Hundred forty-third volume.

VIENNA, 1901.



Critical prepation for the III. and IVth volumes of
the new Tertullian-edition.

Dr. Emil Kroymann.

       Apart from the separate transmission of the Apologeticus, over time three different branches distinguish themselves in the transmission of the works of Tertullian.

        The oldest is represented in the mss. by the Par. Lat. 1622 saec. IX, usually named Agobardinus, and by the editions of Johannes Gangneius, Sigismund Gelenius  and Jacobus Pamelius, which are either (Gangneius) based on mss of a related type, or (Gelenius, Pamelius) have readings of that family written in the margin. Those works of Tertullian, which are only transmitted through this branch are united in the first volume of the new Vienna edition.

        The second branch belongs to to the 11th century. In the mss it is represented by the Montepessulanus no. 307 and the Paterniacensis no. 439 (now in Schlettstadt), both of which however contain only one parts of this transmission. Supplementary to this is the first edition of Beatus Rhenanus  (1521), for which the publisher used the Paterniacensis and the now lost Hirsaugiensis, and his third Edition, for which he requisitioned another collation of the now lost Gorziensis.   Those works which are contained in the Montepessulanus and Paterniacensis will appear in volume III.  


VI. Treatise: Kroymann.


        Finally the latest transmission that of the 15th century, represented by the numerous mss in Italian libraries that I found derive from two mss of the Bibl Naz. at Florence.1  These are the Codd. S . Marco VI. 9 and VI. 10. Close to the latter, are the Vindobonensis 4194 and the Leydensis 2.2  (Also made use of again are the first and third editions of Rhenanus, as both the Hirsaugiensis and the Gorziensis were just as complete as the named mss). The works transmitted only in these mss will appear in volume IV.  Remaining for volume II will be those works which are transmitted  to us in both the Agobardinus and this latest branch of the tradition.

        Now the ms material has been reviewed, 3 clear principles are needed in order to obtain the constitution of the text:

        1. To assess and recognise the relationship of the latest branch to the middle, which belong more closely together than has been previously recognised and the value of the few manuscripts or editions for the constitution of the text.

        2. To test the allied middle and latest transmission on its nature and its value in the oldest portions, together with the Agobardinus.


        When Rhenanus edited Tertullian for the first time in the year 1521 in Basle, he possessed according to his own testimony (Praef. p. 2 and 3) two manuscripts: the Cod. Paterniacensis4  

1. Minutes of the Viennese Acad. CXXXVIII, 3.

2. As well as the works contained in the Montep and Patern., all four contain a set of other works, that are also partly contained in the Agob.

3. The collation of the Paterniacensis is not yet done. The mss is now in my hands.

4. At the Abbey of Payerne (Peterlingen) on the Neuenburger See. Today  it is located in the city library of Schlettstadt, where it came through the Decanus Jacob Zimmermann.

Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


and the Cod. Hirsaugiensis, in two volumes.1  The  Paterniacensis still available today contains the following works :

de patientia dei,
de vera carne domini (= de carne Christi),

de resurrectione carnis,
adversus Praxeam,
adversus Valentinianos,
adversus Judaeos,  
adversus omnes haereses,
de praescriptione haereticorum,
adversus Hermogenem.

For these works based his edition completely on the Paterniacensis : the Hirsau mss, which also contained these works together with 13 others, he used only as a control and supplement, in that he, as the MS shows, noted variants and supplements from the Hirsaugiensis in the margin. The MS then went to the Basle press and there was printed from, so that the marginal comments in the edition exactly correspond to those in the MS. For the remaining 13 works:

de corona militis
ad martyras,
de paenitentia,
de virginibus velandis,
de habitu muliebri,
de cultu feminarum,
ad uxorem libri duo,
de persecutione,
ad Scapulam
de exhortatione castitatis,
de monogamia,
de pallio
and adversus Marcionem

the Hirsaugiensis was the only  source, so the readings in the margin are valuable only as conjectures. For these works the edition may be counted as a replacement for the Hirsaugensis.

       I found that the archetype of almost all Italian MSS was the Cod. S. Marco VI. 10, that breaks down into two parts written by different scribes and  contains these works of Tertullian in the following sequence:

de carne Christi,
de carnis resurrectione,
de corona militis,
ad martyras,
de paenitentia,
de virginibus velandis,
de habitu muliebri,
de cudtu feminarum,
ad uxorem libri duo,
de persecutione,
ad Scapulam,
de exhortatione castitatis,
de monogamia, 
de pallio

- de patientia dei,
adversus Praxeam,
adversus Valentinianos,
adversus Marcionem,
adversus Judaeos,
adversus omnes haereses,
de praescriptionibus haereticorum,
adversus Hermogenem

-- therefore in the same sequence, as they were in the
Hirsau MS. This MS was written


1. These he had been allowed to borrow through Thomas Rappius going bathing at Wildbad.

2. The writings de carne Christi and de carnis resurrectione are omitted as Rhenanus edited them from the Paterniacensis.


VI. Treatise: Kroymann.


according to the signature in the year 1426 by two Franciscans: Thomas von Lymphen and Johannes von Lautenbach, and both in Pforzheim. The observation, that the Hirsaugiensis has the same arrangement of the writings as the MS copied at Pforzheim which I designate as F together with the circumstance, that Pforzheim is in the vicinity of the Abbey of Hirsau, made me assume that F was perhaps a copy of both volumes of the Hirsau MS. However connected to F in the closest way are the Vindobonensis 4194 (V) mentioned above and the Leydensis 2 (L) which, as I showed earlier,1 are derived from the same archetype as F, and in fact F is a direct copy, and V and L through an intermediate copy. This archetype was copied to make F in Pforzheim, and it must be asked therefore, whether this must be the same as the the Hirsaugiensis of Rhenanus. To investigate this I had accordingly my collations of F V L together with the first edition of Rhenanus, which as stated above is in part an impression, if not an exact impression of the Hirsaugiensis (H). I made the word-sample from the first book and the first 15 chapters of the second book of Adversus Marcionem. In addition to H F V L I have included two more MSS, the above-mentioned Montepessulanus (M) and the second Florentine MS, Cod. Magliabechianus S. Marco VI, 9 (N), that I naturally had included in the earlier investigation. This gave the following result:

Common omissions.

582, 102 natura | 592, 30 sit
592, 2  Judaorum - deum | 607, 36 Quis - concupiscet
592, 27 aliquid | 611, 14 aut
600, 8 evangelii | 614, 8 hoc
601, 12 ipso | 618, legi
604, 15 debere | 622, 21 debuerit


1 Cf. my above-mentioned article p. 26-28.

2 I cite these references from Oehler, ed. minor.

Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.



606, 11 deo | 625, 23 exinde - terra
616, 3 et |  
617, 20 o |  
620, 25 eius |  
623, 33 de |  
629, 2 si |  

Common word position.

598, 5 probes eum esse cum probes esse
601, 25 apostoli sententia | sententia apostoli
601, 36 a nobis vincula eorum | vincula a nobis eorum
603, 11 et hic non minus | non minus et hic
604, 13 divina bonitas in terris | divina in terris bonitas
604, 19 ratione desertum | desertum ratione
608, 32 non ut | ut non
609, 4 fieri noluit | noluit fieri
621, 21 scilicet dei | dei scilicet
623, 2 etiam ne | ne etiam
624, 33 et deliquit ex illo | et ex illo deliquit

Common readings.

584, 3 Ponticos | Ponticus
584, 6 praenuntiationis | pronuntiationis
584, 11 obtunsis | obtusis
584, 28 ostendimus | ostendemus
587, 3 plura potest duo | plura post duo
590, 21 indubitato | indubitate
591, 1 ad hanc causam | adhuc causam
593, 20 nulla | nullam
593, 31 res ipsa | rem ipsam
594, 14 fecit et si | fecisset et si
594, 31 quos | quas
595, 12 docens | dicens
595, 14 araneae | aranei
595, 23 suos | suo
595, 29 caelo | caelum
596, 1 substantia | substantiam


VI. Treatise: Kroymann.

596, 7 a XII | a XV
596, 31 in substantia | substantiam
597, 1 et siccis | et sucidis
598, 5 per quem | per quae
598, 11 deducta | deductus
599, 1 deinde negas | dehinc nega
600, 15 inconcussam | inconclusam
602, 24 haeretice | haeretici
603, 6 obvenientia | obventicia
603, 18 si non potest | se non potest
604, 1 sicut | si ut
604, 24 cum ergo | cur ergo
606, 4 qui salvos | quos salvos
606, 24 debui | debuit
606, 36 bonitate | bonitatis
607, 10 perfecta est Alia | perfecte. Alia
609, 21 ut | et
609, 37 diligitur | diligetur
610, 12 circum furenteis  | circi furentis

        The relationship of F V L to the Hirsaugiensis and of N to the Montepessulanus is proven hereby. Nevertheless F cannot be a direct copy of the Hirsaugiensis. For apart from the omissions listed above there are a further 38 common omissions, which the Hirsaugiensis did not have.  This forces the conclusion, that the Hirsaugiensis was first of all copied in the former Franciscan abbey1 at Pforzheim and that then at the time of the Basle councils from this copy a further two copies were taken, of which one is our Florentine MS, while the other, which also came to Italy, was lost, but still survives in V L (both of Italian provenance).2


1. Prof. Stelzner Pforzheim has kindly told me, that the Franciscans had already settled in Pforzheim by around 1270. After the Basle Council the abbey was reformed by Nicolaus Coroli of Heidelberg. The monks had a sum of 400 fl. to spend on the acquisition of books. They seem therefore to have increased their library by copying mss.

2. Rhenanus does not have the word ipsius on p. 605, 30, which in appears in F V L, but this is an obvious error of the typesetter.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


Therefore F is a grandson, and V L great-granchildren of the Hirsaugiensis, and in accordance with this we see that the corruption of the text in V L has gone a stage further than in F. As today we have the text of the Hirsaugiensis, to be sure not without changes, in the edition of Rhenanus, we can omit V L henceforth from the critical apparatus1 and use F as a control everywhere for Rhenanus, where we are in doubt about the text, whether it is that transmitted or a conjecture.

        The third edition of the Rhenanus of the year 1539 made not insignificant progress. This is how he announced in the preface, that after long exertion he had been able to obtain a collation of the MS of Gorze , 'diligentia ac dexteritate Huberti Curtinei, viri cum pietate tum eruditione excellentis, adiuvante Domenico Florentino sodali peractam.'

        This Gorziensis contained the same works, that are contained in Cod. N2 and its traces are in the annotations that Rhenanus placed against each work.
By reason of the multiple agreements between the Gorziensis and N, I earlier believed3 that N was probably a copy of this now lost Gorziensis. To obtain an adequately established view over the relationship between both mss is very difficult, in that we no longer have the collation of the Gorziensis that was made for Rhenanus. Moreover we cannot judge how carefully it was made or to what extent Rhenanus followed  it. We can make a closer check, in that his comments about it in the Annotationes are entirely incomplete. For it appears that in the third


1. Thus the time that I spent on the collation of both manuscripts, trusting the opinion of earlier editors, unfortunately has been entirely wasted.

2. The Annotationes prefacing the Apologeticus are entirely missing.  The reason is that Rhenanus once again copied this text from the Aldine edition of 1515. Thus he did not use the Gorziensis collation for this work.

3. Cf. my article p. 31.


VI. Treatise: Kroymann.


edition of Rhenanus there is an entire group of unannotated new readings, that are certainly not conjectures, but rather from his collation of Gorziensis. It was therefore necessary to compare the third edition with the first and then the results of this collation with the transmitted text from manuscripts N and M. At the same time arose first of all that, as I have already suggested, the Gorziensis is undoubtedly most closely related to the tradition represented by N M.. As a proof, see the following grouping of variants out of the first 25 chapters of the first book of Adv. Marcionem.

p. 582, 10 Pontus qui igitur Euxinus negatur, nomine 
    illuditur H F
Pontus qui igitur Euxinus natura negatur 
    nomine illuditur G M N.
p. 584, 26 de bono praestruendo RF
de bono praeferendo G N
de bono praestru ferendo M.
p. 587, 12 quae H F
qua G N M.
p. 592, 2 Judaorum enim deum dicunt animae deum
om. H F.
p. 592, 6 a certo certus H F
a certo incertus G N M.
p. 592, 27 aliquid G N M
om. H F.
p. 593, 21 nulla H F
nullam G N M.
p. 598, 13 deducta H F
deductus G N M.1
p. 600, 8 evangelii G N M
om. H F.
p. 600, 11 indicantur H (indueantur F)
indurantur G N M.


1.  In fact 'deducta' stands in the text, but Rhenanus in the annotationes notes that the Gorziensis had 'deductus'.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


p. 600, 25 tibi H F
tibi scilicet M N
scilicet tibi G.
p. 600, 35 susceptam H F
suspectam G N M.
p. 601, 12 ipso G N M
om. HF.
p. 604, 1 Sicut H F
Si ut G N M.
p. 606, 29 debui H F
debuit G N M.
p. 606, 36 bonitate H F
bonitatis G N M.
p. 607, 12 perfecte deum ostendere HF
perfecte bonum ostendere G N M.
p. 607, ult. Quis volet, quod non concupiscet H F
om. G N M.1

        This selection of generally agreeing lacunae and readings proves for me that G belongs with NM, yet other circumstances seem to contradict this result. Firstly it appears from the third edition of Rhenanus that the many deviations in H F of word position by comparison with N M are not an isolated case, as appears from the following comparison.

p. 598, 5 eum probes N M
probes eum F Rhen. I und III.
p. 601, 36 a nobis vincula eorum N M
vincula a nobis eorum F Rhen. I und III.
p. 603, 11 non minus et hic N M
et hic non minus F Rhen. I und III.
p. 604, 1 divina in terris bonitas N M
divina bonitas in terris F Rhen. I und III.
p. 600, 25 tibi scilicet M N
scilicet tibi Rhen. III.2


1. On the authority of the Gorziensis Rhenanus in the third edition omits these words.

2. Rhen. 1 does not generally have scilicet. In the Annotationes scilicet tibi appears specifically as a reading of the Gorziensis.


VI. Treatise: Kroymann.



        In two places, we even find gaps, which the Hirsaugiensis had, not filled in, although NM show the missing word:

        p. 604, 14 Aliam illi regulam praetendo, sicut naturalia ita rationalia esse [debere NM] in deo omnia.

        p. 606, 11 Quem enim iudicem tenes, dispensatorem si forte bonitatis ostendis intellegendum, non profusorem, quod tuo [deo N M] vindicas.

        Finally a series of variants are missing in Rhen. III, and undoubtedly they are variants which in accordance with the close relationship of G with N M we would expect to find in it:

p. 594, 9 praesignavit N M
praesignaverit Rhen. I und III.
p. 595, 12 dicens N M
docens Rhen. I und III.
p. 595, 7 aranei N M
araneae Rhen. I und III.
p. 599, 1 de hinc N M
deinde Rhen. I und III.
p. 600, 28 currisset N M
cucurrisset Rhen. I und III.
p. 601, 35 praececinerat N M
praecinuerat Rhen. I und III.
p. 603, 6 obventicia N M
obvenientia Rhen. I und III.

        It would however be incorrect to presume that the Gorziensis did not in fact perhaps tally in all these things with N M. The responsibility for this is undoubtedly owning to on the one hand the collation, which was not arranged according to the demands of modern criticism, and on the other hand Rhenanus himself preferred in many cases to give the advantage to the reading of the Hirsaugiensis.1 It is therefore clear, that the third edition of

1. Incidentally the arbitrariness of Rhenanus in his third edition goes very far. He gave plenty of room to conjecture and his supposed emendations are brought into the text without any explanation. Thus following editors hereby are led astray, vague suppositions of Rhenanus are transmitted as part of the tradition, and this becomes the appropriate place


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


Rhenanus  gives us only a very imperfect picture of the Gorziensis. Codex N becomes for us all the more more important for the constitution of the text in volume IV.

        The question now is, whether criteria are available, to allow a conclusion on the relationship of  G N M. I believe firstly,  that we probably have good reason to say that N in its first part  (p. 1-134 v.) is copied from M, whether directly or through an intermediary.  For the Montepessulanus shows two sets of corrections, one in the hand of the writer himself and a second in another hand, which however in Reifferscheid's judgment is also old. That now in the Florentine manuscript the corrections of the scribe himself appear in the text is naturally no proof of its dependence of M, as in that case the scribe corrected from his model.

to give examples on this sore point:

p. 583, 3 Sedes (sc. gentium Ponti) incerta, vita cruda, Libido promiscua et plurimum nuda; etiam cum abscondunt, suspensis de iugo pharetris, ut indicibus notentur, ne qui intercedat. - Thus the unanimous transmission, that is totally intelligible, when one supplements the genuine circumlocutions of Tertullian simply according to the sense: etiam cum abscondunt (sc. libidinem), suspensis de iugo pharetris (sc. abscondunt), ut etc. But Rhenanus makes three changes: suspensis de iugo pharetris indicibus, ne temere quis intercedat.

p. 589, 5 Id ergo summum magnum, quod deo adscribimus ex substantiae lege, non ex nominis lege, contendimus ex pari esse debere in duobus, qui ea substantia constant, qua deus dicitur, quia qui in quantum dii vocantur, id est summa magna, substantiae scilicet merito innatae et aeternae ac per hoc magnae et summae, in tantum non possit summum magnum minus et deterius alio summo magno haberi. - Here Rhenanus deletes the qui behind quia and the et before summae. That the latter is unnecessary, was already seen by Rigault, who returned it to the text. But also Rhenanus is hardly correct to remove the qui; it seems to me only to have been misplaced to me, and I write: qua qui (= aliquis) deus dicitur, quia etc. qui = quis s. above: ne qui intercedat. - It should be corrected by transposition, not deletion:

p. 592, 30 Adeo inde auctoritas accomodata si falsae divinitati, unde praecesserat verae. Unam sattem cicerculam deus Marcionis protulisse debuerat. - The si, which Rhenanus deletes, should be placed before unde and the full-stop after verae changed to a comma.


VI. Treatise: Kroymann.


However the proof is that the corrections of the second hand which are due to conjecture, not transmission, also appear in N in the text.  In the portion I examined this was the case three times.

       590, 37 Si ita est, ecquid tibi videtur iusta ratione defendi, ut ad normam et formam et regulam certorum probentur incerta? - Therefore M offers: Si ita e <'..' under letter e> (in mg.: sunt, man. al. antiqua) haec quid tibi videtur etc. The corruption haec quid for ecquid allowed the corrector to object to the est and to insert sunt for it so that he therefore read: si ita sunt haec, quid tibi videtur etc.  And so it now reads in the text of Codex N.

       596, 7 M: At nunc, quale est ut dominus a. XV Tiberii caesaris revelatus sit, anno substantia vero a.l XV iam Severi imperatoris nulla omnino comperta sit. There can be no doubt - although no editor has recognised this, that the a  followed by a full-stop is not the preposition, but rather the abbreviation for anno, whence the foolish ad of Pamelius. But already the corrector of M grasped that it was a preposition and supplemented logically  anno which he wrote above it the line. In N, the word is repeated in the text. In the Gorziensis on the other hand it might not have been there, as it also does not appear in the third edition of Rhenanus.

        595, 34 Rosam tibi si obtulero, non fastidies creatorem; hypocrita, muta porrocaracte [XXXX over rocara]  re [circumflex over re] si probes te Marcionitam. The specific testimony of Rhenanus is that the Gorziensis read here: hypocritam ut apocarteresi grobes te Marcionitam, etc., undoubtedly what Tertullian wrote, except that Rhenanus changed the accusative hypocritam into the vocative. What we read now in M, is a correction by the second hand
and Reifferscheid believed that under that he could still clearly recognise the reading of the Gorziensis. In N however we read in the text : hypocrita, muta porro caracterem si etc. A further example I found in de carne Christi p. 918, 24. Here M offers M: Oro vos, si dei Spiritus [filivs over Spiri] non de vulva carnem participaturus

1. For this Pamelius introduces  ad , and Oehler followed him.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


descendit in vulvam, cur descendit in vulvam? The  filius  written above the line is, in Reifferscheid's judgment, in a third hand of saec. XIII or XIV; In N it appears after the omission of Spiritus in the text.1 However with this it becomes as good as certain, that N not only derives from M, but rather is a direct copy of it, and that as now there appear in N a further 17 works, the conclusion is permissible, that also the Montepessulanus existed in two parts, just like the Gorziensis, whose latter part has been lost however. Whether the Gorziensis also derives from this formerly complete Montepessulanus, cannot be determined from our patchy knowledge of the manuscript. However on the other hand nothing has appeared which is contrary to my previous observations and so I consider then that the Gorziensis must have been an earlier copy of M, made before the second corrector of M registered his changes.

        On the other hand the Paterniacensis cannot derive from M although on the other hand both manuscripts coincide almost to the word. For in the 11 chapters of De Carne Christi from which I made the sample, P exhibits an improvement against M in five places.2 In the remainder however the agreement is so complete, that it is easy to see P as a brother of M, derived from the same archetype, and that there are 4 texts in P, which are not in M, is a further confirmation, that the original archetype of our collection contained a complete set.  To determine the relationship of the Hirsaugiensis to M P or better to its archetype, therefore involves difficulties, because Rhenanus follows the Paterniacensis faithfully for the texts which H had jointly with M P, making only notes from H in the margin. Until now, I have found no common lacuna in M P, which is also in H, whereby the origin of H from the archetype of  MP could be concluded.  Therefore I would like first of all to suggest it as a probability, that the Hirsaugiensis is also 


1. Because this is also printed in Pamelius, he may have used N.  But this must be examined more closely.

2.  p. 919, 15 eam.    p. 913, 5 filio.    p. 920, 2 quasi.   p. 920, 5 a. i. ut. p. 921, 2 et.    M contradicts P in an entire series of places.


VI. Treatise: Kroymann.


a descendant of the archetype of M P. If H shows a very considerable number of deviations when compared with M and P, this could be understood from the circumstance that this codex, as we observe from Rhenanus' comments (cf. z. B. p. 2), was corrected right through; or alternatively that one would have to establish another intermediate MS between the archetype and H. That the Hirsaugiensis was in any case younger than the Paterniacensis can be deduced from the fact that Rhenanus, where it was possible, used the latter as the basis of his edition. Moreover a preliminary remark to the old catalogue of Hirsau suggests (Becker, p. 100, 4), that almost all the manuscripts of the abbey were written under the abbots Wilhelm, Bruno, Vollmar and Manegold, i.e. in the period from 1077-1167. Therefore we will not go astray, if we assign the Hirsaugiensis to the 12th century.1

        In view of these results even to me the question must obtrude itself, where the joint source of our manuscripts, to which a very large part of Tertullian's works owe their existence, would most likely be found. Now Peterlingen (Paterniacum), Gorze and Hirsau are Cluniac foundations or at least influenced by Cluny, and therefore it is a small step to look for this archetype in the rich cloister library that existed at Cluny. This theory was not unfruitful. In the old library catalogue of Cluny published by in Delisle2 (1158-1161), to which Ernst Sackur was kind enough to draw my attention, we find under no. 73 and 74 the following notes:

        Nr. 73. Volumen, in quo continentur libri Tertulliani decem ad diversos et apologeticum eius.

        Nr. 74. Volumen, in quo continentur eiusdem libri XVII.

        Without a doubt these are the texts which we see assembled in the Florentine Codex N, and which its model M also probably once

1. This should not of itself hinder acceptance of other points of this work, as in any case the archetype of M P and H must be brothers.

2. Inventaire des manuscrits de la bibliothèque nationale.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


contained when it was complete. For M  and the first part of N in fact contain apart from the Apologeticus 10 texts of Tertullian, if one counts adversus Marcionem as five books, and that it was the same writings, is shown by the comment 'ad diversos'. And in the same way the second part of N, if we count ad uxorem as two books, contains exactly 17 writings. - Therefore the archetype of the collected Tertullian tradition of the 11th and 15th century still rested in the 12th century in two volumes in the cloister library at Cluny, and from M N  we know how the texts of Tertullian were distributed in both volumes.1 It is interesting to reflect that, however orthodox the Cluniac abbots were, the memory of a man whose very name was shunned until the theology of late antiquity has been transmitted by them. The question also naturally arises: from where did Cluny obtain this treasure of which prior to the 11th century we have only a single notice.
In an old library catalogue saec. X of the Abbey of Lorch2 (coenobium Laurisheimense) is itemized: Libri Tertulliani presbyteri : de patientia lib. I de carnis resurrectione lib. I Adv. Marcionem lib. V de carne Christi lib. I in uno codice. Further: liber Tertulliani presbyteri und item alius liber Tertulliani. Finally: item libri Tertulliani in alio codice. - Counting the works in the first-named codex, which are indeed all present in the first part of N and in M, both following in this a manuscript transmitted to them, the number ten appears again; the last-named codex then in fact could have contained the other seventeen writings, that the author of the catalogue found too long-winded to enumerate individually. But if we ignore the fact that these are just hypotheses, even if they were correct we would lack the means to decide

1. The changed order of the writings in the Paterniacensis and Hirsaugiensis is probably only for practical reasons, namely to balance better the contents of the two corpora, which were very irregular. Also the Paterniacensis might have once had a second part.

2. Becker, Cat, ant. 37, 320 and 321, 385 and 386. Mai, Spic. Rom. V, 186. Cf. the comments of Rhenanus on this, Oehler, praef. p. XVII.


VI. Treatise: Kroymann.


whether Lorch obtained its treasure from Cluny1 or Cluny from Lorch or even both from somewhere else. The actual problem, how it was generally possible that the works of heretics were preserved into that time, Harnack has rightly asked,2 and designates as the only solution the study of law of the Franconian theologians
of the 9th century. But associated with this problem, as I hope to demonstrate in the second part of this investigation, there is still another, namely how it was possible, that even in spite of the condemnation by the church an edition of the works of Tertullian could emerge; then, we can see how any such will affect my theory of the transmission of Cluny.


That in fact the transmission of Cluny is different from that of the Agobardinus, can first be seen from the different arrangement of the works in each. To illustrate this, I place the sequence of the works in each side by side:

Agobardinus | Florentinus N
Ad nationes libri II | De patientia
De praescriptione haereticor. | De carne Christi
Scorpiace | De carnis resurrectione
De testimonio animae | Adversus Praxeam
De corona |  Adversus Valentinianos
De spectaculis | Adversus Marcionem
De idololatria |  (Apologeticus)
De anima | De fuga
De oratione | Ad Scapulam
De cultu feminarum | De corona militis
Ad uxorem | Ad martyras
De exhortatione castitatis | De paenitentia
De carne Christi | De virginibus velandis


1. Of course the manuscripts of Cluny need not have been produced only after the foundation of the abbey in 910.

2. Tertullian in der Litteratur der alten Kirche, Sitzungsberichte 1895, p. 560.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


De spe fidelium |  De habitu muliebri
De paradiso | De cultu earum
De virginibus velandis |  De exhortatione castitatis
De carne et anima | Ad uxorem
De patentia | De monogamia
De paenitentia | De pallio
De animae submissione | Contra Judaeos
De superstitione saeculi. |  Adversus omnes haereticos
  | De praescr. haereticor.
  | Adversus Hermogenem.


        While in the Agobardinus works of polemic, apologetic, dogmatic and pastoral content are mingled together in colorful confusion, in the Florentinus it is impossible not to recognise the organizing hand of the collector. Is it by chance, that in the second part there are firstly four texts side by side, which all have events of the Christian life as a theme, that the next four all deal with questions of faith and followed by three which again discuss essentially the same question, namely which demands does the castitas Christiana make? And when now in the first part also, apart from the first text there are assembled purely polemical writings, it seems to suggest that, if we ignore the first work of the first part, and the five last works of the second part, that by some chance the original order is not entirely lost. There is a very simple explanation of how this interference occurred. I myself think that the original shape of this Corpus Tertullianeum was thus:

De fuga
Ad Scapulam
De corona militis
Ad martyras
De paenitentia
De virginibus velandis
De habitu muliebri
De cultu earum
De exhortatione castitatis
Ad uxorem


VI. Treatise: Kroymann.



De monogamia
De pallio
De patientia
De carne Christi
De carnis resurrectione
Adversus Praxeam
Adversus Valentinianos
Adversus Marcionem1
Contra Judaeos
Adversus omnes haereticos
De praescriptione haereticorum
Adversus Hermogenem.

        We now think that from this complete codex containing the common portion of both lines, assuming that it had not been split, some parts would have been intentionally taken out to reduce the bulk of this oversized collection, so it is apparent that related materials didn't have to remain together. However if the above-mentioned arrangement was original, then we have first in de pallio and de patientia two writings linked in this way, both with a strong personal element, and secondly the four polemic writings of the second part go in connection with the five of the first part in such a way that the main part goes forward without the remainder. If the collected polemical writings once again stand side by side in the Paterniacensis, then this looks rather like an attempt to restore the original order, although only half has arrived, as de Patientia is not attached to the other part. I am therefore of the opinion, that the now existing separation of this transmission into two corpora is the result of chance or an act of God, and that there was originally only a single large Corpus Tertullianeum organized from an impersonal viewpoint, i.e. an edition produced by an expert on Tertullian's works. That this corpus emerged in France, must at least be considered as

1. The work after this, the Apologeticus, was originally a stranger to the collection, as the prefatory note in M  shows: Post sex superiores adpositus est elegantissimus liber apologeticus de ignorantia dei in Christo Jesu.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


a possibility; the question however, at which time this happened and at whose instigation, must provisionally remaining undecided. It must now be considered whether the activity of the editor was restricted to simply collecting the works of Tertullian, or whether it represents at the same time a deliberate recensio of the text itself. There is therefore the possibility, to commence this investigation, that in the Agobardinus a tradition has been preserved which is independent of that of Cluny, and which, from the critical results above, should be regarded as a transmission of the text essentially free of intentional changes.

        However the corruption in the transmission of Cluny is in the manuscripts of the 15th century is so considerable, that the investigation can only proceed from the two manuscripts of the 11th century. The only text, which both manuscripts have in common with the Agobardinus, are the ten first chapters of De carne Christi.1 These must therefore form the source material for the following investigation.

        Methodologically it was first of all of importance, to assess, whether there were any incontestable indices to show that both traditions in the final analysis derive from the same archetype, by which of course I do not mean the original manuscript of the author. If this proof were available, it would place the analysis on a more certain ground, as the question of a possible dual edition by the author - i.e. originating with Tertullian himself - would become immaterial.

        As such indices can only be sought in shared corruptions of the text that are only capable of explanation from a common already corrupt source, in my commentary I have especially explored glosses and interpolations in the text.2 - For both however I can show several examples out of the first ten chapters of de carne Christi,


1. In chapter 10, the Agobardinus breaks off.

2. That such exist in the Agobardinus, I believe I have shown in my dissertation:  Quaestiones Tertullianeae criticae (cf. p. 117).


VI. Treatise: Kroymann.


which, as I said, do not form a barrier to the opposite opinion.

        p. 425, 11   Marcion, ut carnem Christi negaret, negavit etiam nativitatem, aut ut nativitatem negaret, negavit et carnem, scilicet ne invicem sibi testimonium responderent (ita A. - M P testimonium redderent responderent). The sense is unambiguous: nativitas and caro support each other and cannot be discarded separately. For here in Tertullian is the terminus technicus: respondere,1 of which Oehler's index shows examples enough. An accusative testimonium is at the same time because of the inherent actual meaning of the respondere entirely intolerable. How it came into the text tells us something about the transmission of MP: the testimonium redderent was a gloss written above responderent; in A, part of it, in M P the whole thing penetrated into the text.

        p.426, 11   Plane nativitas (Christi) a Gabriele adnuntiatur. Quid illi cum angelo creatoris ? Et in virginis uterum conceptus inducitur. Quid illi cum Esaia profeta (profeta om. M P) creatoris? The cum angelo creatoris corresponds only to one thing: cum profeta creatoris; the obtrusive addition Esaia insults every feeling for style. That a corruption exists, is confirmed by the other tradition which leaves out profeta, but retains Esaia. Here therefore the gloss displaced even the correct word, while in A it has crept into the text.

        p. 435, 13    Fuit itaque phantasma post resurrectionem, cum manus et pedes suos discipulis inspiciendos offert, aspicite, dicens (ita M P. - A inquit dicens), quod ego sum. - In order to connect both parts of the sentence more closely, someone had wanted to improve the looser, but with Tertullian very popular inquit with dicens. In A, it penetrated only into the text, in M P it displaced the correct text.

        439, 12   Sed2 temptandi gratia nuntiaverunt ei (sc. Christo) matrem et fratrem, quos non habebat. Hoc quidem scriptura non dicit, alias non tacens, cum quid temptandi gratia factum est erga eum. 'Ecce,' inquit, 'surrexit legis doctor temptans eum,' et alibi : 'et accesserunt ad eum pharisaei temptantes


1. The expression is juristic; respondere actually means to call before a court in order to answer for and to defend himself.

2. An objection of an opponent.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


eum'. Quod nemo prohibebat hic quoque significari [temptandi gratia factum]. The bracketed words give specifically that which is briefly - and after the foregoing quite unambiguously - expressed by the relative quod. That it is not tolerable to have these next to one another, should have been noticed by earlier editors.

        I restrict myself to these four places (although I believe that more than just one has been interpolated), because it appears to me indisputable, and because this would suffice to prove my proposition. The editor of the corpus has therefore for this work at least grounded his text on a manuscript, which is derived from the same already corrupt exemplar from which the Agobardinus derives, i.e. the collection came into being at a time in which the transmission was already no longer intact, and therefore in any case not all soon after the publication of the works.

        A further question is how has the man who assembled the corpus modified the transmitted text? That this transmission is very substantially different to that of the Agobardinus has never been in doubt. Whether however these deviations from the natural sequence are the product of a tradition around two centuries longer, or whether they show features which an incisive corrector had recognized, as they will be most naturally attributed to the collector of the Corpus, it is now appropriate to examine. For clarity I discuss the essential deviations in three sections: 1. Omissions, 2. Additions, 3. Variations.

1. Omissions.

        p. 425, 1    Qui fidem resurrectionis ante istos Sadducaeorum propinquos sine controversia moratam ita (ita Rigaltius. A morata ista. M P moratam) student inquietare, ut etc. Undoubtedly Rigaltius was correct to produce ita from ista, which should not be missing given the following ut. As this word is entirely missing from M P, the suspicion is that the collector intentionally omitted it, because he knew nothing that began with the corrupt ista.

        p. 427, 3    His opinor consiliis tot originalia instrumenta Christi delere, Marcion, ausus es, ne caro eius probaretur. Ex 


VI. Treatise: Kroymann.


quo, oro te? Exhibe auctoritatem. - The last words read in M P: ex qua, oro te, auctoritate? That in fact reads just as smoothly as the phrasing of the Agobardinus sounds unusual, and for that reason alone one must strongly mistrust M P. Over and above that, however the following justifies the reading of the Agobardinus: Si prophetes (M P propheta) es, praenuntia aliquid, si apostolus, praedica publice. After the question ex qua auctoritate? one would expect the continuation: ex prophetae? ex apostoli?, reversed during the exhibe auctoritatem through the imperatives: praenuntia aliquid, praedica publice give exactly the corresponding content.

        p. 430, 4 - - quasi nun valuerit Christus eius (eius om. P M) vere hominem indutus deus perseverare. The eius which P M leaves out (Oehler also, following them), has the following meaning: potentioris dei = dei Marcionis, and is not incidental but rather required. The omission in M P is hardly accidental; for it is not entirely obvious that the deus Marcionis has its own special Christ and this could well be incomprehensible to a naive reader.

        p. 432, 7    Nativitatem reformat (sc. Christus) a morte regeneratione caelesti, carnem ab omni vexatione restituit. A. For this M P offer: Nativitate reformata (morte om.) regeneratione caelesti carnem ab omni vexatione restituit. - It is easy to see how the version of M P appeared. The preposition a was attached to the preceding participle reformat, which left no place for morte. Undoubtedly the collector has edited and made both sentences of A into one.

        p. 430, 20     To what do these words relate: stulta mundi elegit deus. Quaere ergo, de quibus dixerit, etsi1 praesumpseris invenisse:2 num erit tam stultum quam credere in deum natura? A. - For this M P offer: non erit iam stultum credere in deum natura.That this is a manufactured text is clear- the cause

1. I connect et and si and would change the punctuation.

2. M P has te invenisse, hardly correctly, as Tertullian as you know often for the same subject uses the infinitive for the Acc c. Infin.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


is obviously the corruption non for num, and it is just as clear, that this reading does not merely miss Tertullian's meaning, but rather expresses exactly its opposite. For that the words of the work relate to the fact of the birth of Christ is just what Tertullian wants to maintain.

        p. 441, 14    As an argument, that Jesus was not born, the opponent (Apelles) uses the story of Matth. 12, 46 ff., that Jesus on receiving the message, that his mother and brothers were outside and desired to speak with him replied: who is my mother? who are my brothers? Tertullian guards himself against such an interpretation and then adds: Solet etiam adimplere Christus, quod alios docet. Quale ergo erat, si docens non tanti facere matrem aut patrem (aut patrem om. M P) aut fratres quanti dei verbum, ipse dei verbum annuntiata maue et fraternitate desereret? - This omission also is certainly not accidental; in the gospel the text spoke only of the mother and the brothers. But it is also just as thoughtless; for Tertullian relates to the words docens non tanti facere etc. the command of Jesus in Matth. 19, 29, for the correct interpretation of Jesus' question: Quae mihi mater, qui mihi fratres. If Jesus himself taught to respect the word of God more highly than a father and mother and brothers, as he had already with 'God's word in the desert', how did the text become: Your mother, your brothers are here? - Here we have therefore to do with an 'improvement' of the text, which is not in this case caused by an already existing corruption..

2. Additions.

        p. 430, 5 As the sole reason why Marcion denies the actual human corporeality of Christ, Tertullian can only think that he fears that Christ would stop being God if he became truly man. Tertullian holds the opposite, that precisely therein lies the difference between God and man, that God can take another form without losing His own nature. Also the angels of the Old Testament which had appeared to men had remained angels, although they appeared perfectly man-


VI. Treatise: Kroymann.


shaped. If therefore, infers Tertullian following Marcion further, the angels of the Demiurge, the deus inferior, could do this, how much more could the Christ of the Deus potentior of Marcion? Then he goes on:

        Aut numquid et angeli illi phantasma carnis apparuerunt?  Sed non audebis hoc dicere. Nam si sic apud te angeli creatoris sicut et Christus, eius dei (ita A. - M P: eiusdem substantiae) erit Christus, cuius angeli (M P: angeli bis) tales, qualis et Christus.

        To escape from the vice in which Tertullian's logic has placed Marcion, one could say that the corporeality of those angels of the Demiurge was also just a phantasm. But Marcion will not venture to say this. For if the angels of the Demiurgen can be supposed to have a caro putativa, it follows that Christ with his caro putativa is the Christ of the Demiurge and not the Christ the Deus potentior of Marcion. It would make the basis of Marcion's teaching uncertain. Only the reasing of the Agobardinus corresponds to this thoroughly clear thought: eius dei erit Christus, while the eiusdem substantiae of M P is altogether nonsensical. The word substantiae has been interpolated because eius dei in eiusdem had been corrupted. The doubling of angeli in M P certainly gives the correct sense, but is however not only unnecessary, but also rather ugly.

        p. 443, 21   -  -  ceterum quid est sanguis quam rubens humor, quid caro quam terra in figura sua (ita A. - P M conversa in figuras suas). Tertullian wants to represent, that the human body reflects its origin from the moist earth (limus): the blood points at the moisture, the flesh at the earth, and if also the appearance (species qualitatis) is another, one may thus call the blood 'red moisture', the flesh 'earth in its nature'. The idea in figura sua means the same as in suo genere, sua in effigie (de an. 9). Therefore the reading of the Agobardinus gives adequate expression for the thought. What on the other hand terra conversa in figuras suas might mean is not apparent to me. That we have to do with a third interpolation, perhaps caused by the corruption figuras suas, needs no further discussion.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


        p. 440, 9    Poterat enim evenire, ut quos illi nuntiabant foris stare (stare om. A), ille eos sciret absentes esse. - I believe that the stare appearing in M P is not original, for the esse after absentes can be supplemented by foris.

        In some cases in M P an effort can be observed to strengthen an expresssion by additions which reinforce the sense:

        p. 433, 20   Crucifixus est dei filius; non pudet, quia pudendum est. Et mortuus est dei filius, prorsus (prorsus om. A) credibile est, quia ineptum est. Et sepultus resurrexit; certum est, quia impossibile est. - The prorsus added in M P, which the editors have allowed into the text, in my opinion unpleasantly impairs the beauty and power of these words, which contain a clear escalation (non pudet - credibile - certum).

        p. 443, 14   Praetendimus adhuc nihil, quod ex alio acceptum sit, ut aliud sit quam id, de quo sit acceptum, ita in totem (in totem om. A) aliud esse, ut non suggerat, ende sit acceptum. The addition 'in totem' is not only superfluous, but rather strongly illogical. - Entirely barbarous however is
the following reinforcing addition:

        p. 437, 15    Mutuum debitum est inter se (inter se om. A) nativitati cum mortalitate, which I do not need to justify.

3. Variants.

        For clarity I organize this into three groups: 1. those which prove to be arbitrary changes, 2. those which give possibly the original text, 3. those which may be received with security into the text.

        1. Firstly it should be noted, that generally there is a tendency in MP to replace unusual forms and constructions by the usual. For example the substitution of Tertullian's so familiar Futurum instead of Potentialis throughout the present text, e.g. p. 433, 20 and p. 434, 12 est instead of erit, p. 439, 9 licet instead of licebit, and the substitution of the unusual singular predicate by several subjects in the plural, so p. 425, 10 retractantur for retractatur, p. 437, 27 annuntiarentur for annuntia-


VI. Treatise: Kroymann.


retur where I consider both times the singular to be probably admissible. To the same cause belongs also the change of the indicatives into the subjunctive after cum causale p. 442, 13 cum interpretentur where the indicative is good 'Tertullianish' and on the other hand changed the subjunctive back into the indicative after quod p. 435, 13 aspicite, inquit, quod ego sum statt sim.1 The following changes must be judged in the same way: p. 439, 7 Primo quidem nunquam quisquam adnuntiasset illi (sc. Christo) matrem et fratrem eius foris stare for stantes (A). p. 438, 17 Sed etsi de materia necesse fuisset (A fuit) angelos sumpsisse carnem, perhaps also p. 439, 15 alias non tacens, cum quid temptationis gratia factum est circa (A erga) eum.

        Meanwhile the changes of this type are comparatively harmless compared with those which affect entire sentences and sometimes totally obscure the thought of the author. A few examples follow:

        p. 426, 7   Sed et qui carnem Christi putativam inducit, aeque potuit nativitatem quoque phantasma confingere. For this quite comprehensible text M P offer : nativitatis

1. For the subjunctive after quod see ad martyras 4, apolog. 7. - by the way it seems that here already the translation of Jerome [the Vulgate edition of the Bible] - I am aware that I've utilized an incorrect expression - has influenced which has the indicative. Possibly it is also from this that the quia intruded in the words quia Spiritus ossa non habet and almost certainly p. 433, 6 the sapientes instead of sapientia (ut confundat sapientia A). Here I touch on a very important question, for it concerns the recovery of Tertullian from the Itala. I have naturally given this question detailed attention, but I must regret to say however, that it seems impossible to me, here to arrive at a sufficiently established judgment. The transmission of the 11th century can only be compared with the Agobardinus only by this piece of de carne Christi and by de praescriptione haereticorum, which exist in the Paterniacensis. Both works however contain very few literal citations, and moreover it is not for me at all certain, whether the Agobardinus itself is entirely free of correction of the citations. So far however I see the corrections already further advanced in M P and entirely in the manuscripts of the 15th century, i.e. in H F. So then this anyway already very complicated question through the state of our manuscripts becomes still more intricate, and Corssen (Bericht über die lateinischen Bibelübersetzungen, p. 13) is shown to be unfortunately right in his fear, that there would be not much to hope from the new edition with respect to this point also.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


quoque phantasmata confingere. The change is obviously evoked through the not understood, doubled accusative after confingere. What he who made this change, thought was intended under the plural phantasmata, cannot be seen.

        p. 427, 1 Marcion removed the entire birth and childhood story of Jesus from his gospel. In irony Tertullian has promoted this rejection: Sed nec circumcidatur infans, ne doleat, nec ad templum deferatur, ne parentes suos oneret sumptu oblationis, nec in manus tradatur Simeoni, ne senex moriturus exinde contristetur (ita A. - M P: ne senem moriturum exinde contristet). The readings are apparently not substantially different in sense and yet the irony is in the version of A far more cutting; for in truth the old man, when he saw the child, did not sorrow, but rather rejoiced, and this contrast appears in the expression only in the reading of A. With only this change, the text better resembles the original story.

        p. 432, 3 Marcion rejected the birth and the fleshly corporeality of Christ as unworthy of a God. Tertullian holds the contrary, that it is just this human Christ, born and fleshly, who redeems, therefore has loved, and then goes away: Amavit ergo cum homine etiam nativitatem, etiam carnem eius. Nihil amari potest sine eo, per quod est id (A et) quod est. Aut aufer nativitatem et exhibe hominem, aut (aut om. M P) adime (A adhibe) carnem et praesta quem redemit. Incorrectly, all publishers have abandoned the reading of the Agobardinus 'et' for the 'id' of the younger tradition, which to be sure reads more smoothly. For here the subject under discussion is not 'nativitas', but rather 'caro'; the first is expressed through the words 'per quod est', the latter through 'quod est'. The reading 'id' however means that the subject would no longer be 'caro'. On the other hand the second 'aut' has wrongly crept into the Agobardinus. For the first bit means: or (if you deny this) eliminate the birth and create a human being. By this 'aut' the opposition is placed before the consequence of his view which is illogical.. Both clauses are therefore  


VI. Treatise: Kroymann.


equivalent to each other and not mutually exclusive, as would happen with a second 'aut'. M P also correctly have 'adime' for 'adhibe'.

        p. 432, 18    Quaenam haec stulta sunt? Conversio hominum ad culturam (ita A. - M P: hominis ad cultum) veri dei, reiectio erroris, disciplina iustitiae, misericordiae; innocentiae omnis? Haec (M P innocentiae? Omnia haec) quidem stulta non sunt. - The word 'cultura' is found in Tertullian rather rarely for 'cultus', as it seems to be needed first by the church writers for the meaning of 'Divine Service'. It is here therefore a less common word replacing the more usual. In the last place, the not so easily intelligible omnis is changed into omnia and then connected with the following haec. The reading seems acceptable at first sight. Meanwhile it is clear, that, Tertullian in order not to have to individually itemize further Christian virtues wants to finish with a comprehensive expression. A similar expression appears in de pud. 20: sanctitas omnis.

        p. 433, 5    Sunt plane et alia tam stulta duae pertinent ad contumelias et (et om. A) passiones dei. Aut prudentiam dicant (M P  dicam) deum crucifixum. - The 'et' before 'passiones' is missing in A; it does not seem me impossible, that 'passiones' is only a gloss for contumelias which has penetrated into the text,
which M P has then connected with 'et', as earlier 'testimonium redderent' was with 'responderent'. In any case the reading 'dicam', which M P offer and probably understood as a question, is incorrect. For the 'aut' has here the same effect of rejection as above in the words: aut aufer nativitatem.

        p. 434, 4    If birth and carnal corporeality are unworthy of a God - so infers Tertullian taking the sense of Marcion further - how much more is crucifixion, death, funeral? But stulta mundi elegit deus, and who believes in this, is bene imprudens and feliciter stultus. Crucifixus (M P natus) est dei filius; non pudet, quia pudendum est. Et Mortuus est dei filius, credibile est, quia ineptum est. Et sepultus resurrexit; certum est, quia impossibile est.The reading natus for crucifixus cannot possibly be an accidental corruption. However how it could come to replace the correct crucifixus, is not easy to say.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


Perhaps it was caused by the question on p. 433, 8: Quid magis erubescendum nasci an mori. If so, it would certainly be the peak of thoughtlessness. It seems to me more probable that in the original of M P the participle 'crucifixus' had fallen out and then the space was filled at random.

        p. 433, 22    Christ is equally true man as well as truer God. Quae proprietas conditionum, divinae et humanae, aequa utique naturae cuiusque (cuiusque om. M P) veritate dispuncta est eadem fide et spiritus et carnis (caro M P). Virtutes spiritus dei deum (spiritum dei M P), passiones carnem hominis probaverunt. The reading 'caro' and the omission of 'cuiusque' prove to be intentional changes. He that did this understood 'natura' in its usual general meaning and not in the special meaning used here of 'conditio' (sc. humana et divina), and so could begin nothing with the 'cuiusque'. Thus the second change 'caro' became necessary; the sentence 'eadem fide et spiritus et caro' became independent, the preceding parallel phrase/thought therefore supplements the copula. Yet the correction extends still further, however this time, I believe, deriving from an actual mistake in the tradition. The sentences: 'virtutes Spiritus dei deum' - and 'passiones carnem hominis probaverunt' are formed irregularly. Undoubtedly the intention is to prove that Christ was 'vere deus' and 'vere homo'. The first phrase corresponds to that perfectly. The accusative 'deum' in the second phrase must however necessarily correspond with an accusative 'hominem'; instead of this we have 'carnem hominis', and on the other hand we lack a genetive for 'passiones', which should correspond to the genetive 'Spiritus dei' after virtutes. The corruption lies therefore in the second clause, and without any doubt it should read passiones carnis hominis <hominem> probaverunt. He that changed the text in M P, perceived very correctly the intolerable incongruity of the sentences; however because he thought the second phrase was intact, his correction made the damage worse than before.

        p. 442, 14    The followers of Apelles have it that the world was created through an 'angelus inclitus', which however then felt regret over its work: Teste igitur paenitentia institutoris 


VI. Treatise: Kroymann.


sui peccatum (M P delictum) erit mundus, si quidem omnis paenitentia confessio est delicti, quia locum non habet nisi in delicto. The change delictum for peccatum has this obvious reason, that later in the sentence delictum appears twice. In the same arbitrary way nam est nobis et ad illos libellus in adversus illos is changed, probably because in the titles of the polemical writings adversus appears everywhere.

        p. 442, 19    Caro igitur Christi de caelestibus structa de peccati constat (constitit) clementis, peccatrix de peccatorio censu, et par (A pars) iam erit eius substantiae id est nostrae, quam ut peccatricem Christo dedignantur inducere (ita A. - Christus dedignatur induere M P). Oehler has wrongly allowed into the text the reading of the Agobardinus pars, which is totally senseless, but correctly given preference at the end of the reading to the Agobardinus. For Christ himself did was not ashamed to take on human flesh. The variant of M P might perhaps derive from the corruption induere for inducere. - In the remaining, I consider the words id est nostrae, which interrupt the connection between the relative sentence and its related word in most intolerable manner, and which are very obtrusive moreover to be a gloss. - Equally senseless is the following:

        p. 443, 19,    where for utriusque originem elementi (A) read utrumque originis elementum.

        2. I come now to those variants, where it is doubtful whether we are dealing with an arbitrary change or with a better tradition.

        p. 429, 1   If Christ was not born, he should not give the appearance that he was born. Quid tanti fuit, edoce, quoll sciens Christus quid esset, id se (M P: esse; probably a simple corruption) quod non erat exhiberet ? Non potes dicere: ne, si natus fuisset et hominem vere induisset, deus esse desisset, amittens quod erat, dum fit (MP assumit) quod non erat. - Next we must see if this deviation could be a gloss; however then this could only be fit, and we would have in M P the correct reading. However we may presume the contrary, that assumere actually refers to something other than fieri, and that it is quite


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


unsuited for the context. So one must only conclude it necessary, if one becomes something totally different, and not however, if one merely comes across a new form of something earlier. In this way one see the 'assumit' rather as an attempt to supplement an original lacuna.

        p. 432, 5   Si haec (sc. caro et nativitas) sunt homo, quem deus redemit, tu haec erubescenda illi facis, quae redemit, et indigna, quae nisi dilexisset, non redemisset. Nativitatem reformat a morte regeneratione caelesti, carnem ab omni vexatione restituit, leprosam emaculat, caecam reluminat (M P perluminat) paralyticam redintegrat, daemoniacam expiat (A captat), mortuam resuscitat, et nos illam erubescemus? (ita A. - M P: et nasci in illam erubescit?)   That the reading reluminat deserves preference over perluminat is clear. Conversely M P with expiat certainly gives us the original text instead of the corrupt captat. On the other hand the decision about the last variants is very hard. For by themselves both readings are quite comprehensible. Meanwhile there are serious objective reasons to say that Tertullian wrote what we read in M P. For the change of the subject in A is especially noticeable, because Tertullian goes on: Si revera (sc. Christus) de vacca aut sue prodire voluisset; cf. also the preceding: tu illi haec erubescenda facis. Moreover it is noticeable that Tertullian says nos illam erubescemus, where we should expect tu illam erubesces. And finally while with illam we understand only the preceding caro, not also nativitas, as suggested above, on the other hand through the nasci in illam both are expressed. After all that it still seems to me that the reading of MP deserves preference. - That also in the Agobardinus there exist obscurings of the tradition which are not purely mechanical we have proved already above with the falsely inserted text on p. 432, 6 and the change carnem hominis on p. 433, 27, and will be still further confirmed. At this point it is natural to seek the occasion of this crucial change, the primary corruption nos for nasci.

        p. 436, 19    Christ could become man without birth, says Apelles because the angels also did just that:


VI. Treatise: Kroymann.


Agnoscimus quidem ita relatum, sed tamen quale est, ut alterius regulae fides ab ea fide, quam impugnat, instrumentum argumentationibus suis (ita M P. - A argumentorum) mutuetur? - The reading of the younger delivery is so tempting, that it is received into the text by every editor. This still leaves good grounds, in my opinion, to defend the reading of A as original. Let us consider the dative argumentationibus suis, thus the neighbouring instrumentum would generally be understood as 'the equipment'. Then however we must expect unconditionally the plural instrumenta. However it is known that in Tertullian instrumentum can also mean 'the writing', and that seems to suit this place. But it only seems so, for in reality Apelles did not 'borrow' indeed 'the writing' from his opponent, but only from it the rationale for his heretical ideas. Therefore he borrowed the belief that he makes the apparatus of his rationale, i.e. in Tertullian's language: instrumentum argumentorum, exactly as he speaks in apolog. 17 of an instrumentum elementorum: Quod colimus deus unus est, qui totam molem istam cum omni instrumento elementorum de nihilo expressit. - Hereafter I would like to follow therefore A and consider the reading of M P to be an arbitrary change, which had its origin in understanding the changing instrumentum as 'craft thing' and then postulating a dative.

        3. Finally there is one more case, where it seems to me that the text has been arbitarily changed, not in M P, but rather in A.

        p. 433, 21  Tertullian has placed before his opponent the question, whether he would also hold that the crucifixion, the death and the resurrection of Christ were a phantasm. If that is so, he goes on, our belief is an illusion and our entire hope a delusion. Scelestissime hominum, qui interemptores excusas dei (nihil enim ab eis passus est Christus, si nihil vere est passus),1 quid2 destruis necessarium dedecus fidei? Quodcumque deo indignum est, mini expedit. Salvus sim (thus A correctly)

1. I consider this change to punctuation to be necessary.

2. So I write with A. - M P qui which the editors give.


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


si non confundar de domino meo: qui mei, inquit; confusus fuerit, confundar et ego eius. Alias non invenio materias confusionis, quae me per contemptum ruboris probent bene imprudentem et feliciter stultum. Crucifixus est dei filius; non pudet, quia pudendum est. Et mortuus est dei filius; credibile est, quia ineptum est. Et sepultus resurrexit; certum est, quia impossibile est.  Nam (ita A. - M P Sed haec) quomodo vera in illo erunt, si ipse non fuit verus, si non vere habuit in se, quod figeretur, quod moreretur, quod sepeliretur et resuscitaretur, carnem scilicet hanc sanguine suffusam, ossibus structam (MP substructam), nervis intextam, venis implexam, quae nasci et mori novit, humanam (A: humana) sine dubio, ut natam (A: nata) de homine. Ideoque mortalis haec erit in Christo, quia Christus homo et hominis filius. - I have striven in vain, to find a logical connection between the reading of A: nam quomodo vera in illa erunt and the preceding, and cannot grasp just how Oehler was able to follow the Agobardinus. That this sentence can contain the basis for the preceding, is impossible. Or who understands then such connection: - 'and the son of God died; it is believable because it is vulgar. And he revived from out of the grave; it is certain because it is incredible. For how can this be actual in him, if he himself was not real?' On the other hand the connection is thoroughly intelligible, if we read with M P: Sed haec quomodo in illa vera erunt? For Tertullian wants to say: that Christ crucified, died and buried, is incredible and impossible; but for just that reason I believe it. For God's plan of salvation is deliberate foolishness. 'But how,' says he next, 'can these (namely crucifixion, death, funeral) in him be actual, if he himself was not real?' - Therefore: the foolishness of the divine plan of salvation cannot make me waver in my belief, regardless of this phantasm which my opponent has proposed in order to improve the foolishness of God. For if Christ himself and his life and sorrows were not real, my belief is also a delusion. It is possible to so understand the place only, if we read with M P: Sed haec, whereby we also give to vera a clear subject.


VI. Treatise: Kroymann.


But how then does the nam come into A? There it cannot emerge through simple corruption, so there is no alternative but to regard it as a failed attempt to supplement a lacuna in the original. - In the following I consider also that the reading substructam of M P is correct, on the other hand the last part in M P has been trimmed, on the assumption that the sentence is complete. However undoubtedly after implexam, a question mark should be placed and then read with A: Quae nasci et mori novit, humana sine dubio, ut nata de homine, ideoque mortalis haec erit in Christo, quia Christus homo et hominis filius.

        In the preceding discussion, I believe we have gathered essentially all the material coming into consideration for our question. Now the question, whether in the Cluny tradition more exists than a bare collection of Tertullian's writings, must in my opinion be answered in the affirmative. The deviations, omissions and additions are so numerous and peculiar that it is impossible to regard it as the sum of gradual accumulation of corruptions in the course of time. The collector of the corpus has rather subjected the texts of the individual writings handed down to him (for one may be allowed to generalize unobjectionably the judgment) to a correcting examination and has proceeded on the basis of his criticism his own, evidently quite insufficient understanding of the author, which has misled him, in a very arbitrarily manner to athetise, interpolate and make conjectures. To judge by the preceding samples, he might at the same time have used scarcely more than a single manuscript, and in part in any case the tradition serving as his model was of the same family as that of the Agobardinus, which for its part perhaps represents the first attempt, to prevent the loss of the threatened writings of the African, about which Jerome already complains, through making a collection. If the Corpus Tertullianeum we have received really represents a portfolio of works united by the collector - and this question we can neither affirm nor


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


deny - it is certain that he did not know of the existing collection in the Agobardinus. For the corpus only contains a slight portion of the works in that collection, and it is difficult to see, why he would have omitted the rest. A further proof, that the writings united in the Corpus until then led an independent life, is the remarkably different state of preservation. For if one compares e. g. the writing de exhortatione castitatis, that we have otherwise only in manuscripts of the 15th century, against the Agobardinus, the difference is so terrifying - I count in M F in this little writing not less than 92 large and small lacunae - that one sees that the collector of the corpus has received this writing in a already much worse state than is the case for our test work, de carne Christi. And in the same way interpolations and arbitary changes appear. That however the extra three centuries of transmission did not add further to the corruption of the text, may be seen by comparison of other writings of this corpus with the Agobardinus, e.g., De monogamia. In order to summarize therefore, what has been demonstrated on the basis
of the earlier observations, the following state of affairs seems probable: after one attempt had already been made to rescue the works of Tertullian from loss by means of a collection, another man, probably in Gaul, ignorant of that first attempt, has united in a Corpus all the works available to him, which until then had been transmitted in isolation, organised it on an impersonal basis and subjected the text to a thorough revision. The credit for making this corpus available to posterity may be attributed to the Cluniac order. However meaningful the fact of the collector is, equally frankly to him must be attributed the fact, that the condition of the text transmitted to us is lamentable, not only through the misfortunes of transmission, but rather also through the arbitrariness of the editor, although by differing degrees in the different writings. - The criticism of the corruptions of the original wording, where the Agobardinus cannot be requisitioned as a control,


VI. Treatise: Kroymann.


will in most case be impossible, and also, where a corruption makes itself perceptible, the healing will be that much harder to manage, the deeper the wave of corruption has soaked in. So then the critic finds himself vis-á-vis this text in the unedifying position, that he certainly knows that in this situation a 'conservative criticism' is inappropriate, and that nevertheless he must be doubly careful not to lose his footing. Nevertheless the text corruptions emerging from interpolation and glosses will be most easy to remove, and moreover the quantity of lacunae, which distorts this transmission, must be considered, particularly if one does not have the means to certainly produce the lost wording.

        Regarding the Agobardinus, it remains true that it delivers the foundation for the criticism of all its writings. I believe that I have shown in several examples that by comparison the other delivery comes into consideration not only to supplement its numerous lacunae. In any case each variant must be most carefully tested, as this transmission also is not completely free of arbitary changes and additions. In this respect, Oehler's method of swaying helplessly between both transmissions can only serve as an example of what one should not do.

        Still more difficult and ungrateful becomes the task of the critic who has only the first edition of Rhenanus (Hirsaugiensis and in addition the Florentine manuscript F) and the other Florentine manuscript N, the apograph of M and finally the sparse remains of the Gorziensis in Rhenanus to form his critical base text. For it is obvious, that the three century younger tradition did not pass the text along tracelessly. With the availability of the manuscript N, which I have for the first time entirely collated, it is possible to make not to be underestimated progress as against the above witnesses. One needs merely look at the readings given on p.4-6 to see that when compared with H F, that N is a decidedly more trustworthy witness of the text than H (F)


Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


if also, namely because of the lacunae, no manuscript can [[entraten]] the other. That in H the corruption is further advanced, is as I noted above, probably because that there was another intermediary between the joint archetype of itself and M H, perhaps the Divionensis repeatedly noted by Rigault.1 It is a designating characteristic of the Hirsaugiensis, that in it unusual forms are replaced by the usual, as above: praecinuerat for praececinerat, araneae for aranei, and it is quite possible, that Rhenanus has placed too much confidence in the corrector of H, who he specifically attests. This corrector extends itself, as I have noted elsewhere, namely in Bible citations, and he did not refuse himself three interpolations.2 All the same it is clear that in the critical apparatus of Vol. IV
the manuscript N may claim the first place. Meanwhile it may not be concealed that also this manuscript, like the other, has not received unhesitatingly into the text the corrector of M as more authentic (cf., above p. 12), moreover it also shows a lot of things,

1. Oehler praef XIX and XX correctly emphasizes, that the MS contained much more than Rigault says in his introduction. There can be no doubt that it, like Gorziensis, Hirsaugiensis and N, contained the complete corpus. As far as I could see in Oehler, its incidental readings agree unanimously with the Hirsaugiensis. Yet this question needs still further investigation.

2. This chapter 7, end. Here we read in A M P N: Ceterum ad negandam nativitatem alius fuisset ei locus et tempus; in F on the other hand: alius necessarius fuisset, and this interpolated necessarius also was in the Hirsaugiensis, as is proven by Rhenanus' marginal note. Also glosses have here penetrated more frequently into the text than in M P N, so p. 437, 9: comparent velim et causas, ob quas in carnem venerunt processerunt, which Rhenanus notes also for the Hiersaugiensis, and worse still adv. Marcionem 25 (p. 77, 30), where the entire sentence: quis volet quod non concupiscet is without a doubt a gloss on the preceding words, which express exactly that in more precise form. In F, the interpolations are by the way still more numerous than in H, e.g. p. 431, 16: amavit utique quem magno pretio redemit, where the word pretio is missing not only in M P N, but rather also in Rhenanus; just like p. 433, 18, where F Et falso enim against the simple Falso offered by all in the other tradition.


VI. Treatise: Kroymann.


which probably might lead back to the corrector of the scribe himself.

        So p. 427, 3, where Marcion is addressed: Nam et mortuus es, qui non es Christianus, non credendo quod traditum (om. N. - creditum H F) Christianos facit. That the traditum so no sense yields, both the scribe of N and that of the Hirsaugiensis recognized. The latter changed it into creditum, the first however simply left it out. Neither helps, the real correction is found by transposition: non credendo traditum, quod Christianos facit. Compare the shortly preceding words: Si tantum Christianus es, crede quod traditum est. The inclination to leave out in order to remove the problem can also be shown elsewhere, e.g. p. 433, 15: Sed iam hic responde, interfector veritatis, norme vere crucifixus est deus? nonne vere mortuus est, ut (P M F : et) vere crucifixus? nonne vere resuscitatus, ut vere scilicet mortuus? - In the transmitted reading et the words et vere crucifixus were not only superfluous, but rather intolerable, and therefore the scribe of N instead of changing it into ut has entirely removed it. Also single words, which in spite of the clarity of the sense can be omitted, are often certainly not accidentally missing, in N, so p. 432, 17: Stulta mundi elegit deus. Quaenam haec stulta (stulta om. N)? The immediate repetition of the word stulta seems to have displeased the scribe. -p. 440, 9: - - ut quos illi nuntiabant foris, ille eos sciret absentes esse (esse om. N). In fact the expression seems to gain precision through omission of the esse. Also a supplement of the lacunose transmitted text, and indeed a correct one, can be demonstrated in N. p. 441, 9: Cum indignatio parentes negat (ita A N, negat om. M P H F), non negat, sed obiurgat. - The transmission of the Agobardinus proves that the scribe of N has encountered the correct text with his pushed-in negat. But unfortunately that in no way changes the reality, that N does not just impartially transmit its model, but rather, where it appears necessary, believes it must assist it through characteristic combination.

        The criticism hereupon, if also N remains the base of the recensio, will always have to have a watchful eye

Critical Preparation for the III. and IV. Volume of the new Tertullian-edition.


and not trust too unhesitatingly to this manuscript. However the works in Vol. IV, based on such a miserable foundation of manuscripts, in each case remain those the furthest from the authentic wording of Tertullian, and the very worst, or so it seems to me, is de pallio, which has evoked so many controversies, not always handled without bitterness, with its entirely neglected transmission.


This page has been accessed by ****** People since 26th February 2001.

Return to the Tertullian Project Original article About thesis pages