There is a sentence that has been repeated at Herstal for generations: that Charlemagne was born here. It appears on leaflets, in guidebooks, on shop signs. As things stand, it cannot be verified.
Our workshop occupies a farm raised on the former palatial estate. We have an obvious interest in the legend being true. All the more reason to look at what the sources actually establish, and to find that the documented truth is more interesting than the tradition.
A date that has changed sides
For centuries, Charlemagne's birth was fixed at 2 April 742. The date goes back to a reading of Einhard, the biographer who wrote the Vita Karoli Magni around 830. It passed into encyclopaedias, textbooks and commemorative plaques.
It is now widely disputed. German scholarship has argued for the years 747 or 748, and this reasoning has convinced much of the research community. The argument turns on family chronology: Charlemagne's younger brother, Carloman, was born in 751. A gap of nine years between a Carolingian couple's two eldest sons is unusually long.
A second clue comes from a text closely tied to the Capitulary of Herstal. Around 775, a cleric named Cathwulf wrote to Charlemagne and reminded him that the clergy had prayed, before his birth, for his parents to be given a child. This implies that Pepin the Short and Bertrada of Laon were already married. Their religious union, however, is dated to around 743 or 744, after the traditional date of 742.
Research carried out at the University of Liège sums up the current position: the eldest son of Pepin the Short is first mentioned in 754, on his first official appearance, and the annals point to 2 April 747, more probably 748.
A place no one can establish
Einhard gives a date. He gives no place at all.
The candidates proposed down the centuries say a great deal about each region's stake in the memory: Quierzy-sur-Oise, Ingelheim am Rhein, Aachen by the pen of Victor Hugo, Herstal and Jupille. No contemporary source settles the question.
Research from Liège is explicit on this point: there is no trace of the king of the Franks having been born at Herstal. It allows for a possible line of reasoning, but a conditional one, which depends precisely on the abandoned date. Had Charlemagne been born in 742, a birth at Herstal would be plausible, since that year Pepin the Short and his brother Carloman were setting out to fight the Alemanni and the family was staying in the Meuse region. The shift of the date towards 747 or 748 weakens this hypothesis.
The Town of Herstal maintains a careful wording, which deserves to be quoted for what it is: a claim, not a certainty. It states that Charlemagne may have been born there, and names the presumed site, La Préalle. That is the district where our farm stands.
We will not turn this proximity into proof.
What is established, and it is worth more
Tradition loses on the birth. It gains considerably on the rest.
Herstal first appears in the texts in 723, on the occasion of a stay by Charles Martel. The place was then a villa, a large rural estate. Its status changed under Pepin the Short: from his accession to the throne, the sources no longer speak of a villa but of a palatium. Carolingian stays there are attested from 723 to 920, across nearly two centuries.
The most solid fact is also the most overlooked. From 770 to 784, Herstal was Charlemagne's favourite residence. Fourteen years, at the very moment he was building his kingdom before turning it into an empire. A favourite residence, within an itinerant power that moved constantly from one estate to another, is no small detail: it is the place one returns to, where one governs, where one decides.
It was there, in 779, that he promulgated the first great capitulary of his reign, twenty-three articles reorganising the religious and public life of the kingdom. The text is the first to use the word capitulare in a legislative sense. Every great text that followed, up to the imperial coronation of the year 800, proceeds from it.
A sovereign is born somewhere once. He chooses where to govern for fourteen years.
The palace was a farm
Here is the point that touches us most directly, and it is thoroughly documented.
The word palace is misleading. In the Carolingian period, a palatium was not a building of marble and colonnades. It was an agricultural estate where the prince resided, a large fortified holding that had to feed a court, officials, servants, delegations. The bishops gathered at Quierzy put it plainly: it fell to the king to have the means to live decently and hold his rank through his estate, and to receive there those who came to the palatium.
The reference study, published in 1975, reconstructs the site, for lack of physical remains, through topography, old land registers and written mentions. It describes a scattered complex: a stone dwelling for the sovereign, an oratory dedicated to Our Lady, and various working buildings, among them a mill. The palatial estate covered more than 2,500 hectares, taking in the former parishes of Wandre, Vivegnis and Herstal.
A farm, a mill, working buildings, an oratory. That is what Charlemagne's palace was.
We make chocolate in a farm built on that estate. The continuity is not symbolic, it is functional: this ground fed an imperial court before it fed us.
The mill and the water
The palatial estate was not confined to its centre. Covering more than 2,500 hectares, it required outlying holdings, satellite farms, and above all water.
Herstal had it. Three streams ran down from the surrounding heights, meeting near what is now Place Beck before continuing towards the Meuse. They were called the Rieux des Mollins, the mill streams. The place names outlived the streams: the Rue des Moulins still climbs towards Vottem, and the watercourse now runs underground, at the bottom of the gardens along Rue Basse Préalle.
That is precisely where our farm stands. Local history devoted to Herstal's old topography states it plainly: the farm now housing the chocolate workshop began life as a mill.
We do not claim our mill was the palace's own mill. We note that the site was given over to grinding grain on an estate whose primary purpose was to feed, and that it has kept that calling.
What the stone says, and does not say
A word on a legend that circulates, and which we do not repeat.
Tourist guides report that a large stone near the entrance gate bears the mark of Charlemagne, who is said to have sat on it. It is false. It is a Meuse pebble, extracted from the building more than ten centuries after the Emperor.
We apply the same principle to the birth. Herstal's own Tourist Office, in presenting the farm, states that the attribution to the emperor's birth is not confirmed by any serious historical source. We share that caution.
A dig under way, and a discovery
For two and a half centuries, locating the palace remained a scholarly hypothesis. The 1975 study placed it near the present-day Place Licour, with no dig able to confirm it.
That has just changed. In 2024, during works to extend the Liège tramline, the Walloon heritage agency opened an archaeological excavation at Place Licourt. In January 2025, it announced a major first discovery: the foundations of a church, along with limestone sarcophagi dated to the 7th and 8th centuries.
The quality of the burials points to high-ranking figures of the Pippinid era, the very lineage from which Charlemagne descends. Is this the palatial church, the oratorium that the 1975 study placed at the heart of the estate without being able to prove it? Caution is called for, but the convergence between the 1975 topographical hypothesis and the 2025 discovery is remarkable.
After thirteen centuries, the palace is beginning to emerge from the ground.
What we take from this
We could have settled for the convenient version. Charlemagne was born here, and the chocolate is ennobled by it.
The documented reality suits us better. The ground we occupy belonged to an estate that was, for fourteen decisive years, the place a sovereign chose to return to. He legislated there for a Europe still taking shape. That estate was an agricultural holding built to feed a court, and the buildings we live in first served to grind grain.
It is a filiation of place and function, not of blood. It can be checked, and that is what gives it its worth.
Research continues, at Place Licourt as in our own archives. This article will be corrected as the sources speak. That is not a weakness. It is the opposite: a house that dates from 1974 can afford to wait for history to become clearer.