#How long were Maupassant's short stories?

##Introduction

Guy de Maupassant wrote about 300 short stories, most of them in barely ten years, between Boule de suif in 1880 and his illness at the end of the decade. Reading them again, I was struck by how regular they are: a few pages, one situation, a twist, and it's over. Was it a style, or a format?

Most of these stories were first published in daily newspapers, before being collected in books. So I counted the words of each one, and compared them with where and when they appeared.

##Data

The site maupassant.free.fr lists the short stories in chronological order, with, for each, the newspaper or book where it first appeared, the date, and the pseudonym used, if any. It also has the full text of each story, so I downloaded the 306 of them.

import requests
import pandas as pd
from bs4 import BeautifulSoup

stories = pd.read_csv('maupassant.csv')  # url, title, publication, date, pseudonym

for url in stories['url']:
    r = requests.get('http://maupassant.free.fr/' + url)
    soup = BeautifulSoup(r.content, 'html.parser', from_encoding='iso-8859-1')
    text = soup.find('div', class_='texte').text.strip()
    with open('transcripts/' + url.split('/')[-1].replace('.html', '.txt'), 'w') as f:
        f.write(text)

Then the length of each story, without its title (the first line of the file):

import re

def count_words(path):
    text = open(path).read()
    body = text[text.find('\n') + 1:]
    return len(re.findall(r"[\w’'-]+", body))

##A newspaper format

Each dot is a story, placed at its date of publication and at its length, on a logarithmic scale:

The median story is 1,900 words long, and 80% of them fall between 1,200 and 2,800 words, the grey band: about ten minutes of reading. That's the space that a daily newspaper gave to its conte, the short story it printed alongside the news, often several times a week.

The two newspapers where most of them appeared show it well. Gil Blas published 162 of them, with a median of 1,800 words; Le Gaulois 93, with a median of 1,850 words. In Gil Blas, until 1885, Maupassant often signed with a pseudonym, Maufrigneuse: 107 stories in all.

The long stories appeared elsewhere. Boule de suif (13,400 words) came out in a collective book, Les Soirées de Médan, in 1880, and made Maupassant famous overnight. Yvette (23,900 words) and L'Héritage (23,200 words), the two longest, were published as serials, over several weeks. The 9 serialised stories have a median of 10,000 words, five times more than the others.

At the other end, the shortest is La Bécasse (680 words): the frame of a collection, in which a group of hunters tell each other stories after dinner, rather than a story of its own.

##Ten years, with three busy ones

The output is concentrated in a few years: 188 stories between 1882 and 1884, more than one a week, while Maupassant was also writing his first novels, Une vie (1883) and Bel-Ami (1885). From 1885 on, the pace slows down: the novels take more of his time, the stories get a little longer, and they move from the daily newspapers to Le Figaro and L'Écho de Paris.

After 1890, only a handful remain. Ill with syphilis, Maupassant was committed to a clinic in Paris in January 1892, where he died in July 1893. The last story in the list, Le Colporteur, came out in Le Figaro in March 1893, while he was there.

##So, style or format?

Both, probably. The newspaper set the length: between 1,500 and 2,500 words, the story had to fit in its column. But Maupassant made this constraint his style: one situation, few characters, no digressions, and an ending that falls in the last lines. When he had more space, in serials or books, he wrote longer stories, and some of them, starting with Boule de suif, are among his best known.