The question occurred to me in 1997 as I was preparing an exhibit called A Diary of One's Own, created by the Association pour l'Autobiographie at the Lyon public library (Lejeune and Bogaert). My approach was didactic: I wanted to construct a story where the spectator would follow the different phases in the life of a diary, just as in the good old days, in primary school, they used to show us the workings of the digestive system, beginning with a mouthful of bread. A story, Aristotle will tell you, must have a beginning, a middle, and an end. Here, it needed to follow the diary writing process represented in the exhibit.
WHAT IS THE END OF A DIARY?
The problem arose at three different points in the sequence we were planning
for the exhibit.
-- The beginning of a diary is almost always indicated: it is rare to begin one without saying so. In one way or another, you mark off this new territory of writing--with a name, a title, an epigraph, a commitment, a self-presentation.... We had plenty of such beginnings, so we wondered if similar rituals existed for ending a diary.
-- In the exhibit's section on time, we examined the first and the last pages of diaries that had been kept throughout an entire life: the diary of Amiel (1839-1881, forty-two years worth of writing, 173 journals, 16,800 pages), of Jehan Rictus (1898-1933, thirty-five years, 153 journals, 34,800 pages), and of Claude Mauriac (1927-1995, sixty-eight years, we have yet to count the total number of pages, but the journal measures three and a half meters)  The section on time was meant to demonstrate the immense duration of an existence, to show the transformation of diary-writing over time. We didn't expect it to tell us much about endings, since the diary writer is often not the author of the diary's end and doesn't even know that "this" page would be the last.
--The end of our sequence led us to examine the idea of an ending, leaving aside those de facto endings (the most numerous kind) that weren't experienced as such, and trying to grasp the ending as an act, in all senses of the word:
a) a voluntary and explicit stop (to a journal that has not been destroyed);
b) the destruction of a diary (an energetic and definitive closure);
c) a rereading (subsequent annotation, table of contents, indexing);
d) publication (a transformation that assumes some sort of closure).
We were overwhelmed. What a contrast between the simplicity of a diary's beginning and the evanescence of its ending: the multiple forms ending can take (stopping, destroying, indexing are all different, even opposite actions); the uncertainty of point of view (is the ending the act of the person writing--and at what moment of writing?--or of the person reading?); arid the impossibility, most of the time, of grasping this death of writing.
Classic texts on the diary, in French at least, offered no help: they ignore the problem. Nor did "How to" manuals provide any assistance. They are frill of good advice on how to wrap up an autobiography. But it wouldn't occur to anyone to explain how to end a diary. It would be like writing a treatise on suicide.
To unravel this knot, I made a compilation of sixteen endings from diaries. I have followed Descartes' method, to "divide these difficulties into as many fragments as necessary to best resolve them," but without intending to resolve them. I will thus distinguish among:
1) The ending as a horizon of expectation. I will try to show how the diary is experienced as writing without an end;
2) The end seen in relation to finality, or rather, to the possible finalities of a diary (I will distinguish four);
3) The end as reality, the diary faced with the death (natural or voluntary) of its author.
These will be three meditations on writing, life, and death.
HORIZON OF EXPECTATION
"To be continued in the next episode...." "Stay tuned...." When you write the entry of the day, you don't know what will happen in the next episode, nor whether it is already inscribed in the Big Book, as Jacques the Fatalist would say. But by writing today, you prepare yourself to be able to live tomorrow, and to piece together, in a predetermined framework of writing, the story of what you will have lived. All journal writing assumes the intention to write at least one more time, an entry that will call for yet another one, and so on without end. I once had a look at the hospital journal of the poet Louis Guillaume--a small, unfinished notebook. Each day, after finishing his entry, he would write down the next day's date. One day, without even knowing it, he wrote down the date of his death, and the diary remained on hold. This is why it is so pleasurable to purchase a new date-book every January. It's an annual life insurance. The diarist is protected from death by the idea that the diary will continue. Th ere is always writing to be done, for all eternity. The intention to write one more time presupposes the possibility of doing it. You enter into a phantasmagoric space where writing runs into death. The infinite post-script....
Here, for example, are the last lines from Andre Gide's last text, a meditation in the form of a diary, written six days before his death. In it, you can observe two contrary yet complementary tendencies: in the first paragraph, the refusal to finish, the desire to write, to write anything at all, in order to go on living; in the second, a theatrical "final word," doubtless pre-meditated. If you really have to finish, better do it yourself, and carefully polish your exit:
No! I cannot admit that with the end of this notebook, everything will be over; that it will be done. Maybe I will want to still add something more. Add I don't know what. Just add. …