lf.A personal publicationItaliano
Thresholdsliminalfinds.com

The Alarm Rings at the End

A dream seems to take shape around the sound that ends it. In 1946, Jean Epstein explored the puzzle through cinema; Pavel Florensky had confronted the same question in 1922, in an essay that would not be published until 1972.

In the dream you are crossing an enormous city, certain that you have to get somewhere and with no idea why. Traffic holds you up, the streets grow tangled, and then a house appears and you know at once that this was where you were going. There are still stairs to climb, corridors that knot into one another, a door. You press the bell. The bell is the alarm clock, which has been ringing for a few seconds in the real room.

Jean Epstein tells this dream in 1946, in a book called L'Intelligence d'une machine. Epstein was a filmmaker, the director of La Chute de la maison Usher, and the book is a reflection on what cinema does to time when it speeds it up, slows it down or runs it backwards. The dream comes in almost in passing.

It starts with a child. On the screen a car moves steadily forward, but its wheels seem to turn in jerks, now one way, now the other, and at moments slide without turning at all. The child asks an adult to explain, gets an answer he does not quite follow, and from that day on, Epstein writes, this "twelve-year-old philosopher" keeps a certain distrust of a spectacle that offers such a capricious portrait of the world.

Then comes the film projected in reverse. Dead leaves rise from the ground and fasten themselves back onto the branches, raindrops climb towards the clouds, a locomotive swallows its own smoke. "The effect has become cause, the cause effect." Epstein stresses that inside that small reversed world, watched as many times as you like, the sequence repeats with the regularity of a law. There, everything happens as if the smoke were the cause of the fire. Hence the sharpest sentence in the book: causality seems to be only "a mental colour" that we give to certain degrees of probability.

This is where the alarm-clock dream arrives. For Epstein the sound that wakes the sleeper is both the end and the cause: "the pre-established end and the posterior cause" of the whole series of dreamed events. His explanation runs like this. The stimulus takes a few seconds to pass through the thickness of sleep, and during those seconds the sleeper's inner time runs far faster than the time outside, fast enough for the imagination to build a story that makes the sound plausible. The chapter closes on an image: a film in which "the plough pulls its oxen and the smoke falls into its chimney".

Epstein presents the dream as a common one and cites no one. Sleep researchers of the nineteenth century knew the type well. In 1875 F. W. Hildebrandt reported that the ring of his alarm clock had entered a long, coherent dream "perhaps a hundred times", "as if the whole dream had been laid out for it alone", and gave three examples; in one, the alarm becomes the bell of a village church. Freud reproduces them in The Interpretation of Dreams, next to the dream of Alfred Maury, who lives through the Reign of Terror all the way to the scaffold, feels the blade of the guillotine and wakes because the headboard of his bed has fallen onto his neck.

In 1922 an Orthodox priest who was also a mathematician and physicist, Pavel Florensky, wrote an essay on the icon, Iconostasis, and opened it with the same dreams. He cites Hildebrandt's three alarm clocks and the guillotine dream, "which has made the rounds of every psychology textbook". Florensky sees in these dreams something Epstein would observe through cinema twenty-four years later: a sequence in which whatever comes last seems to determine everything before it. The guillotine closes a perfectly coherent story that began in the spring of the revolution, and it coincides with the blow of iron on the neck that woke the sleeper. So, he concludes, in dream time the cause appears after all its effects, closes the series and determines it "as final cause, telos". In the dream, he writes, time runs towards the present.

When Epstein published his book, Florensky's essay was still unpublished, and its author, sentenced to death while held in the Soviet camps, had been shot in 1937. The essay would appear only in 1972, in a theological journal issued by the Moscow Patriarchate, nineteen years after the filmmaker's death. There is no record of any contact between the two. They belonged to different worlds, divided by language, faith and profession. Florensky reads final causality within a theological vision: the dream is the first step of life in the invisible, and art is "condensed dream". Epstein uses cinema to question whether our categories of time and cause are necessary. And yet both work from a kind of dream already described by nineteenth-century psychology, and draw from it the same figure: an end that acts on the beginning.

What the dreamer experiences must be distinguished from what they remember upon waking. Memory is already a reconstruction. Neither writer proves his thesis; both describe the shape of that experience with precision.

That shape is what matters here. In the dream, something that has not yet happened seems to arrange a story that leads, without a misstep, straight to it: the bell is expected, prepared for, like the last line of a well-written scene. Epstein had a machine that made something similar visible to people wide awake, in the dark of a cinema, with the plough pulling its oxen. What remains is the ordinary gesture of every morning, the hand reaching for the alarm clock, and a question their interpretations cannot close: whether the story of the dream was already there before the ring, or was built around it in an instant.

02 / A Note on Sources

A Note on Sources

Jean Epstein's L'Intelligence d'une machine (Paris, Jacques Melot, 1946) was read in the digital transcription of Les Classiques des sciences sociales (Université du Québec à Chicoutimi, 2002), made from the 1946 Melot edition. No printed copy or scan was consulted, and no page numbers are given. The passages used (the car and its wheels, in the section "Signes"; the reversed film, "Tête-à-queue de l'univers"; the "mental colour", the alarm-clock dream and the image of the plough that closes the chapter "L'envers vaut l'endroit") were checked in French against that transcription. The date of La Chute de la maison Usher (1928) is taken from the Cinémathèque française record; the date of Epstein's death (2 April 1953, Paris) from the authority record in the catalogue of the Bibliothèque nationale de France. Hildebrandt's dreams (Der Traum und seine Verwertung fürs Leben, 1875) and Maury's (Le Sommeil et les rêves) were read in Freud's Traumdeutung, chapter I, in the section on external sensory stimuli, in the textlog.de transcription; Hildebrandt's and Maury's own texts were not consulted. The Russian text of Pavel Florensky's Iconostasis was read in the 1260.org transcription; the critical edition edited by A. G. Dunaev (Iskusstvo, 1994) and the English translations were not consulted. Its first publication, in no. 9 (1972) of Bogoslovskie trudy, a theological journal of the Russian Orthodox Church issued by the Moscow Patriarchate, is taken from that issue's table of contents. The death sentence of 25 November 1937 and the execution of 8 December 1937 are given in the Gulag-Memorial biographical record (Vera Ammer). Translations from the French, German and Russian are mine.

03 / The find

L'Intelligence d'une machine, 1946

Jean Epstein, L'Intelligence d'une machine (Paris, Jacques Melot, 1946), chapter "L'envers vaut l'endroit". Set beside Pavel Florensky, Iconostasis (1922; first published in Bogoslovskie trudy, no. 9, Moscow, 1972).

Epstein

Freud, Hildebrandt, Maury

Florensky

If you made it this far, leave your stamp.

Like an old library card: each stamp marks a reader who made it to the end.

LIMINAL FINDS Library loan card

If this was worth your time, you can support Liminal Finds (opens in a new tab).