in the structure, the throw of the dice does not eliminate hazard, as Derrida argues accusing Lacan for overlooking this issue. 9) However, Cristie takes a next step to account for which is not possible neither within the Lacanian nor Derridaean hermeneutics.

Not only because even after MacWhirter's story our detective cannot act legally: the evidence is still inconclusive for, as everybody knows, the criminal cannot swim. Hence the necessity of the tour de force as a last resort: Superintendent has to provoke the criminal, 10) to push him overboard in order to see whether he is really able to swim across the river, as he was said to do on the night of the crime.

But this impossibility to reach the solution by "internal" means is escalated to the point of total collapse of all hermeneutics. The hermeneutical collapse with which the story ends is staged as a two-step affair. The first step being MacWhirter's admission:

"I did not actually saw a man ... - indeed I could not have done so for I was up on Stark Head on Sunday night, not on Monday ..." From red Audrey had gone white. She said incredulously: 'Your story was all a lie?'" (190)

The second step being Audrey's conclusion that it was not simply a lie, but a twofold, mutual one:

"'Battle worked upon your story, yes. But I don't believe you fooled him. He deliberately shut his eyes'

'Why do you say that?'

'When he was talking to me he mentioned it was lucky you saw what you did in the moonlight, and then added something - a sentence or two latter - about its being a rainy night' MacWhirter was taken aback. 'That's true. On Monday night I doubt if I'd have seen anything at all'" (190-191)

It strikes the eye how neatly this picture tallies with our reconstruction of Freud's theory of anxiety. Anxiety sets out of work the basic device of intertextuality - the mechanism of Nachtraglichkeit, of belated understanding. On the level of content this device is pungently used against itself: Audrey recollects what Battle had said in order to understand the failure of understanding. On the level of form this means that the two scemes one of them being primal have no bearing on detective's technique: whatever he might think, the real solution is attained on entirely different level. Which has its exact counterpart in the Freudian notion of the impossibility of backward conversion of anxiety into libido to be understood semiologically as the invalidation of the libidinal intertextual prototypes.

Instead of functioning according to the rules of libidinal economy, ensuring intertextuality and the interpretative attitude, the two scenes are governed by anxiety and thus involved in the self-deconstructive movement which is even more dangerous for it can be assumed that these scenes are actually doubled, as Derrida would like them to be. The result being that the interpretative attitude implied in the text is a subversion of the very possibility of interpretation.

That anxiety diminishes understanding is readily admitted by psychoanalytic community so long as clinic is concerned. 11) But literature, as the saying goes, is a restorative, reparative affair (cf.Gagnebin 1994). However, the evidence provided by detective fiction - the genre where "literariness" is at its nakedest - contradicts this assumption. In the same movement in which anxiety subverts intertextuality, it blasts the "implied" position of the reader in the text represented by the detective. Which explains why the latter in our case cannot use the intertextual mechanism, to wit, to utilize structurally the alleged prototype, Vorbild of anxiety. The result being that he himself comes to be infected by anxiety, forced to acknowledge his inability to understand - already at the first, initial, primal stage.

"'No, Sylvia/Superintendent's daughter/, I don't understand, because I'm not made this way'" (21)

So long as clinical psychoanalysis is involved it is readily granted that anxiety precludes the possibility of transference. But we see that this preclusion defines the textual reality for which the applied psychoanalysis has to account. The self-deconstruction of the intertextual mechanism of Nachtraglichkeit is followed by the analogous self-deconstruction of the transferential mechanism thanks to which, as Lacan and Derrida argue, the reader comes to be envolved in the text. 13) This envolvement is secure while hermeneutical transference and intertextual Nachtraglichkeit are,in the final analysis,one and the same thing both being grounded in tropology. Every understanding, as Lacan and Harold Bloom have taught us, is a misunderstanding. But misunderstanding is the only way to secure communication, to avoid the danger of the impossibility to understand at all. But it is precisely this impossibility which in our case comes to be transfered from one personage onto another. Which explains why the final scene, from which we have cited, by the very movement of referring to the primal scene subverts intertextual hermeneutics: even after the "revelation" the personages are unable to free themselves from anxiety, 14) which comes to be transfered onto the reader.

The result being that the reader finds himself excluded from the primal scene, to wit, excluded from the text. The notion of the primal scene remains useful for purposes of describing the textual reality if we stress with all radicalness the features defining the primal scene in psychoanalytical clinic. The exclusion of the child from the primal scene, his inability to understand what is taking place gives rise to anxiety, neurosis. The same holds for textual reality - contrary to the restorative view propounded by the psychoanalytic orthodoxy (Pederson-Krag), as well as by the psychoanalytical revisionist (Lacan) or even by the "heretic" (Derrida). All of them foreground the restorative character of literature which is allegedly governed by repetition compulsion/transference/intertextuality. In our view it is governed precisely by anxiety neurosis that subverts the mentioned notions derived from the clinical picture of hysteria. 15) Therefore, contrary to the common assumption that the detective fiction "pretending to rouse the reader, in fact, reconfirms him in a sort of imaginative laziness and creates escape by narrating" (Eco 1983 : 113), or, to say the unsaid, that it is totally dominated by the pleasure principle, I think that this genre where some basic rules of literary textuality are encountered in the most pure form cannot be described better than with the words used by Graham Greene to characterize poetics of another pretender for the title of the Queen of Crime - that of Patricia Highsmith who "has created a world claustrophobic and irrational which we enter each time with a sense of personal danger" ( Greene 1984 : 9).

NOTES