She has weaknesses, certainly, but they are, if anything, ‘respectable’ ones. She is slightly greedy and self-indulgent; she is too tolerant of her husband’s frailties and generally too easygoing; most of all, she fails to apprehend the extent of the threat that Lantier represents to her happiness. But these are minor faults. There is no inherent vice in Gervaise (as Zola insistently tells us there is in Nana). She is clearly a victim of circumstances.

These circumstances can and ought to be changed: this is the moral that Zola wanted readers to draw from the book, as he insisted in his letter to Albert Millaud (written in response to Millaud’s article in Le Figaro on 7 September 1876):

… if you wish to know the lesson that will emerge by itself from The Drinking Den, I would put it more or less into the following terms: teach the worker in order to give him morals, remove him from the poverty in which he lives, strive against overcrowding and promiscuity in the slums where the air is thickened and poisoned, and above all prevent drunkenness which decimates the common people by killing the mind and the body.

Again, in an extensive letter to Yves Guyot, editor of Le Bien public (published in that paper on 13 February 1877): ‘If I were absolutely obliged to do so, I would say that the whole of The Drinking Den might be summed up in this sentence: Shut the drinking-houses, open schools…’

The question of housing is crucial, he goes on to say: the stenches, the overcrowding, the tiny rooms in which fathers and daughters, brothers and sisters sleep side-by-side… The evil is compounded by heavy work that turns men into animals, and low wages that deprive them of hope, filling the drinking dens and the brothels: ‘Yes, the people are like that, but because society accepts it.’ Even though Zola insists (to Millaud) that his book ‘is not a work of propaganda’, he also claims (to Guyot) that it is ‘a useful work’ – and the importance of the message is illustrated by the fact that the writer’s very first outline for the Rougon-Macquart, given to his publisher in 1869, speaks of a novel that will show the decline of the Parisian workman ‘because of the deplorable influence of… the drinking dens’ and a demand for ‘air, light and education for the lower classes’. From the very start, this was intended to be a book directed at the social conscience of its readers. The underlying motive is not to excite feelings of repulsion or horror at the grotesqueness of an underclass that is defined as other, as alien and threatening, to be made safe through repression or repentance. On the contrary, these are human beings who are suffering ills brought on by identifiable and preventable causes.

What Zola carefully avoids is the suggestion that depravity has much to do with the misery he describes. Other writers about the urban working class had described the poor quarters of the city and told of the crime and degradation that festered in them in a way that would cause a little frisson of excitement in respectable drawing-rooms: Eugène Sue’s Mystères de Paris, for example. Zola’s novel excites pity, not anxiety; and he quite justifiably protests (in his letter to Le Bien public) against the charge that he has depicted working-class characters who are all uniformly debased and vile. On the contrary, he insists, there is only one real scoundrel in The Drinking Den, and that is Lantier – and even he is not a workman, but wears a jacket instead of a smock and has pretensions to gentility. As for the other characters, the Lorilleux may be miserly, but they are not idlers or drunkards; they show what happens to ‘the slaves and victims of the sweat shop’; nor are the Boches or the Poissons idlers and drunkards; and as for Goujet, he is ‘the perfect worker, clean, thrifty, honest, who loves his mother and does not miss a day’s work, but remains great and pure to the last…’

So, are Gervaise and Coupeau idlers and drunkards?

Not at all: they become idlers and drunkards, which is a different thing altogether… Gervaise is the most appealing and the sweetest character that I have ever created; she remains good and kind to the end. Even Coupeau, in the frightful illness that gradually takes hold of him, preserves the good-natured side to his personality. They are patients, nothing more.

In other words, alcoholism, poverty and the despair they breed are social diseases.

It may sound, by now, as if Zola was being slightly disingenuous in protesting that his novel was not a work of propaganda; but consider the means that he uses to convey his message and to gain our sympathy for his characters. Critics of the book were quick to attack its use of slang, especially of taboo terms or obscenities. In doing so, they overlooked the novel’s most original linguistic feature, which is the use of the device known as style indirect libre, a form of narrative that stands somewhere between direct and indirect (or reported) speech: ‘Gervaise didn’t want a big wedding. What was the sense in spending all that money?’ (Chapter 3) – where the second sentence (my italics) conveys Gervaise’s opinion on the matter without the addition of ‘she thought’, ‘she asked herself or any similar phrase.

One might imagine that having more than one narrator would be confusing; but Zola achieves the effect of a single, harmonious narrative by the rapidity of his transitions between this form of subjective narrative and the more usual objective narrative – that is to say, between the words of an authorial narrator and the voice (or, more often, inner voice) of one of the characters:

As it happened, the Boches had left the Rue des Poissonniers on the quarterly rent-day in April, to become the concierges at the big house in the Rue de la Goutte-d’Or. What a small world it was! One of Gervaise’s main worries, after living in peace without a concierge in her place in the Rue Neuve, was that she would once again fall under the tyranny of some fierce dragon with whom she would have to argue over a little spilled water or a door closed too noisily one evening. Concierges are such a vile breed! But with the Boches, it would be a pleasure. They knew each other and always got along well. They were like family. (Chapter 5)

The subtlety of the proceeding is well illustrated by this paragraph, where almost every sentence could be said to represent a different form of narration from the preceding one: a mere statement of fact is followed by an interjection (‘What a small world it was!’) not attributed to any narrator; then an omniscient author’s explanation of how his character feels (‘One of Gervaise’s main worries… was…’), then what, in the light of these feelings, is clearly an exclamation by Gervaise herself (‘Concierges are such a vile breed!’); and, finally, three short sentences in which the voice shades gently from Gervaise’s into one that could just as well be the author’s (‘But with the Boches, it would be a pleasure… They were like family.’).

If the voice of the omniscient narrator in most nineteenth-century fiction has the effect of making us feel that we can observe the central characters’ reactions and feelings, here the smooth flow from one voice to another invites us to share them. Gervaise’s mind mediates the action, not in the somewhat artificial way of a first-person narrator, who is supposed to be telling us the story as it happens, but with the immediacy of an omniscient author, from moment to moment abolishing the distance between the novelist and his character. Zola is the facilitator who allows Gervaise to tell us her story – and this, of course, is precisely how he wanted us to see him, as the artist whose work serves to give a voice to those who cannot express themselves, the cry of the silent masses, those who in most nineteenth-century fiction are depicted as mindless, grotesque or comic. The proceeding is all the more shocking to his bourgeois readers because it reveals the inner life of this woman whom most of them would suppose to have none. Can a laundress think and feel and hope? Can a laundress be touching and lovable? Can a laundress be tragic?

She thinks and feels in the everyday language of her class, which is why Zola also felt obliged to offend the sensibilities of his readers by the use of slang. This was not his own everyday speech and he researched it as conscientiously as he did the locations for the novel. He turned mainly to two monographs, Alfred Delvau’s slang dictionary, Dictionnaire de la langue verte (1866), and Denis Poulot’s Question sociale, le sublime, ou le travailleur comme il est en 1870 et ce qu ’il peut être (1870), a curious work that analyses different types among the members of the working class, from the ‘true worker’ to various types of idler whom Poulot classifies under the general title of ‘le sublime’. Looking at his definitions, it is not hard to apply them to the characters in The Drinking Den, with Goujet as the ‘true worker’, Lantier as a ‘sublime’ of the type Poulot calls ‘Son of God’ (who ‘wears a jacket… talks politics, reads the newspapers, very idle… allows others to support him’) and Coupeau as the true sublime who works half a week and is always between two glasses of spirits. The names of Coupeau’s drinking companions, Bibi-la-Grillade, Mes-Bottes and Bec-Salé, are taken directly from Poulot (something that helped to encourage the charge of plagiarism).

Clearly, much of what Poulot says might have been applied to urban workers elsewhere, but conversely a good deal is specific to the time and place about which he was writing.