Pirating Mare liberum (1609)

In: Grotiana
Mark Somos Max Planck Institute for Comparative Public Law and International Law, Heidelberg, Germany,

Search for other papers by Mark Somos in
Google Scholar
Dániel MargócsyDepartment of History and Philosophy of Science, University of Cambridge, United Kingdom,

Search for other papers by Dániel Margócsy in
Google Scholar
Full Access

Two pirated editions form a vital but neglected part of the printing and reception history of the first edition of Grotius’s Mare liberum.


Two pirated editions form a vital but neglected part of the printing and reception history of the first edition of Grotius’s Mare liberum.

* With many thanks to Hans Blom, Gonzalo Bustamante Kuschel, Alberto Clerici, Paul Dijstelberge, Ioannis Evrigenis, Erik Geleijns, Marina Garone Gravier, Gaby Mahlberg, César Manrique Figueroa, Leonidas Montes, Eric Schliesser, Kristof Selleslach, Steven van Impe, Stijn van Rossem, Rodrigo Villalobos Alvarado, Ken Ward, the two anonymous reviewers, the librarians at the institutions listed in the census, and the Alexander von Humboldt Foundation for generously supporting Mark Somos’ work.


Mare liberum has been called a landmark in the genealogy of capitalism, liberalism, imperialism, and a few -isms besides. 1 Aspects of its composition, publication history, political context and impact have been extensively researched. Given its importance, it is striking that existing scholarship neglects the fact that three distinct editions of Mare liberum were published in 1609, or at least carry that date. Nor has there been any attempt to estimate the editions’ print run, or gain insights into this seminal work’s reception by examining the surviving printed copies’ physical characteristics, such as their location, provenance marks, handwritten annotations, and the choice of other texts they were bound with. This article describes the three editions, provides a census of known surviving copies, and offers new insights into the political context and early reception of Mare liberum.

A brief summary of the extensive literature on relevant aspects of Mare liberum is sufficient here. The work, best known for arguing that all nations had a right to trade freely, grew out of a substantial revision of Chapter xii of De iure praedae commentarius (ipc), which Grotius probably composed between September 1604 and November 1606 to justify the Dutch seizure of the Santa Catarina, a Portuguese carrack of immense value captured by the Dutch East India Company (voc) in February 1603. The full ipc was rediscovered only in 1864, and first published in 1868. 2

In a 4 November, 1608 letter, the Zeeland voc directors commissioned Grotius to publish Mare liberum. 3 On 23 November, Grotius wrote to Daniel Heinsius, his colleague and friend, that he was working toward a ‘quick publication’ to transform Chapter xii of ipc into Mare liberum, and asked Heinsius to find a publisher. Lodewijk Elzevir in Leiden agreed to publish it, and subcontracted the printing to Joannes Balduinus. On 11 December Grotius accepted the proposed number of presentation copies, and asserted that the book must appear before February 1609, still in time to influence the negotiations that prepared the Twelve Years’ Truce. In his 18 December, 1608 letter, Grotius complained about the ‘neglect and idleness of the printer,’ who refused to fix a publication date or even guarantee printing before February. On 11 January, 1609 Grotius wrote to Heinsius that he was pleased about the typeface, but not the ‘ugly paper.’ Further exclamations of dissatisfaction with the publisher followed. 4

Grotius sent the preface and appendix to Heinsius on 18 February, 1609, and discussed arrangements to apply to the States of Holland for a privilege in order to discourage pirated editions. While the sea was free to everyone, the world of printing had to be regulated by monopolies. Due to the ongoing peace negotiations, on 7 March, Johan van Oldenbarnevelt, Land’s Advocate for Holland and effectively Prime Minister of the Dutch Republic, asked Grotius, his protégé, to suspend the publication of Mare liberum until the Twelve Years’ Truce was signed between the Dutch Republic, and Spain and the Southern Netherlands. Nonetheless, Mare liberum already appeared in the spring 1609 catalogue of the Frankfurt Book Fair. 5 The treaty was signed in Antwerp on 9 April, 1609. Probably at the end of April, Grotius wrote to Heinsius that the book was available for sale. 6

The Mystery Edition

Though for other reasons, Henk Nellen rightly notes that ‘[t]he history of the composition and publication of Mare liberum is a mystery.’ 7 In addition to Elzevir’s much-discussed Leiden version there exist two further editions with different dimensions, fonts, frontispieces, illustrated initials, and other variations. While the printing history, dating, and political context of the Leiden edition have attracted close scholarly attention, the existence of two non-Elzevir editions from 1609, let alone their distinctive features and the stories they tell, have been ignored. One non-Elzevir edition features briefly as item 542 in the 1950 bibliography of Grotius’s works by Ter Meulen and Diermanse. They in turn refer to Willems’s short and unsupported 1880 description of it as a pirated version of the Leiden edition. 8 More recent scholarship does not discuss this edition, except for a footnote that describes the volume as a pirated edition printed in Antwerp. This attribution is brief, unsupported, and contrary to the majority opinion of the experts we consulted, including the same expert who is cited in the Feenstra edition’s footnote. 9 We have also discovered a third edition in Chile, which has not been noted in the literature.

Historians have long emphasized the significance of the immediate political context for the publication of the original edition. Grotius’s intervention was a political act in the ongoing war between Spain and the Northern Netherlands, and Oldenbarnevelt’s intervention attests to his belief that the volume might have been read and taken seriously enough to jeopardise negotiations. The analysis of two other, possibly pirated, editions of Mare liberum may offer an opportunity to revise our understanding of its significance. If the two non-Elzevir editions were pirated, that suggests that even in its original context Mare liberum could serve purposes other than what scholars generally attribute to it. If, as we conclude, a Dutch printer is likely to have produced the better known mystery edition, it is even possible that he did so with the agreement of Grotius, who may have gone against Oldenbarnevelt’s will, and effectively supported a free market not only in maritime, but also in the printing trade. 10 If, however, this edition was printed in Antwerp, it raises the question of how it could have benefited printers and political actors in the Spanish Netherlands. And if the lesser known mystery edition, now in Santiago, Chile, was printed in Salamanca, it raises even more questions about the quick spread of the work to Spain, and the Salamanca school’s interest in it. If printed in Cologne, it highlights the book's German reception. Our analysis provides the first steps towards answering these questions.

A close comparison between the two editions that are noted in the scholarly literature shows several differences. Both editions are in octavo format, but neither the fonts nor the printed surface are the same. The Leiden edition is 66 pages of Grotius’s text, plus a two-page appendix of Philip iii’s letters to Martim Afonso de Castro, Portuguese viceroy of India until his death in June 1607. Its title page shows an eagle and the publication details, ‘Lugduni Batauorum, Ex officina Ludovici Elzeverij, ANNOCIƆ.IƆI.IX.’ The dedication is followed by an errata section, which Feenstra suggests might have been compiled by Grotius himself. 11 In this edition, page number 66 is misprinted as 42.

The mystery edition is 67 pages of Grotius’s text, followed by Philip iii’s letters on 3 pages. This edition uses a smaller font than Elzevir, and it is often printed on cheaper, inferior paper. Ter Meulen and Diermanse measure the Leiden edition as 15,5 x 9,5 cm, and the unidentified edition as 16,5 x 10 cm. The title page of the latter shows a fantastical lion head device with the text, ‘Impressa primum, Lugduni Batavorum, In officina Ludovici Elzevirij, ANNOCIƆ.IƆI.IX.’ 12 The design is repeated at the end of the book, unlike in the case of the Leiden edition. There is no errata, and the designs for the two illustrated capitals, E in the address to the readers, and P at the beginning of Chapter i, also differ from Elzevir’s. There is considerable variation in the use of ‘v’ and ‘u’ and ‘i’ and ‘j’ between the two editions, with no consistency in either. Page number 15 of this edition is misprinted as 13.

Although the mystery edition corrects the mistakes listed in the Leiden errata, it is not obvious that it was created by simply pirating the Leiden edition. Evidence gleaned from variations between the two editions, and their correspondence with the known manuscript, is inconclusive. There are cases when the mystery edition matches the manuscript better; and cases when the Leiden edition is closer. Importantly, both editions seem to have taken into account the midpoint of the text that Grotius erroneously established in the surviving manuscript, Ms. bpl 917 (sometimes assumed to be the draft from which the Elzevir edition was printed) in order to allow two typesetters, and perhaps two copyists, to work on the book simultaneously. 13 While variations and the attention to the midpoint do not exclude the possibility that the mystery edition is based on a manuscript, the force of these two sets of evidence is also undermined by the strong probability that the typesetters used cleaner copies, not Ms. bpl 917. Let us examine these arguments in turn.

Firstly, both editions diverge from the manuscript’s spelling and contraction conventions. While Grotius preferred ‘i’ to ‘j’ and ‘u’ to ‘v’ in cases where both were acceptable, both editions frequently contravene his preferences. Nor does either edition follow closely Grotius’s placement of commas in the manuscript, or his choices between et and the ampersand. When either edition prints the enclitic –que as q or q́, or uses a curved macron to indicate an omitted m or n (on the first page of the main text alone, regulã for regulam, cũcta for cuncta or for non), it is equally likely that it follows or departs from the manuscript. No pattern emerges from these features.

When Elzevir is Closer to Ms. bpl 917

Compared with the official, Elzevir edition, the mystery edition departs in another three ways from Grotius’s manuscript: it introduces ten new typographical errors, it abbreviates words when it tries to replicate the pagination and midpoint that Grotius calculated, and it italicises the few quotations that Grotius failed to mark as such. Variations will be noted in the following format: references to Ms. bpl 917 begin with f, indicating the folio number. L refers to the Leiden edition, and M to the mystery version, usually followed by a page number. Since the page numbers of L and M are often the same, lm is used to refer to words or passages printed on the same page in both editions.

Although the mystery edition corrects the mistakes noted in the Leiden errata, it also introduces new typographical errors. The address to readers misprints praesidum for praesidium, and vos for nos. L3 itinera becomes M3 itinrae, and L8 inter alios turns into M8 in alios. The last sentence of Chapter iv prints sequimur instead of sequitur (M13). The mystery edition misprints subiere as subire in the Virgil quotation, and insultavere as insultare in Ovid (M17). In the long Chapter v, the mystery edition omits an et (M26, quod aliae in the penultimate line), and it has inaturalis instead of naturalis in the penultimate sentence on M44. At the beginning of Chapter x, the mystery edition misprints su juris (M56) for sui juris (f114r, L54). In citing Aristotle’s division of agreements into those that are concerned with transporting merchandise by land and those concerning transport by sea, the manuscript and the Leiden edition both use devehit (f114v, L54), while the mystery edition prints vehit (M54). Finally, M65 misprints ratem as ratim in the Praetor citation.

The unidentified edition’s printer abbreviates igitur as g. in the first paragraph of Chapter v (M13). Other unusual abbreviations include v. for what both the manuscript and the Leiden edition have as vero (M16, 20, 26), n. for enim (M16), a. for autem (M17, 18, 22), and e. for etiam (M20). These unusual abbreviations are concentrated on M16-20, suggesting that they were part of the printer’s strategy to match the pagination of the Leiden edition, which was based on Grotius’s calculations in the manuscript. Other indications of this deliberate strategy, such as widening the gap between words, are discussed below.

Grotius underlined dozens of citations in the surviving manuscript. Both editions printed these passages in italics. On f103v Grotius forgot to underline a citation from Ulpian, and another from Labeo. The Leiden edition does not italicise them, and its errata does not note that they are citations (L23). The mystery edition italicises them in keeping with Grotius’s practice (M23). Though technically this is an instance when the Elzevir is closer to the manuscript, it suggests that the mystery version conforms better to Grotius’s intent, or that the mystery printer worked from a cleaner manuscript. The same thing happens in the next paragraph, when Grotius signals that he is citing Uti possidetis interdictum from Paulus, but contrary to his custom does not italicise the phrase (f103r). The Leiden edition adheres to this, but the mystery edition uses italics in accordance with Grotius’s general practice (M24). Again, on f105r Grotius makes it clear that he is citing the terms interdictum and de loco publico fruendo, but fails to italicise them. Neither does the Leiden edition – but the mystery version does (M28).

When the Mystery Edition is Closer to Ms. bpl 917

In some cases, the mystery edition matches the manuscript better than the Leiden edition. There is no need to repeat the Leiden errata here; it is sufficient to note that the mystery edition corrected the text by taking all the errata into account. In addition to the more than one dozen mistakes listed in the 1609 errata, Feenstra’s 2009 edition notes another twenty-eight discrepancies between the manuscript and the Leiden edition. 14 Many of these are printed marginal references to previous passages in Mare liberum itself, and they are absent from the manuscript. However, there are also several differences that neither the 1609 errata, nor Feenstra and the critical tradition of editing Mare liberum that he drew on, seem to have noticed. They are listed here together with a comparison with the mystery edition.

The Leiden edition’s argumnto (L4) is spelled correctly in the mystery edition. To draw an inference from Doneau, Grotius writes sequitur (f104r), which L25–6 misprints as sequuntur, but M25 corrects. Discussing the common use of the sea in Chapter vii, Grotius writes usum, qui … on f109v. The same words appear in the mystery edition, as opposed to usum quae in the Elzevir (L40). The Leiden version mistakenly prints ex ex on L45, which the mystery edition corrects to ex. None of these mistakes are in the errata, or in Feenstra’s notes. The printer of the mystery edition could have corrected these by carefully reading the Leiden edition, without access to a manuscript.

The pagination and spacing of words, however, are harder to explain, unless we allow that the unidentified printer might have worked from a manuscript, or had otherwise learned about Grotius’s preferences and instructions. As mentioned, the size of fonts and the print surface of the mystery edition differ considerably from the Elzevir. In the Elzevir edition, the address to the readers and the errata together take up 10 pages. Only 6 pages are needed for the same address in the unidentified edition, which has no errata. Nevertheless, the difference in total page numbers is minimal: the Elzevir is 66+2, and the unidentified edition is 67+3. How is this possible?

The unidentified edition tries very hard to approximate the Elzevir. Even though the page and font sizes differ, the typesetter tweaks the layout to make the first 3 pages of the main text begin and end on the same word, even though the lines often differ (e.g. L3 line 3 ends “que-” in Elzevir but hinc in the unidentified, in which querelae shifts to line 4). The size differences force a bifurcation in pagination after page 4. By page 9, the end of Chapter iii, the unidentified edition squeezes in 4 lines more than the Elzevir. Lines begin and end differently now. Yet something strange happens. The printer keeps adjusting margins and columns, and by page 23 the difference between the two editions is reduced to 2 lines. By page 32, the two editions are almost perfectly aligned again, with only one line’s difference at the top. Now the divergence resumes, but the unidentified printer starts to plan ahead. By page 36, the beginning of Chapter vi, it is the unidentified edition that is 2 lines shorter than the Elzevir. Here the printer relaxes, and a mere 2 pages later the unidentified version runs ahead again. Long pauses between words are added, and an alternative letter “m,” one with a long tail, first introduced on M13, comes into increasingly frequent use. But the unidentified version pulls ahead, and by M44 it is more than half a page longer than the Elzevir. Spatial correction resumes, and by page 52, the beginning of Chapter viii, the Elzevir is 5 lines ahead. This advantage remains. Elzevir ends on L66 (misprinted as 42), and the tapering text reaches the bottom of the last page. The unidentified edition runs over slightly into M67, and the last page is filled up with a reproduction of the lion’s head ornament.

Why this extraordinary attention to the Elzevir edition’s pagination and length? A possible explanation might be that the attempt to match the Elzevir edition is an epiphenomenon or false correlation, because both editions are following a manuscript and/or detailed printing instructions by Grotius or someone else who participated in the production of the Leiden Elzevir. When Grotius prepared Chapter xii of ipc for the press as Mare liberum, he divided the text into two parts. This allowed two typesetters to work simultaneously, and perhaps two scribes to prepare a fair copy. Interestingly, Rabbie points out that Grotius’s manuscripts for Meletius (1611), Ordinum pietas (1613), De satisfactione (1615), and De imperio summarum potestatum circa sacra (1616) were also copied for the press simultaneously by two scribes, usually, but not always, Leiden students. 15 Could this arrangement have started with Mare liberum? In any case, as Ittersum has shown, to determine the midpoint of Mare liberum, Grotius counted the lines in Chapter xii of ipc in the manuscript, added the new conclusion on an inserted bifolium, noted the number of lines on each page, added them up (erroneously) on f121v, and determined the midpoint as nisi ivissent eo on f107v of the surviving manuscript. Though the printer may have redone the calculations, this midpoint shaped the production process. 16 Despite the considerable size differences and fluctuation in comparative alignment, the two 1609 editions are quite well matched at Grotius’s midpoint. It is on L34, line 21 of 29 lines in the Elzevir, and M34, line 19 of 29 lines in the unidentified version. The alignment of Grotius’s midpoint, and the consistent effort to correct for the size differences, raise the possibility that both editions were produced from a manuscript. If so, this need not have happened with Grotius’s approval. After all, Petrus Bertius had an unauthorised copy of Ordinum pietas made in 1613, while it was being copied in preparation for the press. 17 Johann Boreel, a close friend of Grotius who saw Mare liberum in manuscript and later advised the Zeeland voc to encourage Grotius to publish it, exemplifies the small circle of politically engaged friends who would have been in a position to arrange for an unauthorised edition based on a manuscript copy, or from the Elzevir edition but with Grotius’s instructions.

It is also worth noting that the mystery printer did not set out to produce copies that would fool people into thinking that they were buying the Leiden edition. The mystery Mare liberum (which, unlike the Elzevir, calls itself the first edition on its title page) has different ornamental devices and pagination, which make it easy to distinguish it from the Leiden version. An attempt to produce an identical edition was not the reason why the printer paid close attention to Grotius’s rough calculation for the pagination, and to the midpoint.

This raises the further possibility that Ms. bpl 917 is not the final version from which fair copies were made for the printers. This manuscript has several messy and hard-to-follow parts, which would have made the work of copyists, let alone typesetters working directly from it, very difficult indeed. The hypothesis that at least one cleaner copy followed Ms. bpl 917 before the copyists took over is reinforced by differences between this manuscript and both printed editions that cannot be traced to the manuscript. All the differences noted in the next section are inexplicable without interim drafts between Ms. bpl 917 and the copy given to Elzevir and Balduinus, and an interim draft or instructions to observe the Elzevier edition’s length and midpoint having been given to the mystery printer. 18

When Both Editions Diverge from Ms. bpl 917

On f96v Grotius describes the principle in the law of nations that any state can trade with any other as perpetua atque immutabilis, but the first page of the main text in both editions prints it as perspicua atque immutabilis. 19 The word perpetua is clear, legible, and unamended in Ms. bpl 917, but it is not what appears in print. On the same page, a question mark at the end of a question on f96v is moved up to the interrogative, Quo ista? in print. On f98r Grotius uses another question mark to ask how the Portuguese could have discovered India, when Horace already mentions it; the printed editions replace the question mark with a full stop (lm6). On the same page Grotius spells idolatrae, while both editions print idololatrae. On f98r, at the end of Chapter ii, Grotius cites Vitoria on the Indians not being amentes aut insensati, which becomes amentes et insensati in both editions (lm7). In Chapter iii, when according to both printed editions Grotius disputes that the Portuguese are lords of all those places (dominos eorum locorum, lm8), the manuscript has dominos orienta instead (f98v). In Chapter iv, Grotius considers whether it is lawful to attack those who refused to convert even after learning about Christianity. The printed editions have respicere (lm11) while the manuscript has recepere. Next, Grotius inserts a long quotation from Cajetan, in which subiciendos eos (f99r) becomes subjiciendos illos in print (lm12). On f102r, Grotius cites from Seneca’s Octavia as exstruxit urbes; lm18 omits a letter and changes capitalisation to print Extruxit urbes . When he argues that occupation may not hinder common use in things designed by nature for common use, Grotius writes nam cum ita se res habet (f103r); the res disappears in both editions (lm21). Discussing the principle of uti possidetis, on f104r Grotius rewrites solam iniuriae actionem to … iniuriarum …, but only the printed editions change solam to solum (lm24). On the same page, f104r, Grotius cites Doneau’s position that the sea is not tradeable, and cannot be properly owned. Both editions print Unde sequitur si proprie loquamur (L25–26, M25), even though in the manuscript Grotius replaced this phrase with, Unde censent eruditiores si proprie et recte loquamur. 20

The phrase, ne illum quidem on f105r becomes ne quidem illum in both editions. Similarly, ita mare illud on f106r becomes sic mare illud in print (L31, M30). On f106v, when Grotius marshalls ancient sources to show that the Portuguese did not discover new routes, he writes quod Cornelius Nepos tradidit. The two editions change the last word to testatus est (lm32). In Chapter vi, Grotius points out that the Pope is not the temporal lord of the whole world in the opinion of all men of sound judgment (ex omnium sani judicij hominum, lm37) – a phrase that is missing from the manuscript (f108r). Grotius rejects Angelus’s position that even in things that are common by the law of nations, prescription can confer dominion on the possessor (etiamsi ad dominium praescriptio proficere non potest, tamen dandam esse possidenti exceptionem, lm41). However, the manuscript gives the conditional, with posset instead of potest, and fore instead of esse (f110r). On the same page, the printed versions omit res from the manuscript’s correct citation of Vázquez’s phrase, tale jus cessat res agitur. Still in this long citation, Grotius’s manuscript correctly gives est ut a nemine possit (f111v), while the printed version drops the a (lm47).

This is a minor omission compared to the next sentence, where both printed versions cut praetextu praescriptionis, cum impedienti id minime profit, which Grotius’s manuscript cites faithfully from Vázquez. In another long citation, this time from the Digest, Grotius gives terram vel flumina (f112r), which both editions render as terram et flumina (lm48). On f113r posse quidam becomes quidam posse in both editions (lm50). When Grotius writes Ex occasione … (f113r), lm51 merely has Occasione. The Latin translation of the Greek passage from Aristotle’s Politics in Chapter viii is underlined in the manuscript (f113v, note F), but both editions fail to italicise it (lm52). Describing the primitive condition, Grotius uses the phrase, nec adhuc vero tamen permutatio erat on f114r, from which both editions redact vero (lm53). On the same page, the print versions drop autem after dictus, before Grotius’s invocation of an Aristotelian term. Both editions add usque to the penultimate sentence of Chapter xi (f115r, cf. lm59), and both omit esse after quisque malit in Chapter xii (f116v, lm60). The printed versions supply a question mark after societate ferendumne est that is missing from the manuscript (f116r, lm61). Grotius forgot to underline a quotation he attributes to Demosthenes on f117r, but both editions italicised it (L63, M64). 21

The Mystery Printer

So far we established that when the mystery and Leiden editions depart from the manuscript, they do so mostly in the same ways. When they diverge from the manuscript in different ways, the mystery edition seems to postdate the Leiden version. It incorporates the Leiden errata, makes additional corrections that do not need a manuscript source, and introduces a few typographical errors of its own. However, the printer’s attention to matching the Elzevir pagination and implementing Grotius’s miscalculated midpoint, despite the printer’s disinterest in producing copies that could be mistaken for the Elzevir, suggest that the mystery edition might have been based on a manuscript originating from Grotius, or at least with a knowledge of Grotius’s instructions.

As mentioned, a scholarly reference to the mystery edition describes it as a pirated version printed in Antwerp; but the literature offers no evidence. Extensive consultation with experts on seventeenth-century printing yielded no conclusive proof either, but it did reveal that there is in fact no solid reason for the Antwerp attribution. A rare book curator noted that in Antwerp, V and I would have been more probable choices for printing the title than the U and J that feature in the mystery edition. An archivist kindly advised us that the E and P in the mystery edition belong to an arabesque series that was in wide use from the 1590s in Antwerp, but also in the Dutch Republic.

There are good reasons not to expect certainty when using physical characteristics to identify the printer. Another colleague pointed out that printers borrowed and swapped stylised initials, such as Mare liberum’s E and P, therefore a simple match between Mare liberum and other books cannot provide definitive identification. However, a match could at least identify the town where the mystery edition was printed. Based on stylistic considerations, most experts and rare book curators feel that Amsterdam is the most likely place of printing. One specialist suggested the workshop of Jan Janssonius, given his use of cheap paper and notoriety for pirated editions. However, there were two printers by this name, father (?–1629) and son (1588–1664). The father, based in Arnhem, was appointed official printer to the state of Gelderland in 1599. 22 The son moved to Amsterdam in 1612, and enrolled in the Amsterdam guild as a bookseller only in 1618, too late for the unidentified edition. At the same time, Janssonius Sr. owned the lion head’s device and the stylised E and P, and in 1609 Janssonius Jr. was old enough to engage in the printing practices he later became known for.

Our next step was to suspend the secondary literature, which effectively amounts to a dubious footnote, and try to identify the printer from distinctive features of the mystery edition. This study focused on three such features: the stylised capitals E and P, and the lion’s head ornament (figs. 1–3). The obvious starting point is publishers that Grotius and Heinsius already had a relationship with. By 1609, Heinsius published with four Leiden presses: Jacobs Jan Paets, Joannes Maire, Andries Clouck, and the Plantin branch run by Raphelengius. 23 Heinsius also published in Heidelberg, with Verdussen in Antwerp, and with Dirck Pietersz in Amsterdam. 24 Grotius published short pieces or book-length works in Leiden with the Plantin branch run by Raphelengius, Guyotius, Thomas Basson, Andries Clouck, Jacobs Jan Paets and Jan Claesz van Dorp in Leiden, at the Hague with Meerman, Cornelis van der Nieustadt, Jacob Hillebrand, Beuckel Cornelisz Nieulandt and Albert Henrick, and in Paris and Heidelberg. 25 As far as we can establish, none of these printers used the lion’s head ornament or either the E or P initial before 1610. The next step was to test the hypothesis that matching distinguishing features of the mystery edition can help identify the printer, then cast a wider net, and focus on books printed between 1607 and 1611.

Figure 1
Figure 1

The title pages

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

Figure 2
Figure 2

Stylised initial E from the address to readers

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

Figure 3
Figure 3

Stylised initial P from Chapter i

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

Like Mare liberum, Heinsius’ extraordinary 1611 edition of Aristotle’s De poetica also appeared with Elzevir, also subcontracted to Balduinus. It features the same eagle device (fig. 4). 26 Though the Leiden Mare liberum has only two stylised initials, and De poetica has only four, one of the letters, P, is the same, allowing for a close comparison (fig. 5).

Figure 4
Figure 4

Title Page of ed. Heinsius, Aristotle’s De poetica

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

Figure 5
Figure 5

Stylised capital P from Elzevir’s 1609 Mare liberum and 1611 De poetica

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

If it makes sense to assume that the same printer would use the same devices and capitals, then the task is simple: use the unidentified Mare liberum’s features to identify the printer. This is difficult to do conclusively, as there are only three salient distinguishing features, namely the stylised capitals E and P, and the lion’s head ornament. Only the P appears in Adam Sartorius’ printing of the Jesuit professor Matthias Mairhofer’s Calvinische Andacht Bezogen … (Ingolstadt, 1610). We found the same P in the first edition of Althusius’ Politica, printed in 1603 in Herborn by Christophorus Corvinus, in the second edition of Johann Pincier’s Catechesis religionis Christianae … also printed by Corvinus in 1603, and in William Whitaker’s Praelectiones, printed by Corvinus in 1601; but neither the E nor the lion’s head device. The 1601 Whitaker, Pincier and Althusius volumes employ a range of stylised initials with different designs, and Corvinus may not have had the full set at the time. However, there is a 1603 Herborn printing of Whitaker’s Praelectiones without the printer’s name, which has the same P; and a 1609 version of the same edition, which has both the P and the E. Though the printer’s name is absent, the printer’s device on both title pages belongs to Corvinus. In other words, Corvinus in Herborn certainly had both the E and P in 1609 when he printed Whitaker’s Praelectiones, which Grotius cites in the 1613 Ordinum pietas and later in De imperio.

The lion’s head device, without the E and P, appears in Georg Widmanstetter’s printing of Außschütt: und Steüberung …, a furious piece of anti-Protestant polemic by the probably pseudonymous Cleophas Distelmayer (Graz, 1608). The same device features in the Cologne printer Arnold Quentel’s 1599 and later editions of Jodocus Coccius’s Thesaurus Catholicus, and in Quentel’s 1603 Opera omnia of Ephrem the Syrian. Quentel’s 1603 printing of Prognosticon futuri status ecclesiae by Johann Paul Windeck contains only the E (pp. 147, 331, 397), similarly to his 1604 edition of Windeck’s De theologia iureconsultum (pp. 55, 153). However, Quentel’s 1603 printing of Windeck’s Controversiae de mortis Christi efficacia … contains the lion’s head device, the stylised P (pp. 6, 111), and the E (pp. 16, 17, 45, 130, 190), as well. Quentel is a strong candidate as the mystery printer, but a few details cast doubt on this identification. Quentel’s type, consistent through these books, is unlike the mystery Mare liberum. For instance, his capital Q has a longer tail, and the words set in all capitals hover in the line slightly above the uncapitalised words.

We also found only the lion’s head emblem in Verdussen’s 1606 edition of Laurens van Haecht Goidtsenhoven, Chronycke van de Hertoghen van Braband; in Verdussen’s 1607 printing of Medici Hispania … by Ludovicus Nonnius, an Antwerp physician and antiquarian of Portuguese descent and Rubens’ friend; in the 1608 Den wtersten wille van Lovvys Porquin; and in Verdussen’s 1608 printing of Aubert Le Mire’s edition of the Chronicon of Sigebertus Gemblacensis. It further appears in Janssonius’s 1617 edition of Ptolemy’s Geography. This volume, printed in Arnhem, also contains stylised initials with the exact same design and background as the E and P in the mystery edition. 27 Although we were unable to confirm that Verdussen used the E and P initials that appear in the mystery Mare liberum, the initials D, G and I from the same series are printed in Medici Hispania. This is of course not conclusive – recall that the initials in the Elzevier edition come from two different series (figs. 2 and 3), which at least raises the possibility that Balduinus did not have complete sets.

Evidence from Grotius’s life lends additional circumstantial force to the Janssonius scenario. As is often recounted, from 1632 Janssonius, Jr. competed fiercely with Blaeu and produced cheap editions of several works by Grotius, including Sophompaneas and De iure belli ac pacis. 28 In item 544, Ter Meulen and Diermanse describe the 1632 Mare liberum that Janssonius, Jr. printed in Amsterdam. However, as early as 1625, when the first edition of De iure belli ac pacis appeared, Grotius requested that copies be sold through Janssonius’s shop in Amsterdam. 29 Grotius seems not to have had a comparable relationship with the Verdussen family.

As mentioned, Heinsius and Grotius discussed arrangements for a privilege for the Elzevir edition of Mare liberum. In the end they did not obtain one, so the printer was effectively free to create his own version. Janssonius’s business practice of issuing rival editions on cheap paper, the absence of a privilege, and Grotius’s interest in working with Janssonius in 1625, do not constitute decisive evidence, but together with the experts’ opinion and the mystery edition’s distinguishing features they tip the balance of probability from Antwerp to Arnhem or Amsterdam as the place of publication. 30 At the same time, one should note that Verdussen came to be known for specialising in works in Spanish and Portuguese, and also for publishing for the Spanish and Portuguese book markets. For instance in 1610, the year after Mare liberum, Verdussen published Relaciones del origen, descendencia y succession de los reyes de Persia y de Harmuz by Pedro Teixeira, an explorer probably from a Portuguese Jewish family, who set out from Lisbon on expeditions to Goa, Ceylon, Malacca, Borneo, Manila, Mexico, Baghdad, Aleppo, and Venice, before he settled down in Antwerp to write.

Though it is not possible at this stage to settle the printer’s identity with certainty, the relative probabilities of the possible printers lead us to another question, namely the motivation for publishing the mystery edition. One scenario is that the province of Gelderland encouraged Janssonius, their official printer, to produce an edition. An alternative explanation of the Janssonius scenario is that the unidentified edition was instigated by someone privy to Grotius’s composition of Mare liberum (or even by Grotius himself) in order to stop the Dutch and perhaps the French delegates, led by Pierre Jeannin, from compromising on trade. In this case, they might have sent the manuscript to Janssonius or another Dutch printer in hope of a quick turnaround. It would also explain the mystery edition’s layout, length, and attempt to adhere to the Leiden edition’s pagination. Another possibility is that Catholic authorities may have encouraged Verdussen or Quentel to publish a version as quickly as possible, perhaps to sabotage negotiations by revealing the Dutch strategy to Spain (exactly what Oldenbarnevelt intended to prevent). In Verdussen’s case, Antwerp would have also been interested in alerting the Flemish audience to the threat of the Dutch claim to free trade, while closing the Scheldt to navigation. In Quentel’s case, authorities would have preferred to sabotage the Twelve Years’ Truce in the wake of the 1583–88 Cologne War, which reestablished the city’s Catholic allegiance. Moreover, Cologne’s own rise as a member of the Hanseatic League and a long-distance trade centre on the Rhine made Mare liberum an attractive justification of the right to free trade, possibly creating interest in a locally produced and accessible edition.

The political context of the pirated edition, and indeed of Mare liberum in general, cannot be settled with certainty without confirming the place of publication. Whether it was printed in Arnhem or Cologne makes a great difference. The number of surviving copies and their dissemination may tell us more.

The Census

The next promising line of inquiry concerns the print run and early reception of the two editions. Universal catalogues, such as Worldcat and the Karlsruher Virtuelle Katalog, and national union catalogues, are not wholly dependable guides. They often fail to recognise the difference between the two editions, and simply copy and paste the details of the better known Leiden version. In the course of this research, several copies recorded with Elzevir characteristics turned out to belong to the mystery edition. Many of the union catalogues’ and even the library-specific catalogues’ online entries that specified original 1609 printed editions have turned out to be facsimiles (e.g. in the Vatican Library, the Finnish Parliament’s library, at the University of Antwerp, at Whitman College in Pennsylvania, at King’s College, London, and at Dijon-bu). However, the questionnaire that we sent to all institutions that hold a copy has also uncovered uncatalogued copies, including a mystery edition at the Harvard Law School.

We located as many copies as possible using a range of online tools. Then we sent a questionnaire to the libraries and institutions that own copies, asking whether their exemplar belonged to the Elzevir or the unidentified edition, whether it was bound with other works (and if so, which ones), whether it contained annotations or added materials, such as the author’s portrait, whether they had provenance information either directly from their copies or from library records, and whether they could photograph for us the two stylised initials, so we could confirm the edition and perhaps narrow down the possible printer based on woodblock damage. Searches in universal and national catalogues, scanned card catalogues, online databases of rare book dealers and auction houses, and the questionnaire sent to libraries, together led to a total of 78 copies, including both editions, as well as the Santiago copy discussed separately below. From the 78 total, 52 copies belong to the Leiden edition, 23 to the mystery edition, 2 are lost (Germany, Greifswald ub 542/Pa 705 adn1 (Vermißt 1945); Munchen J.publ.e. 237), and one, the Santiago copy, is unique (more about this below). Setting the unverifiable and exceptional 3 copies aside, the mystery edition counts for more than 30% of the surviving copies. It is not a rare curio, but an important and unduly ignored part of the dissemination and reception of Mare liberum in and after 1609.

The process of locating copies and estimating the print run of each edition is complicated by the fact that many surviving copies are bound in Sammelbände on nonce volumes, collections of several shorter works bound together. The survival of early modern printed texts depends on numerous factors, but in terms of disappearing uncatalogued in a nonce volume, Mare liberum has a higher risk than a full-length book.


The distribution pattern from our census, confirmed by provenance marks and early handwritten marginalia, shows the highest concentration of copies of both the Leiden and the mystery editions in German towns, and especially north German ports (see Figures 6 and 7). Most of these copies were acquired in the early seventeenth century by municipal authorities. This finding raises the question what the function of Mare liberum was in Germanic lands, which were not major players in early modern globalisation. While our maps confirm that Grotius’s work was much appreciated in the Netherlands and on the British Isles, where it could be used to justify these countries’ colonial endeavours, the maps further reveal that the Mare liberum editions were also read by smaller Germanic towns that could only aspire to maritime trade, but did not actually have a major colonial project at the time. Grotius’s book therefore was not simply political propaganda for colonial powers, but also a political work that opened up the dream – not to be realised – for states of the Holy Roman Empire to engage in free trade across the globe. The north German concentration may also be a sign of Hanseatic towns’ interest in new justifications of their right to trade, and in the success of their Dutch colleagues and competitors. The fact that two thirds of the surviving copies of the mystery edition are in German libraries, and no copies of this edition are preserved in the Netherlands, Belgium, France, or the Iberian peninsula, gives further support to the hypothesis that Quentel in Cologne was the printer.

Figure 6
Figure 6

Elzevir editions in Europe

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

Figure 7
Figure 7

Mystery editions in Europe

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

By contrast, we found very few copies in Spain and Portugal, probably because the book was placed on the Index in 1610. Given the Iberian reception of Mare liberum, the distribution of surviving copies leaves open questions concerning the means through which Iberian authors accessed this work, and the arrangements they had for owning and holding this forbidden book. The mystery edition in particular is much more scarce in Catholic countries than the Elzevir. Given the trend that most copies stayed where they were initially purchased, this strengthens the probability that the better known non-Elzevir edition was not printed in Antwerp, nor by a Catholic printer who either tried to profit before Mare liberum was placed on the Index (if the 1609 date of publication on the title page is correct), or to circumvent the papal prohibition (if the date of publication is false). However, the prevalence of the mystery edition in Germany suggests that Quentel in Cologne may indeed have been responsible for the printing.

The Santiago Copy

The questionnaire uncovered an extraordinary copy in Santiago in the Biblioteca Nacional de Chile, call number Sala Medina, B1,T27(7). As far as we know, it is unique. It is also intriguing in many ways. It is in folio, measuring approximately 28 x 20 cms, making it almost four times the size of the other two editions. Its decorative features compound the puzzle of its printing history, but it was probably pirated from the Leiden Elzevir by a printer who had no access to Greek type. It is condensed to fewer pages, introduces over 70 typographical errors, omits some of the text, and alters spelling in several ways that suggest orthographical conventions foreign to the other two editions’. This version also changes capitalisation to show greater contempt for popular sovereignty and republics, and greater respect for lords and theologians, than Grotius’s manuscript and the other two editions. Let us examine these conclusions in turn. Variations will be noted according to the following format: L refers to the Leiden edition, M to the mystery version, S to Santiago.

Like the other two editions’, the Santiago edition’s title page claims that it was published in Leiden by Elzevir in 1609 (fig. 8). However, the printer’s device belongs to the Gryphius family. The dynasty began with Michael Greyff (1445/1450–1512), who learned printing in Strasbourg, and opened his workshop in Reutlingen. His son, Sebastian (c. 1492–1556) learned the craft from his father. In addition, he also studied printing in Venice before he settled down in Lyon, acting on behalf of Venetian booksellers. Noted for his editions of Erasmus, Budé and Poliziano, Sebastian set up his own printing house in 1536 with the well-known gryphon mark and the motto, Virtute duce, comite fortuna. His heirs continued his Lyon imprint, and published a string of famous editions of classical and contemporary authors under the collective name, apud Haered. Seb. Gryphii. Sebastian’s son, Antoine (1527?–1599) also published in Lyon under his own name. Moreover, Sebastian had two brothers: François, who became a printer in Paris, and Johann, who published books in Venice. Sebastian, and his heirs, who were still publishing in and after 1609 in Lyon, used Greek type in several books. (So did Antoine, and Johann in Venice.)

Figure 8
Figure 8

Santiago title page, with kind permission of the Sala Medina of the Biblioteca Nacional de Chile

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

Several versions of the gryphon device were in circulation, but it remained one of the most recognisable imprints in Europe. 31 Both Sebastian and his heirs used the exact design that we find on the Santiago edition, except for the border. 32 Sebastian, Antoine, and Sebastian’s heirs in Lyon, and Johann in Venice, did occasionally use borders around the same gryphon device, but with different designs than the one in the Santiago copy. 33 It is as unlikely that Elzevir used the famous gryphon as a Gryphius printer crediting Elzevir, and not using Greek type. The most probable explanation therefore is a piratical printer, who almost courted discovery. Why combine Elzevir’s name with the Gryphius device otherwise? Another possible explanation is that the printer pirating the copy now in Santiago, with its eliminated spaces, no Greek type and over seventy new errors, confused the two Lugduni, and matched a Lyon device to the Leiden name. One of the experts on early modern printing who kindly advised us suggested that the book may have been printed in Salamanca by the heirs of Juan de Junta, who also ‘borrowed’ the Gryphius device. 34 This is an exciting possibility, given that Grotius’s engagement with the school of Salamanca, including his extensive citations from Salamanca scholars in Mare liberum, has long been one of the most debated aspects of his thought. Since Mare liberum was placed on the index of prohibited books by January 1610, a Salamanca printing in 1609 could have been motivated by state interests and/or the printer’s hope for high sales, while a later, fully piratical printing could have only been produced for commercial reasons while risking legal liability. At the same time, placing Mare liberum on the index in 1610 could have motivated piratical printers to forge not only the printer’s name and place of publication, but the date of publication, as well. After all, there is no independent verification that the two unauthorised editions were in fact printed in 1609.

The two stylised initials differ from both the Leiden and the Arnhem/Cologne editions, and they are not from the same series, but they are also not obviously useful for identifying the printer (fig. 9).

Figure 9
Figure 9

Stylised capitals, Santiago copy

Citation: Grotiana 38, 1 (2017) ; 10.1163/18760759-03800002

The Santiago version counts the folios, not the pages, and it runs to 25 leaves, i.e. only 50 pages. The first, unnumbered folio shows the printer’s device and the table of contents. The address to readers is printed on the second and third folios, also unnumbered. Chapter i begins on the fourth folio, which has “Fol. i” printed in the recto’s upper right margin. The fifth folio carries the number 5. Subsequent folio numbers run sequentially to 25, except folio 14 is misnumbered as 10, and folio 15 as 12. Grotius’s miscalculated midpoint phrase, nisi ivissent eo, is just above the middle of S12r, i.e. on page 23 out of the 50.

In addition to cutting the Leiden edition’s 68 and the mystery edition’s 70 pages down to 50, the Santiago copy shows further signs of compression. Spaces between words and sentences are often eliminated entirely. Citations in the Leiden and mystery editions are reduced by one to three lines. This is accomplished by moving connecting terms, such as Item that joins two lines from Virgil, from a separate line into the same line as the quoted verse (Chapter i, S4v); by moving phrases, such as Et Horatius that introduce a citation (in this case, to corroborate Cicero), which other versions print in a separate line, into the same line as the citation; or by placing citations that are printed in two lines in other versions into a single line (e.g. Avianus and Seneca passages on S9r). Quod often becomes ȹ, tamen turns into tñ. Occasionally less obvious abbreviations are used, for instance domini becomes dñi (S16r, cf. 108v, lm37).

Surprisingly, the printer of the Santiago copy had no access to Greek type. Accordingly, the dozens of Greek citations that appear in the manuscript and the other two editions are transcribed into Latin (S6v, S9v, S12r, S20r, multiple passages on S20v, S21r, S23r, S23v, S24r). Interestingly, the three dates at the end of Chapter vii (1477, 1595 and 1519) are spelled out and printed with letters on f113v and lm51, but S20r uses Arabic numerals. As in the case of the other two editions, punctuation marks and the use of ‘i’ or ‘j’, or ‘u’ and ‘v’, do not reveal a meaningful pattern. They are as likely to be identical with the Elzevir’s as with the mystery edition’s. Hyphenation at the end of lines is often lacking and slapdash throughout.

Several features suggest that Santiago was copied from the Leiden Elzevir. 35 Errors that the mystery edition corrects, but Leiden does not, also appear in the Santiago version. 36 Errors introduced by the mystery edition do not appear in the Santiago. 37 Though the Santiago version uses a great deal of contraction, it does not abbreviate when the mystery edition does, but either replicates the Leiden edition’s contractions, or creates its own. 38

Two substantial omissions also suggest that the Santiago edition was pirated from Elzevir’s. On the fourth and last page of the address to readers, five words from Grotius’s description of nature as an equal parent, bountiful to all and whose authority extends over everyone (in omnes munifica, cuius imperium …), is missing. This could have happened when the typesetter’s eye skipped a line in the Leiden edition. Because they are aligned differently, this explanation would not work for either the manuscript or for the mystery edition. Similarly, the sentence on f107v, lm35: Et si quicquam eorum prohibere posset, puta piscaturam qua dici quodammodo potest pisces exhauriri, at navigationem non posset, per quam mari nihil perit, is truncated on S15[12]r to Etsi quicquam eorum prohibere posset, per quam mari nihil perit. The most obvious explanation is that the copyist or typesetter was copying from the Leiden edition, and missed two entire lines. Neither the manuscript, nor the mystery edition align in a way that would make this a plausible explanation.

Though comparing capitalisation across the three editions and the manuscript yields no meaningful pattern, changes in the capitalisation conventions of the Santiago edition alone might be due to relatively consistent preferences. On the first page of the address to readers, patremque in lm becomes Patremque. In a long citation from Cajetan in Chapter iv, preachers, praedicatores, begins with a small p in all versions except the Santiago, which capitalises the word (f99v, lm11, cf. S7v). In a key passage in Chapter v, where Grotius argues that ancient references to shores enclosed within the Roman empire were owned by the people, not privately, the manuscript and the other two editions capitalise the Roman People, Prince, and Praetors. The Santiago version decapitalises the people, but keeps the other two (f103r, lm22, cf. S11r). Populus Romanus (f104v, lm26) becomes populus Romanus (S12r) again, and again (f104v, lm26, cf. S12v). The one exception is on S12r, where the capital initial of Populus is printed the same as on f107v and lm35. Conversely, when f109v and lm41 print Principis, S17r gives principis. In the next sentence, in the phrase that no man may be the lord of all mankind, f110r, lm41 dominus becomes Dominus on S17r. Similarly, f105r, lm28 Reipublicae is merely reipublicae on S13r. Mundi on f116v, lm61 becomes mundi on S23r. The manuscript, and the Leiden and mystery editions all have doctores Hispani at the end of Chapter iv (lm13); Santiago has Doctores (S8r). F113r, lm49 doctores, referring to legal authorities cited by Vázquez, becomes Doctores on S19v. When f113r, lm50 prints Doctores, S19v retains the capital. Natura in all other versions becomes natura in Santiago (f97r, lm16, S9r), Naturae in all other versions (f96v, lm2) becomes naturae (S4v; f108v, lm37, cf. S16r; and f111r, lm45, cf. S18v). It is difficult to draw conclusions from generally erratic early modern capitalisation practices, but the systematic alterations in the Santiago copy may reflect preferences that were more orthodox and monarchist than Grotius’s, who emphasised popular sovereignty at several points in Mare liberum, and whose use of the Bible and religious authorities in this text was challenged among others by Welwod. 39

Another aspect of intentional variation is the decision by the Santiago edition’s printer to eschew all-capitals. Unlike the transliteration of Greek characters, which was probably due to a lack of access to Greek type, the decision not to use all-capitals was a stylistic choice, as the printer owned the Latin capital letters, and every other version – the manuscript, the Leiden and the mystery editions – capitalises the same few dozen words that the Santiago refuses to. 40

The Santiago copy’s printer also has a probably intentional preference for varying m’s and n’s. The third page of the address to readers gives utrunque for what both other editions print as utrumque. Quantumcunque (lm10) is printed in the Santiago copy as Quantumcumque (S7r), numquam as nunquam in the long citation from Cajetan (lm11, S7v), and utrumque (108v, lm38) as utrunque (S16r). F109v has quam diu, turned into quamdiu on lm40, but quandiu on S16v. Philip iii’s letter contains a cuiuscunque in the Leiden version, cujuscunque in the mystery edition, and turned into a cuiuscumque on S25r. 41 A few spelling conventions may offer further clues to this printer’s identity. Piratas becomes pyratas (S5r), and piratarum becomes pyratarum (f107r, lm33, cf. S14[10]v). On the last page of the address to readers auctoritas in all other versions becomes authoritas in the Santiago copy. On f109r, L39, M40 auctoritatem becomes autoritatem on S16v; the same happens again on 111r, lm49, cf. S19v. In addition, f109v, L40 Iohanni and M40 Johanni become S16v Ioanni, just as f112r, lm47 Iohannis becomes S19r Ioannis, and f113v, L51 Iohanne and M51 Johanne become S20r Ioanne. Finally, the Santiago version introduces well over 70 new typographical errors. 42

So far we have not been able to identify the printer, but an Iberian or Spanish American origin is a real possibility. Specialists have kindly advised us that printers there conservatively preferred numbering folios (with numbers only on the recto), and there was no Greek printing in Spanish America until the late eighteenth century. The obviously fake title page, lack of Greek type, the compression without regard to Grotius’s midpoint, the multiple features that consistently indicate copying the Leiden Elzevir, and its own apparently deliberate alterations in capitalisation and spelling, may suggest a Spanish or even Latin American press; but at this point this remains pure speculation.


The 1609 Mare liberum appeared in at least three editions. Piratical printers took Grotius’s argument to heart, and treated the printing world as a mare liberum. Based on the books’ features, the mystery edition was probably printed in Arnhem or Cologne. The Santiago copy was printed by someone who at the time of printing had no access to Greek type. Unlike the mystery printer, the Santiago edition’s printer was not concerned with the number of pages and midpoint that Grotius tried to calculate in the surviving draft manuscript. There is no obvious reason to doubt that all three versions appeared in 1609, and that the Arnhem/Cologne and Santiago editions were both pirated from the Leiden Elzevir. However, given its effort to follow the Leiden pagination and Grotius’ midpoint, we cannot rule out the possibility that the Arnhem/Cologne edition was also based on a manuscript, or at least the printer was familiar with Grotius’s instructions to Elzevir.

The north German ports, where the largest number of copies of both the Leiden and Arnhem/Cologne editions were acquired early and are often still held, should become part of the detailed analysis of Mare liberum’s early reception.

In addition to insights into the early reception and political context of Mare liberum, this article provided a summary census of surviving copies, in the hope of encouraging readers to find many more. Comprehensive analysis of additional copies, and of the handwritten marginalia, may help to confirm where the mystery edition was printed; the origin of the Santiago copy; whether the two pirated editions were really published in 1609; and what annotations can tell us about the early reception. By noting the existence of three distinct editions dated to 1609, describing them in some detail, and providing a list of known copies, this article is hopefully a stimulating first step in the rediscovery of Grotius’s seminal work.


See e.g. China Miéville, Between Equal Rights: A Marxist Theory Of International Law (Brill, 2005). Eric Wilson, The Savage Republic (Martinus Nijhoff, 2008). Martine van Ittersum, ‘Dating the Manuscript of De Jure Praedae (1604–1608): What Watermarks, Foliation and Quire Divisions can Tell us about Hugo Grotius’ Development as a Natural Rights and Natural Law Theorist,’ History of European Ideas 35 (2009), 125–93, 142–43. John D. Haskell, ‘Hugo ­Grotius in the Contemporary Memory of International Law: Secularism, Liberalism, and the Politics of Restatement and Denial,’ Emory International Law Review 25:1 (2011), 269–98. Marti Koskenniemi, ‘International Law and the Emergence of Mercantile Capitalism: Grotius to Smith,’ The Roots of International Law (Martinus Nijhoff, 2014), ed. by Pierre-Marie Dupuy and Vincent Chetail, pp. 1–37. Christopher R. Rossi, Sovereignty and Territorial Temptation: The Grotian Tendency (Cambridge, 2017).


For this context see Martine van Ittersum, Profit and Principle: Hugo Grotius, Natural Rights Theories and the Rise of Dutch Power in the East Indies (1595–1615) (Brill, 2006).


This and the next paragraph rely on Martine van Ittersum, ‘Preparing Mare Liberum for the Press: Hugo Grotius’ Rewriting of Chapter 12 of De iure praedae in November-December 1608,’ Grotiana 26–28 (2005–7), 246–80, at 248–57; and Henk Nellen, Hugo Grotius: A Lifelong Struggle for Peace in Church and State, 1583–1645 (Brill, 2015, translation and revision of Hugo de Groot: Een leven in strijd om de vrede, 1583–1645, Balans, 2007), pp. 107–09.


‘Characteres sane placent, et libelli forma. utinam chartam non pessimam adhibeant.’ ­Grotius, Briefwisseling van Hugo Grotius, ed. by P.C. Molhuysen, B.L. Meulenbroek and H.J.M. Nellen, 17 vols. (The Hague, 1928–2001) (henceforth bw), letter on i.134-5, citation from i.135. Accessed online at Also see Ittersum, ‘Preparing,’ p. 257.


Ittersum, ‘Preparing,’ p. 256. Nellen notes that despite Oldenbarnevelt’s instruction the ‘embargo was not very strict, and the printer was allowed to send advance copies to the spring book fair in Frankfurt.’ Hugo Grotius, p. 108.


Hugo Grotius, p. 107.


Jacob ter Meulen and P.J.J. Diermanse, Bibliographie des écrits imprimés Hugo Grotius (The Hague, Martinus Nijhoff, 1950) (henceforth tmd) p. 213, no. 542. Alphonse Willems, Les ­Elzevier Histoire et annales typographiques (Brussels, 1880), p. 21, no. 56.


Jeroen Vervliet, General Introduction, in ed. and annot. Robert Feenstra, Hugo Grotius, Mare Liberum, 1609–2009 (Brill, 2009), p. xxxv n39.


Contravening a request to suppress a book would not have been unprecedented for the age. Also in early 1609, Franciscus Gomarus published Vvaerschouwinghe … in Leiden, his speech to the States General against Arminius, in spite of the States’ actual ban. Edwin Rabbie, Introduction to Hugo Grotius, Ordinum Hollandiae ac Westfrisiae Pietas … (Brill, 1995), pp. 5–6.


Feenstra, Mare liberum, pp. l-li n109.


Ter Meulen and Diermanse, p. 213, no. 542, mistakenly write that the mystery edition prints the date as ‘ANNO M.DI.IX.’


Ittersum, ‘Preparing,’ pp. 257–58.


Feenstra, Mare liberum, pp. 24n2, 24n3; 31nb, 38n3, 58n2, 72n3, 74n1, 76n1; 80n1, 94n2, 97na, 112n2, 112n3, 114n2, 114n3, 114n5, 116n1, 120n1 fecit, 122n1, 122n2, 129nb, 129nc, 130n2, 130n3, 133na, 133nb, 136n1, 137na.


Rabbie, ‘Introduction,’ pp. 39–40.


Ittersum, ‘Preparing,’ pp. 257–58. In commenting on this article, Hans Blom pointed out that the midpoint had to be recalculated, otherwise it would be at the end of a page.


Rabbie, ‘Introduction.’


As mentioned, we are not reproducing differences between Ms. bpl 917 and the 1609 Leiden edition that Feenstra has already noted. Some of these also cannot be explained by the copyists’ or the printer’s errors or corrections based on editorial decisions, Grotius’s style, or by the copyists or printers checking Grotius’s citations against print versions of Vázquez, Vitoria, Aristotle and other texts cited in Mare liberum. Such differences should be added to the list in this section, strengthening the probability of at least one more manuscript. See e.g. Feenstra, Mare liberum, p. 72n3 (where Feenstra asks, ‘why does it not occur in Ml?’, given that the manuscript is clear), p. 74n1, p. 122n1, and p. 122n2. One anonymous reviewer kindly pointed out that the practice of destroying the typesetter’s copies after printing, and the return of the surviving Mare liberum section of Ms. bpl 917 into the full ipc manuscript, further raise the probability that the book was printed from a different manuscript.


Also noted by Feenstra, Mare liberum, p. 24n3.


Feenstra notes the first omission in Mare liberum, p. 72n3, but not the second.


The passage is in fact from Isocrates: Feenstra, Mare liberum, p. 146.


Ferry Reurink, ‘Jan Jansz (?–1629), boekverkoper, uitgever en drukker’, in Biografisch Woordenboek Gelderland (deel 10, 2014), pp. 86–88.


With Paets: poems in a 1602 collection, In nuptias nobilissimi …; a 1607 edition of paraphrases on Aristotle’s Nicomachean Ethics attributed to Andronicus of Rhodes; 1607 edition and translation of Maximus of Tyre’s Dissertationes; 1609 poems on the marriage of Gaspar Vosberg. With Maire: 1603 Elegiarum … With Clouck: Auriacus, and a 1609 edition of his funerary oration for Scaliger, apparently printed by Clouck, Elzevir and Plantin in collaboration. With Raphelengius: his Crepundia Siliana in the 1601 edition of Silius Italicus; his own orations in 1602; Hesiod in a 1603 edition and commentary; a poem in Clusius’s 1605 Exoticorum …; and his 1607 panegyric on Scaliger.


Heidelberg: Theocritus, Moschus, Bion and Simmias, 1603. With Verdussen: a poem in Otto van Veen’s 1608 Amorum emblemata.


Plantin: tmd 12, Raphelengius: tmd 8, 9, 13, 242, 246, 407, 411, 413, 414, Basson: tmd 11, 31, 261, 264, 405, 406, Dorp: tmd 265, Henrick: tmd 20, 21, 244, Heidelberg: tmd 27, 256, 259. Paets: tmd 268, 408. Maire: tmd 53, 262, Clouck: tmd 54, 248, 251, 252, de Roy: tmd 245, Meerman: tmd 247, Elzevir: tmd 257, Hillebrand: tmd 258, 260, Guyotius: tmd 249, 250, Nieustadt: tmd 253, 254, Nieulandt: 255, Paris: tmd 263, 266, Antwerp: tmd 267. Of course, Grotius may not have had a direct relationship with these printers, given the practice of ‘borrowing’ poems and shorter pieces published elsewhere. Nonetheless, checking all these printers seems methodologically safer than assuming that the surviving correspondence and circumstantial evidence is sufficiently complete to exclude the possibility that Grotius had direct contact with any of them.


On this book’s significance see Somos, Secularisation and the Leiden Circle (Brill, 2011), Chapter iii.


See e.g. the C in the introduction on pp. 27, 29, in Book i, 13, and throughout.


Nellen, Hugo Grotius, pp. 377, 588–91.


Nellen, Hugo Grotius, pp. 375–76.


Janssonius Sr. subcontracted extensively, and employed printers in Amsterdam, as well. Thus it is possible that the unidentified edition was printed in Amsterdam; but the surviving evidence is inadequate for such a line of inquiry.


Slightly different gryphon design than the Santiago copy’s, printed by Sebastian Gryphius: Juan Luis Vives, Excitationes animi in Deum (Lyon, 1543). Same design, printed by Antoine: Testamenti novi (Lyon, 1569). Another slightly different design, printed by the heirs of Sebastian: Testamenti novi (Lyon, 1564). Yet another slightly different design, printed by Johann Gryphius: Ovid, Metamorphoseon (Venice, 1586).


Same device, no border, printed by Sebastian Gryphius: Erasmus, Apophthegmatum opus … (Lyon, 1541). By Antoine Gryphius: Plautus, Comoediae viginti … (Lyon, 1581). By the heirs of Sebastian: Jacopo Sadoleto, Epistolarum libri sexdecim (Lyon, 1560).


Same device as the Santiago copy’s, with border, printed by Sebastian Gryphius: Gerolamo Cardano, Contradicentium Medicorum (Lyon, 1548). Yet another slightly different design, with border different from the Santiago copy’s, printed by Antoine: François Valleriola, Observationum medicinalium libri sex (Lyon, 1583). Another different gryphon design, with border different from the Santiago’s, printed by Johann: Ambrogio Calepino, Dictionarium … (Venice, 1590).


Lorenzo Ruiz Hidalgo, La imprenta en Salamanca (1501–1600) (Madrid, Arco Libros, 1991), pp. 57–58.


In Chapter v, the mystery edition’s foedere is federe in both the Leiden and Santiago versions. F101v, L17 and S9r give occupatio, but M16 has Occupatio. Atque on f101r, L18, S9v is merely atque on M18. F102v, M19 Jurisconsultis is L19, S10r Iurisconsultis; and f102r, M20 Aer is L20, S10r Aër. F109v, M40 occupatio: non … becomes L40, S16v occupatio. Non … 109v, M40 qui became quae in both L40 and S17r. In Philip iii’s letter, printed as an appendix, a has in the Leiden edition appears as hac in the mystery edition, but has in S25r. On the last page of Philip iii’s letter, Leiden prints indicens. So does S25v, while the mystery ­edition has indicent. F108v, L37 and S16r has a mare near the end of Chapter vi, printed on M37 as Mare. The same happens again a few lines later, on f108v, L38, S16r, cf. M38, as well as with mari and maris (f109r, L39, S16v), compared to Mari and Maris on M39. F111r, L45, S18v dominio maris is dominio Maris on M45. On f116v and M62 Lucra is lucra on L62 and S23r. F118v, M65 Ait is ait on L65 and S24r. That said, Mare commune on f109v and lm40 becomes mare commune on S17r; Mare on f110v, lm43, becomes mare on S18r; and Maris on the last page of Philip iii’s letter becomes maris on S25v. F110v, lm44 Thesi Loca turns into Thesi loca on S18r. The Santiago copy’s printer, in short, tended to follow the Leiden spelling, but was also able to decapitalise words independently from Elzevir.


Grotius fails to italicise the Ulpian and Labeo citations on f103r. They are printed without italics on L23 and S11r-v, but correctly on M23. The same goes for the Paulus citation on f103r, L24, S11v, cf. M24, and to two passages from the Digest on f105r, L28, S13r, cf. M28. Where f104r and M25 print sequintur, L25 and S12r give sequuntur. An at on f104v and M27 is ut on L27 (the errata notes that it should be at), but ad on S12v. The one exception is the ex ex on L45, corrected to a single ex on both M45 and S18v, even though unnoted in the Leiden edition’s erratum. Though not a correction but a capitalisation preference, f116r and L61 have Terrarum after the Hesiod citation in Chapter xii, but M61 and S23r print terrarum. Three cases that do not fit this pattern are Santiago printing sequitur at the end of Chapter iv, like the manuscript, while the other two editions give sequimur (lm13); f102r, S10v giving Quid for lm20 quid; and at the end of Chapter v, f108r and S15[12]v have legibus, but lm36 print Legibus. On balance this is far more likely to be a misprint than an indication that the Santiago followed the manuscript in these three cases.


E.g. subiere in the Virgil quotation on f101r, L17, S9v is subire only on M17; insulta vere in the Ovid passage on f101r, L17, S9v is insultare only on M17; f101r coperunt and M18 coeperant is coeperunt on L18 and S9v<!--GreaterThan-->. F104v quod et aliae is quod & aliae on L26 and S12v, but quod aliae on M26. F144v, L56 has a sui in the second sentence of Chapter x, misprinted as su on M56, but given correctly on S21v. The penultimate sentence of this chapter ends with a question mark on f115r, L57 and S21v, but M57 prints a colon instead. A hinc on f115v, L58 and S22r is missing from M58. One exception is accommodari on f113r and L49, which becomes acommodari on both M50 and S19v.


E.g. a. on M17 and M22 is autem on f101r, L17, S9v and on f103v, ll22, S11r; M20 sc and e. are f102, L20, S10r scilicet and etiam; f102r, L20, S10v vero is M20 v.


Welwod, An Abridgement of All Sea-Lawes … (London, 1613), in Grotius, The Free Sea (ed. David Armitage, Liberty Fund, Indianapolis, in, 2004), pp. 66–67.


There are 11 transliterated words in 4 groups in the long Cajetan quotation in Chapter iv, S7v; 3 words in 2 groups in a Horatius quotation in Chapter v, L15, M14, S8v; 2 words in the Aratus passage and 1 in the Seneca citation on S8v; 7 words on S9r; 7 words on S9v; 8 words on S10r; 2 words on f103r, L24, M23, S11v; 5 words on f104v, lm25, S12r; 1 word on f111r, lm44, cf. S18r; 20 words in the long quotations from Vázquez (f111r-112r, L44-48, M45-48, S18r-19r); 4 words in the Hesiod citation (f116r, L60-1, M61); 2 words in the Propertius passage on f118r, lm64, cf. S24r; and 2 words on f118v, lm66, cf. S24v.


One might dismiss such changes as accidental had it not been for their clear consistency, and for evidence that authors paid attention. Wtenbogaert and Grotius exchanged ­letters about printing cum or quum and foedus or fedus in Ordinum pietas. Rabbie, ‘Introduction,’ p. 45.


In the table of contents, propriam is misspelled as proprium in the title of Chapter xi. On the first page of the address to readers, a Huic is misprinted as Hui. The second page gives utriusque for utrisque, and contiguisset for contigisset. Israel’s iusta bella against the Amorites becomes iuxta bella, while the Saracens’ access to the lands of the Jews was aditu arcerentur in all versions, except for Santiago’s auditu arcerentur (S5r). The former misspelling reoccurs in the Boetius quotation in Chapter iv, where Non est iusta satis turns into … iuxta … The Santiago main text begins with capvt primum, probably by mistake. All other chapter numbers are indicated with Roman numerals, as they are in the manuscript and in the other two editions. Chapter i ends with a full stop in Santiago, but a question mark in all other versions. On Java, Molucca, Sumatra have their own rempublicam in all versions (f97v, lm5), but the Santiago drops the first m (S5r). On S5v the grammarians’ occupare is misprinted as ocupare, idololatriae loses the r, and what is Mahumetani in all other versions becomes Mahometani. In the title of Chapter iii, Lusitanos is misprinted as Lusicanos, while commercijs remisisse in Chapter iv becomes commertijs remississe. In Grotius’ exposition of why it is unlawful to wage war on pagans even if they refuse to convert, noluerint (lm11) is misprinted as nolluerint (S7v). In the long citation from Cajetan in Chapter iv, movere is misprinted as movera. The first sentence of Chapter v ends with a full stop in all versions except the Santiago, which prints a question mark. A quiddam in Chapter v (lm14) is misprinted as quidam in Santiago, and innuentes (lm15) is given as inuentes on the same page (S8v). The alludit before a Seneca citation in Chapter v becomes alliudit (S9r). F102r, lm19 utente is S10r vtentente. On the same page, esse becomes S10r esset. F102r sicut et usum is lm20 sicut & usum, but on S10v it turns into sicut vsum. In the citation from Plautus on f103r and lm21, rete is misprinted as recte on S10v. F104v, lm25 littore is S12r lottore. F105v, lm29 mancipatus becomes S13v pancipatus. F105v et ditionem, and lm30 & ditionem, becomes editionem on S13v. On the same pages, notam is misprinted as notant, and effugeret as affugeret. F106r and lm30 give accenderit in a citation of Ennius in Cicero’s De officiis, but S13v misprints it as accederit. From f106v and lm32 Aethiopiam, S14[10]r drops the h. This is an error, not a spelling convention, as the Santiago version prints the word with the h when it reoccurs on S10v. F106v, lm32 Oceanum turns into S14[10]r Occeanum. F106v, lm32 Carthaginis is misprinted as Cathaginis on S10r-v. F106v, lm32 oceani becomes S14[10]v Occeani. F107v, lm35 Brittonum is misprinted as S15[12]r Britonum. F108r, lm37 provincias is S15[12]v provintias. F109r, L38, M39 validius becomes S16r validus. F109v, lm40 animadverit is printed as animadvertit on S16v. F110r, L42, M43 additamentum is S17v aditamentum, and on the same pages dicturi becomes S17v dictuti. F110v, lm44 iure is misprinted as S18r inre. VASTISSIMUM IMMENSUMQUE PONTUM on 111v, L45 and M46 becomes vastissimum immensumque punctum on S18v. F111v, lm46 non Regnum Hispanicum contra semetipsum becomes S18v … semetipsam. F112r, lm47 communia is misprinted on S19r as ocmmunia. F112r, L47, M48 consuetudine becomes S19r consuetadine. At the start of the next sentence, Vtroque becomes, on S19r, Viroque, and in the same sentence the word relatas in all other versions is misprinted as relata in the Santiago copy. In a long citation from Vazquez, terras on f112v and lm49 is misprinted as terra on S19r. F113r, lm49 vide becomes S19v videt. In the next sentence the phrase, adjundgendum est etiam …, drops the est on S19v. F113r and lm50 print exstet and saeculi in the same sentence, but they become extet and seculi on S20r. F113r, lm51 at the end of Chapter vii give Moluccas and Asiatici, which S20r misprints as Molucas and Assiatici. F114r, lm53 victus becomes S20v virtus. F114r, L53, M54 Philosophi is misprinted as Philisophi on S20v. F114v, lm55 mercandi becomes mercendi on S21r, and a neque turns into a nec on the same pages. F115r, lm57 give primigenio in the second sentence of Chapter xi, while S21 has primogenio. F116r and L60 have iuri, and M60 prints juri, while S22v gives the incorrect iure. In a Hesiod citation on f116r, L60 and M61, universorum is omitted on S23r, and PRINCIPUM in the same citation on the same pages becomes Principii. F116v and lm61 perniciosum becomes pernitiosum on S23r. F117v, L62-3 and M63 oportet turns into S23v opportet. On the same pages, tranquillam is misprinted as tranquilam on S23v. On f118r and lm64, a Propertius passage has prorae, misprinted as prora on S24r. In the same line, f118r and M64 give the correct Centaurica, L64 misprints it as Centaurcia (noted and corrected in the errata), while S24r has Centaurca. F118r, L64, M65 innoxius is misprinted as inoxius on S24r. F118v, L65 citation from the Digest has ratem, M65 gives ratim, while S24r has ratë. F118v, lm66, citation from Cicero has the word, disceptationem, misprinted on S24v as disciptationem. F118v, L66[42], M67, the very last word of Grotius’s text, oppugnant, became oppugnat on S64v. On the last page of Philip iii’s letter, proposui in lm becomes propusui on S25v, redibis becomes redibus, and S25v drops a suum from the final sentence.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 242 0 0
Full Text Views 1304 422 52
PDF Views & Downloads 556 241 45