On almost every map of the Germanic peoples a name sits along the southern Baltic: Suardones. In the manuscripts from which that name is taken, it does not appear. What they read is “Suarines”. So how does a people enter the books that the tradition does not actually know? The answer leads straight into the workshop of textual criticism – and shows that every map of antiquity is, first of all, an editor’s decision.
Open any book on the Germanic peoples, look for the group of seven listed in chapter forty of Tacitus’s Germania, and you will almost certainly find seven names: Reudigni, Aviones, Anglii, Varini, Eudoses, Suardones, Nuithones. Six of those stand more or less like that in the manuscripts. One does not. The authoritative witnesses read Suarines at this point. The form Suardones appears, according to the available collations, in a single manuscript – and there not even in the text, but in the margin.
This is not a detail for specialists. It is the textbook case of how a scribe’s note becomes a people, which then travels through dictionaries, translations, school essays and maps as though no alternative had ever existed. This article traces that road – from a lost codex through humanist copies and the great critical editions to the search engine of today.
The sentence in question is short. After the Langobardi, whom Tacitus praises for their small numbers, it runs: Reudigni deinde et Aviones et Anglii et Varini et Eudoses et Suardones et Nuithones fluminibus aut silvis muniuntur. Seven names, protected by rivers or forests, followed by the famous remark that nothing about them individually is worth noting – together, however, they worship Nerthus.
The cult itself is told elsewhere; it is the subject of our piece on the Nerthus group. What concerns us here is the list of names alone. For it is the densest passage in the whole work: seven peoples in a single line, two of whom later made world history – the Angles gave England its name, the Varini survive in an early medieval law book – while others never surface again. Why one name survives and another does not is the question we pursue in the article on the Reudigni. Here we ask something different: how certain are we that we are reading the names correctly at all?
The Germania barely survived the Middle Ages. The one codex to which everything goes back is the Codex Hersfeldensis of the ninth century – and it is lost. We know of it through letters and descriptions: from 1425 Poggio Bracciolini negotiated for it with a monk from Hersfeld; in 1455 Pier Candido Decembrio saw a manuscript in Rome answering exactly to that description; in the same year Enoch of Ascoli brought such a manuscript to Italy. After that the trail breaks off. The codex was evidently dismembered, sold and scattered.
What remained were copies – and one stroke of luck. In 1902 Cesare Annibaldi found a composite manuscript in the private library of the Counts Balleani at Jesi near Ancona: the Codex Aesinas (now Rome, Biblioteca Nazionale Centrale, Cod. Vitt. Em. 1631). Bound into it were eight leaves in Carolingian minuscule, which Bernhard Bischoff dated to the middle of the ninth century and placed in the West Frankish region of the Loire. They are generally taken to be remnants of the lost Hersfeldensis itself.
Here lies the point that popular accounts almost always miss. The Carolingian leaves of the Aesinas (fols. 56–63) contain not the Germania but the Agricola – roughly from chapter 14 to chapter 40. The Germania stands on fols. 66–75, and it is written entirely by the hand of the humanist Stefano Guarnieri, probably in the 1460s. Guarnieri took care to imitate the older script; the result looks venerable, but it is a fifteenth-century copy.
For our question this means that there is not a single witness to the Germania from before 1400. All we have are humanist copies which derive from one another, partly directly and partly through intermediaries. How many exactly cannot responsibly be reduced to a figure – there are several dozen, and their relationships are debated to this day. Annibaldi published a facsimile of the Agricola with collations in 1907 and a diplomatic edition of the Germania in 1910; Rodney Potter Robinson produced in 1935 what remains the fullest critical edition.
It is worth pausing to explain the tools philologists use. A reading (Latin varia lectio) is what actually stands in a given manuscript, whether or not it makes sense. A conjecture, by contrast, is a modern scholar’s proposal: a form found in no manuscript but intended to explain the nonsense that is transmitted. An emendation is the admission of such an improvement into the printed text.
Which reading stands where, and what editors have otherwise proposed, is recorded in the apparatus criticus – that small-print band at the foot of the page which almost nobody reads and in which the actual work resides. Finally there is the rule of thumb known as lectio difficilior: of two readings, the harder and stranger one has, other things being equal, the better claim to be original, because copyists smooth things out and rarely complicate them. A codex descriptus, lastly, is a manuscript demonstrably copied from another surviving one; it carries no independent weight for establishing the text.
Applying this grid to Germania 40 produces a clear picture. The text of the manuscripts reads Suarines. In the margin of a single manuscript stands Suardones. So this is not a case of pure conjecture – nobody invented Suardones at a nineteenth-century desk. It is the case of a marginal reading that found its way into the main text.
That matters, because it makes the editors’ decision more intelligible than it first sounds. They did not guess. They chose between two transmitted forms – only against the majority and against the main text. Whether that choice was right is a separate question.
A layman would say: what stands in the text counts. Scholarship takes a more differentiated view, and for an explicitly stated reason. The marginal notes in the Aesinas are, by general agreement, contemporary with the copy – and, as several students of the manuscript have recorded, they are often better than the main text precisely where proper names are concerned. The observation is found in Jacques Perret (1950), in Erich Koestermann’s Teubner edition (1964) and in Charles Murgia (1977).
Palaeographically this makes sense. Proper names are the most dangerous spot a copyist can meet: he does not know them, he cannot infer them from context, he cannot correct himself. Someone who later collates an exemplar and notes a divergent form in the margin is doing exactly what an attentive corrector should do. A marginal reading attached to a tribal name is therefore not a curiosity but an argument to be taken seriously.
Anyone who looks at the collations of the Germania soon notices that Suarines is no isolated case. Practically every ethnic name in the work fluctuates. The familiar Dulgubnii appear as Dulgicubuni, dulgitubini, Dulgibini; the Chasuarii as Thasuarii, tasuarii, occasuarii. The Naharvali are consistently Nahanarulos and the like in the manuscripts; the current form is Karl Müllenhoff’s restoration. The Lugii appear as Legiorum and leugiorum, the Cotini as Gotini, the Naristi as Narisci or maristi, the Lemovii as lemonii.
Two opposite conclusions can be drawn from this. One: the names are hopelessly corrupt and every map is fiction. The other, soberer one: names are damaged in copying because they are foreign bodies in the sentence – and that is precisely why textual criticism developed rules for handling the damage. The Marsigni, who appear exactly once in chapter 43, are a related case; readers who want to follow that will find them in the article on the Marsigni.
How did Suardones become canonical? The trail leads into the nineteenth century. In 1837 Johann Kaspar Zeuss published Die Deutschen und die Nachbarstämme, the work that first treated the ethnic names of the ancient authors systematically from a linguistic standpoint. Karl Müllenhoff built on it in his Deutsche Altertumskunde. And Rudolf Much wrote the entry for the first edition of the Reallexikon der Germanischen Altertumskunde (vol. 4, 1918/19) – under the headword “Suardones”.
That settled the matter, not through a new argument but through authority. A headword in a standard reference work is a set of points on a railway: anyone writing afterwards looks it up, finds “Suardones”, and writes “Suardones”. The second edition, incidentally, has reset those points. The entry in today’s Germanische Altertumskunde Online reads “Suarines/Suardones” – both forms, in that order. The correction has long since been made; it has simply not reached the places where most people read.
For that is exactly the mechanism by which an editorial decision becomes invisible. The standard English translation by Alfred John Church and William Jackson Brodribb (1877) prints Suardones. The freely available text collections online print Suardones – the Latin Library, for one, offers chapter 40 in precisely that form. And these digital texts are now the source for nearly everything: encyclopaedia articles, mapping projects, school essays, language models.
The decisive point is this: a digital text has no apparatus. What appears on the screen looks like “the text”, not like “one of several defensible decisions”. The footnote which would say that the manuscripts read something else was lost in the copying. That does not make the consensus wrong – but it makes it unopposed, which is a different thing.
The same chapter contains the most instructive comparison in the entire work. The goddess’s name also fluctuates considerably in the manuscripts: transmitted forms include Nertum, Nerthum, nethum, Neithum, Neherthum and Herthum. In 1519 Beatus Rhenanus opted for Herthum – an obvious choice, since it chimed with the German word for earth and Tacitus expressly glosses the goddess as Terra mater.
That decision held for roughly three centuries. Only in 1817 did Franz Passow restore Nerthum, and only Jacob Grimm supplied the decisive argument: Nerthus is, by regular sound change, exactly the ancestor of the Old Norse divine name Njörðr. In the meantime an editorial choice had produced a goddess who never existed. R. W. Chambers summed it up drily in 1912: the false goddess had for many years driven the rightful one out of the fortieth chapter of the Germania.
The consequences of that error are still written into the countryside. On the island of Rügen a lake and a hill fort bear the names Herthasee and Herthaburg – named after a goddess derived from a printed edition of 1519. Antiquarians of the eighteenth and nineteenth centuries declared the Herthasee the site of the Nerthus ritual, which was circular: the lake was called that because scholars called the goddess Hertha, and the goddess was located there because the lake was called that. Theodor Fontane wrote the place into Effi Briest in 1896; several sports clubs carry the name to this day.
Then came confirmation from outside. In 1902 the Codex Aesinas was found – and it read Nertum. Jacob Grimm had died in 1863; he never lived to see his choice of reading vindicated. John Lindow has pointed out how strong the coincidence is: a form preferred on linguistic grounds turned up barely forty years later in the best witness. That Lotte Motz in 1992 regarded the correspondence as accidental, and that Rudolf Simek followed her, belongs to an honest account – but it is not the majority view.
So within a single chapter we have both possible endings of textual work side by side. With Nerthus, a reasoned choice of reading was confirmed by a later manuscript find. With Hertha, a decision meant just as reasonably produced a phantom for three hundred years – one that named lakes and populated novels.
Where do the Suarines stand in this scheme? Soberly put: in between, and without prospect of resolution. There is no second witness to confirm Suardones and none to refute it. There is no mention in Pliny, none in Ptolemy, no inscription, no coin. Unlike the case of Nerthus, no manuscript discovery can help here, because outside this one line the word does not occur anywhere in antiquity.
The temptation is to settle the matter linguistically. Suardones can be attached to Proto-Germanic *swerdaz, “sword” – “the sword people” would be a plausible warrior name. Suarines could be connected with a root meaning “heavy, weighty, honoured”, or leaned against the name of the Suebi. Both sound good. Both are unproven.
The handbook of Old Germanic ethnonyms by Alexander Sitzmann and Friedrich Grünzweig (2008) therefore states expressly that the meaning and etymology of the name remain unclear – and for two reasons at once: the single attestation and the unstable transmission. Here lies the methodological trap. To choose the reading Suardones because it etymologises more attractively, and then to offer the etymology as confirmation of the reading, is to argue in a circle. How far etymology can carry when nothing else is available is the subject of our article on the Aviones.
There is a single attempt to approach the matter from outside, and it is itself another problem in textual criticism. The Old English poem Widsith lists in line 61: “Mid Seaxum ic wæs ond Sycgum ond mid Sweordwerum” – I was among the Saxons and the Sycgan and the Sweordweras. In 2013 Leonard Neidorf suggested that this name might be a corrupt form of the same people Tacitus has in mind.
The appeal is obvious: Sweordweras would mean literally “sword men” and would fit the reading Suardones. But this is not proof; it is a chain of circumstantial evidence with two weak links. First, the Old English name may itself have arisen by reinterpretation – a scribe who hears a foreign sound and makes a familiar word of it is the commonest source of error there is. Second, Widsith is poetry, and roughly half a millennium younger. That the same catalogue also names Varini, Angles and Jutes makes the neighbourhood interesting; it does not make it probative.
Two proposals exist for their location, and each hangs on the respective form of the name. The Reallexikon places the Suarines in northern Jutland, among the other peoples of the Nerthus list. The Danish scholar Gudmund Schütte suggested instead that the name lives on in the town of Schwerin.
That second proposal is no longer tenable, and one can show precisely why. Schwerin appears in the sources as Zuarin – Thietmar of Merseburg reports that the Obodrite prince Mistislav fled there. The form belongs to Polabian zvěřin, “game enclosure, animal park”, and carries the -in suffix typical of Slavic place names. The German refoundation under Henry the Lion took place in 1160. The name is therefore Slavic and roughly a thousand years younger than Tacitus. Archaeologically the matter is entirely silent: there is no site, no pottery group, no cemetery that can be connected with this people.
What remains is a single sentence, and it stands not in chapter 40 but immediately after it. Tacitus closes the whole list by classifying it:
Et haec quidem pars Sueborum in secretiora Germaniae porrigitur.
And this part of the Suebi extends into the more remote regions of Germania.Tacitus, Germania 41.1; translation: Glanz und Gravur
That is the most solid statement we possess about the Suarines: Tacitus reckons them, together with all six neighbours, to the wider Suebic body. About the internal order of that body he says nothing; about language nothing; about rule nothing. The Semnones and the Langobardi, whom he treats immediately before, belong in the same frame. One should not overstretch this: in Tacitus “Suebi” is a collective term, not a constitutional statement.
At the end stands an insight that reaches far beyond this one name. Every map of the Germanic peoples, every list, every entry in a work of reference is first of all an edition: the result of decisions somebody made and justified. Whoever says “Suardones” is not quoting Tacitus but Zeuss, Müllenhoff and Much. Whoever says “Suarines” is quoting the main text of the manuscripts. Both are defensible. Pretending that there is no choice is not.
The Suarines are therefore a better lesson than many a grand theme. They force the reader down to the small print – to the line that records what is actually transmitted and what somebody made of it. That is how we handle the Germanic peoples throughout our articles, and where there is doubt we name both forms. The Eudoses from the same list show that this is no isolated case. And Germanic religion as a whole can only be presented honestly in this way: by keeping the distance between source and interpretation visible.
The Aviones · The Eudoses · The Reudigni · The Nerthus League · The Angles · The Warni · The Suebi.
