To to answer this question, we'll need to go back in time and look at the primary sources, and I'll limit the scope to the French corpus of 17-19th century coffee literature, so the focus will be on when and why the French started filtering their beloved coffee.

[Note: making coffee involves three fundamental operations that determine the quality of the final product: roasting, grinding, and brewing. The latter is done by mixing the ground coffee with (hot) water, which separates the desirable compounds from the rest (grounds, dregs, marc). I collected exclusively the parts about the brewing process, but, just like today, the early coffee makers were well aware of the importance of each operation, and discussed them at length.]

Part 1: 17th century

The habit of coffee drinking had spread in the Islamic world in the 16th century, reaching Europe a century later. The first coffee-house was established in Oxford in 1650, and before you knew it many Europeans were drinking it on a regular basis.

The most quoted source describing the arrival of coffee in France is Jean de la Roque's Voyage en Arabie Heureuse (1716), where the author tracks the history of coffee consumption in France. According to La Roque, his father had brought coffee beans and ustensils in Marseille in 1644 from Constinople, where the drink was available everywhere. This first attempt was short-lived, but Mediterranean traders kept bringing coffee to Marseille, where the first coffee house opened in 1671. In Paris, there were several false starts until the arrival in 1669 of Soliman Aga, Ambassador of Sultan Mehmed IV, who popularized coffee drinking. Entrepreneurs, many of them from the Levant, soon opened coffee-houses throughout the country. By 1676, the guild of the limonadiers (literally lemonade sellers, but actually sellers of hard alcoholic drinks like brandies) were given the right to sell coffee in grain, powder, or drink. A few decades later, there were hundreds of coffee houses in France, and it was also prepared in people's homes. But how did they prepare it?

The earliest extensive source about coffee in France is a small opuscule titled De l'usage du caphé, du thé et du chocolate (1671), written by erudite pharmacist Philippe Sylvestre Dufour or by his friend the physician Jacob Spon. The chapter about the coffee compiles several previous texts - notably one by the Rome-based Maronite Syrian Antonius Faustus Naironi, an avid defender of coffee - and deals mostly with its medical properties. When describing the way coffee is prepared by the Turks, it explains that ground coffee is put in boiling water and left to boil until it loses its bitterness.

Afterwards, this composition is poured into small porcelain bowls, of which I have spoken, to be drunk, as hot as the mouth and gullet can tolerate, being swallowed only little by little, and at different times, because of its present heat, that after it has taken on the flavour, the colour of this powder, whose mass descends, remains at the bottom of the jug.

In this description, the coffee is not filtered but the grounds are shown to be clearly separated.

Twelve years later, in 1685, Dufour published a revised edition of the book with more focus on practice. This time, Dufour provides a precise description of the brewing part of coffee making: boil water in a coquemart (the generic term for a boiling pot), throw the coffee powder in it, remove the pot from the fire when the mixture rises to the surface, replace it on the fire and repeat the process a dozen times. After that, keep the pot on warm ashes.

It should be left there for as long as it is deemed necessary so that, the marc having sunk to the bottom, the decoction remains clear and of a russet colour: it should then be poured into the cup by tilting it, lest the marc, which is completely worthless, should be remixed.

It is clear here that the dregs must be discarded and absolutely not be mixed in the final beverage. Dufour also introduces the word cafetière to name the coffee-making pot, possibly the first time that the word appears in writing in its modern sense in French (not the restricted US one of "French press").

In 1686, Anicet Caufapé, a physician from Toulouse, published a treaty on the "nature, preparation, and use of the most common foods". The entry on caphé explains that the coffee bean contains

two substances of very different nature: one is volatile, aperitive, detersive, subtle, and penetrating; the other, on the contrary, is crude, terrestrial, and astrigent.

Roasting liberates the former from the latter, which remains at the bottom of the caphetière. Only the "tincture" is useful, and Caufapé gives a long list of its admirable properties. The dregs are only good to "tighten a little the belly". His description of the brewing process is similar to that of Dufour, without the repeated boiling. In the end, the capethière is tilted and the tincture poured slowly in a cup.

The following year, Nicolas de Blégny, physician of Louis XIV, wrote a book about the "proper use of tea, coffee, and chocolate for the prevention and cure of diseases" (La Roque, writing 30 years later, accused him of plagiarizing Dufour without adding anything new). Blégny resents the fact that coffee, a "medical food" is prepared by limonadiers with no experience in the art of healing. His theoretical analysis of the nature of coffee is similar to that of Caufapé: the tincture obtained after extraction contains the beneficial "fatty and ethereal parts", while the "terrestrial particles" remain in the marc. If the coffee was to be consumed whole, it will

tire the stomach extraordinarily, decrease the movement and fluidity of the mass of the blood, and would necessarily result in obstructions in viscera full of capillaries, like the liver and the spleen.

A badly purified tinctured with "too many terrestrial parts" would result in the same problems while a well-prepared tincture will contain subtle and ethereal parts that will turn into a vapor going straight to the head.

Blégny then describes several methods for coffee preparation as well as the devices required at each step. Brewing is done in a cafettière that he says to be similar in France and England. Like the previous authors, Blégny recommends to pour the liquid slowly to prevent the dregs from going into the cup. It is also possible to accelerate the precipitation of the marc by throwing a little bit of cold water in the pot before removing it from the fire.

It is again very important that [the tincture] contains no marc of crude particles at all.

Still, he admits that some "igneous corpuscules" and "acid particles" should be part of the mix to give taste to the drink. Also, the "volatile substances" have to be bound to some remaining "terrestrial corpuscules" otherwise they would just dissipate.

Blégny then takes the opportunity to do some product placement for two devices of his own invention, a roasting machine and a caffetière. The latter has an enlarged bottom that "serves marvellously to retain the marc", and "some sort of filter on its spout" that only let the liquor pass through (top left picture).

There is little doubt that by that time coffee was meant to be consumed as a dregs-free "tincture". An "apology for the coffee", published anonymously in 1696 in the Mercure Galant, is a more literary take on Blégny's recommendations:

Coffee should be taken like potable gold, without mixing anything in it. It is even necessary to separate the marc from it, by precipitating it at the bottom with a few drops of water, because the marc is the dregs and the coarse part of the Coffee, which would weigh on the stomach, and would do it much harm. One must therefore only take the tincture of Coffee, a simple and very pure tincture; and this tincture being well made, is marvelous; because it contains only the most subtle, the sweetest, and the most sulphurous parts of this so salutary bean, with the exception of a few igneous corpuscles, which volatilize its parts, a few acid particles which make the flavor of this tincture, and some earthly substances, which serve to bind volatile matter, and to give it consistency. We must drink this tincture of coffee as hot as we can, in several sips, as we see the birds drink from their little troughs.

And here we are! Coffee, brewed by boiling in what was often called a cafetière du Levant, had been on French tables for less than 15 years, and people were taking pains to remove the dregs, supposed to be not just useless but detrimental to health, and the King's physician had invented a coffee maker with a built-in filter.

We can mention here a treaty by physician Nicolas Andry de Boisregard concerning the suitability of foods for Lent: the entry for Coffee proposes a new way to make coffee that "nobody had thought about before", which is simply to boil the raw beans and collect the resulting yellowish juice. For Andry, this drink contains the most light and ethereal natural extract, "sweet and subtle". It is unlikely that Andry's method was ever used - it would make a very weak coffee indeed -, but the reason he offers for doing this is interesting. For him, regular coffee is well-suited to Orientals, because they "eat in a very simple fashion" and can thus benefit from a strong coffee. Westerners, on the contrary, love eating foods that "irritates the taste" and they "heat their blood by thousands of sophisticated seasonings": coffee is thus too powerful for them. While Andry does not mention the dregs issue, his theory is in line with those of other authors who claim that the beneficial properties of coffee are mainly contained in its "ethereal substances" extracted by roasting and boiling.

->Part 2: 18th century