Somerset Historical Essays/Appendix B
The Two Earliest Glastonbury Charters
The early Glastonbury charters have never been systematically examined. The texts are readily accessible in Birch's Cartularium Saxonicum; and, though it is possible that further research might add to their number or improve their quality, we have enough material already to challenge the attention of the diplomatic expert whose trained experience would enable him to discern the precious from the vile and establish for the historian and the topographer the reasonable certainty of facts which at present they must needs view with suspicion. We are being taught in other directions that it is not enough to say that a charter is genuine or spurious. Fragments of genuine charters are embodied in late copies which have been remodelled for reasons often hidden from us. The compiler of a chartulary will sometimes abbreviate his documents, change the spelling of placenames, recast the bounds, add the date of the Christian era—all for practical purposes and with no fraudulent motive of any kind. The Glastonbury charters offer several examples of the combination of two charters in a single document in the tenth century, perhaps for mere convenience of use in the courts of law. Fraud is a motive which seldom can be proved, though there are a few undoubted instances, where privileges are claimed or where legendary history called for corroboration. The critical study of our early English charters has been worthily initiated by the able editors of the Crawford Charters: already it has begun to bear fruit of historical value here and there. But Glastonbury is neglected still, and the purpose of this Note is to call attention to some exceptional features in two charters which suggest that its ancient muniments are peculiarly worthy of investigation. The two earliest of the Glastonbury charters are grants of land by K. Coenwalch and Bishop Haeddi, bearing the dates 670 and 680 respectively. It will be well to take the latter first, as there is reason to think that it is a more faithful representative of its lost original. Both charters are printed by Birch from the fourteenth-century Secretum of Abbot Monyngton in the Bodleian Library.
B. C. S. 47.
Bishop Haeddi grants to Abbot Hemgisl Lantokai (Leigh in Street) and an island: 680 (for 677).
Regnante ac gubernante nos domino nostro Jhesu Christo: mense Julio, pridie nonas, indictione quinta, anno incarnationis ejusdem.DCLXXX. Nichil intulimus in hunc mundum, verum nee auferre possumus: ideo terrenis caelestia et caducis aeterna comparanda sunt.
Quapropter ego Eddi episcopus terram quae dicitur Lantokai, tres cassatos, Hemgislo abbati libenter largior: nee non terram in alio loco, duas manentes, hoc est in insula quae giro cingitur hinc atque illinc pallude cujus vocabulum est Ferramere.
Denique solerter peto ut nullus post obitum nostrum hoc donativum in irritum facere praesumat. Si quis vero id temptaverit, sciat se Christo rationem redditurum.
Ego Eddi episcopus subscripsi.
Let us begin by looking at our earliest evidence as to the gifts of Bishop Haeddi, who ruled the undivided see of Winchester from 676 to 705. In the Liber Terrarum (J. of G., p. 370) the fifth entry runs thus:
Hedda episcopus de Lantokay, i. Leghe. dat. Glast. II.
This indicates that there were two charters dealing with this property which were copied into this ancient Land-book. One of these no doubt was the charter used by William of Malmesbury, who in the De Antiquitate (p. 50) writes thus, under the heading De Leghe:
Eodem anno Hedde episcopus Lantocay vi hidas, Kentuino eciam et Baldredo consencientibus, dedit Glastoniae: quam donacionem Cedualla confirmavit, et propria manu, licet paganus, signum crucis expressit.
This is plainly not the charter which we have before us, for it grants six hides at one place, as against three cassates in one place and two manses in another. William of Malmesbury has placed it 'in the same year' with Baldred's grant of Pennard, that is, in 681. This is probably an error, for it seems that this was the year in which Caedwalla was in exile and came into contact with Wilfrid: but it is possible that his confirmation was obtained later.
Our charter is probably that which is mentioned in the list of charters still preserved in 1247 (J. of G., p. 375):
Hedde episcopus de Lantokay et Ferremere. Hemgillo abbati.
Here we have the same combination of properties. When we look at the map we see that Leigh in Street is two miles south of Glastonbury, whereas Meare is nearly four miles to the north-west. It is possible that two grants are here combined in one charter: this would account for the different designations ('tres cassatos' and 'duas manentes') occurring in the same charter.
Nevertheless our charter presents very primitive features. Its first eight words are, as we shall see, identical with those of K. Coenwalch's charter (B. C. S. 25), and are such as we might expect. The addition of the year of the Incarnation may have been made when the charter was remodelled, say in the tenth century; or it may have been inserted by the compiler of the fourteenth-century Secretum. It agrees neither with William of Malmesbury's date (681), nor with the indiction, which points to 677. The brevity of the proem (Nichil intulimus, &c.) is in its favour, and the closer examination of it which we shall presently make will suggest that it retains its original form. This again is a point which links this charter to the charter of K. Coenwalch. There are other parallels of language between the two charters—such as 'libenter largior' and 'hoc donativum', though K. Coenwalch's charter is elaborate and fanciful where Bishop Haeddi's is brief and plain. But it is time to read what claims to be the earlier document.
B. C. S. 25.
K. Coenwalch grants to Abbot Beorhtwald Ferramere and two islands: 670 (for 671).
✠ Regnante ac gubernante nos domino nostro Jhesu Christo.
Nichil intulimus in nunc mundum, verum nee auferre quid possumus: ideo terrenis caelestia et caducis aeterna mercanda sunt.
Quapropter ego Ceduualla terrain quae dicitur Ferramere, unum cassatum, Beorhtuualdo abbati libenter largior; nee non duas parvas insulas; hoc est cum captura piscium in utraque parte stagni, cum paludibus, silvis, pascuis apium, et omnibus ad se pertinentibus dabo ei, ut habeat diebus vitae suae et post obitum suum cuicumque voluerit derelinquat.
Corroboravimus nunc crucisque signo confirmato hoc donativum stabili jure gratum et ratum decerno durare quamdiu vixero poli terras atque aequora circa aethera siderum jusso moderamine volvet. Si quis autem nisus fuerit hujus meae donacionis testamentum confringere aut adimere conatur, ipse acrius multatus sit infernalis ergastuli poena demersus, quern eo daemon vel diis dampnatorum paravit.
✠ Ego Coenuualla basilleos Westsaxonum propriae manus subscripcione sanctae crucis designavi effigiem, ut nemo qui se regeneratum in Christo noverit hujus largicionis donum mutare praesumat.
✠ Signum manus Theodori archiepiscopi.
✠ Signum manus Leuteri episcopi.
✠ Signum manus Hedde abbatis.
✠ Signum manus Aldhelmi abbatis.
Scripta est haec cartula privilegii anno incarnacionis Christi dclxx.
We must begin by attempting some emendation of this very corrupt text, which suggests that the charter from which this copy was made was in parts wellnigh illegible. First of all, 'Cenuualla' must be read for 'Ceduualla': the signatures make this plain. Then 'pascuis apium' presents a problem: I have nothing better to suggest than 'pascuis, aquis', a sequence which is found in some charters. For 'confirmato' in the last paragraph we must read 'confirmatum '. This we discover when we turn to K. Cuthred's confirmation of privileges (B. C. S. 169), which enables us to restore sense to the remainder of the sentence by emending the impossible 'vixero' to 'vertigo'. The whole passage from K. Cuthred's charter must be quoted:
… sieque propriae manus subscriptione crucisque signo confirmatum hoc donativum stabili jure gratum et ratum regum praedictorum decerno durare, 'quamdiu vertigo poli terras atque ecora circa ethera siderum jusso moderamine volvet'. Si quis autem hujus meae donacionis testamentum visus fuerit confringere vel gressum pedis nobis Hengissingum traditum urbemque glebam extra terminos prefixos vel definitos limites seu constitutes adimere, ipse acrius multatus sit infernales ergastuli in pena demersus violentiae suae presumpcionem luat in evum. Amen.
✠ Ego Cudredus rex Westsaxona propriae manus subscripcione sanctae crucis designavi efligiem, ut nemo qui se regeneratum in Christo noverit presumat mutare hanc donacionem.
Here we have an almost equally corrupted text, but the blunders of the one charter can to some extent be set right from the other.
We must now look to see what our earlier authorities have to tell us. The Liber Terrarum had a charter which is thus described (J. of G., p. 370):
Carta Cenuualli de Ferramere dat. Glastoniae.
This may point to a grant of Ferramere only. No charter of K. Coenwalch is recorded as existing in 1247. We turn, therefore, next to William of Malmesbury (De Antiq., p. 49):
Anno dominicae incarnacionis sexcentesimo septuagesimo Cenwald, qui et Kenuualchius, qui a Cerdicio septimus apud Westsaxones et per beatum Birinum in Christum credidit, anno regni sui xxix°, Berthwaldo abbati, interveniente Theodoro archiepiscopo, dedit Ferramere II hidas. 'Ego Theodorus subscripsi.' Dedit eciam idem rex Beokerie, Godenie, Martynesye et Andreyesie.
Here again we seem to have evidence of a charter granting Ferramere only. It was dated in the 29th year of the king's reign: this no doubt was the original dating, and the year A.D. 670 is prefixed by William of Malmesbury, probably on his own calculation; for he used a D E type of the A.S. Chron., which placed the accession of Coenwalch in 641 (not in 643 as in A). The signature 'Ego Theodorus subscripsi' further shows that his form of the charter differed from ours, which has 'signum manus'—itself an early feature.
He may have seen a separate charter which granted Beokery, Godney, Martinseye and Andredseye. The 'two small islands' of our charter would be two of these. Among the earliest requirements of the monastery would be the security of its fishing rights in Meare and the 'islands' of this marshy region.
Looking now at our charter again, we observe that it opens with the same words as B. C. S. 47 (Bishop Haeddi's grant), and has the same proem with but two variants—the insertion of 'quid' and 'mercanda' for 'comparanda'. Other parallels have already been noted.
Its most striking feature is the parallel with K. Cuthred's privilege (B. C. S. 169). From that or some similar charter the elaborate attestation of the king has been taken over, with the addition of ' basileus ', which suggests the tenth century. The contrast with the primitive attestations which follow ('signum manus …') is noteworthy.
We conclude that we have here fragments preserved of an original charter of the seventh century, which has undergone more than one modification. William of Malmesbury saw a form which recorded a grant made by K. Coenwalch to Abbot Beorhtwald in the 29th year of his reign. Accepting 643 as the year of the king's accession, we may date the grant in 671 or 672. There seems no reason to doubt the historical fact which is thus recorded.
We have now to consider a feature of these two charters which is of uncommon interest. The proem runs thus:
Nichil intulimus in hunc mundum, verum nee auferre [+ quid B. C. S. 25] possumus: ideo terrenis caelestia et caducis aeterna comparanda [mercanda B. C. S. 25] sunt.
The quotation from 1 Tim. vi. 7 does not, as we might have expected, follow the text of the Vulgate. For there we read:
Nihil enim intulimus in hunc mundum: haut dubium quia nee auferre quid possumus.
On the contrary, we have here an ancient form of the Old Latin version. Three times St Cyprian, who died in 258, quotes the text exactly as we have it in B. C. S. 47:
Nihil intulimus in hunc mundum, verum nee auferre possumus.
Twice it is quoted in the same way by St Paulinus, bishop of Nola, who died in 431. The fact is so curious, and the variants that we shall find in other charters so interesting, that it is worth while to add some further pre-Vulgate evidence.
Ambrosiaster: Nihil enim intulimus in mundum, verum quia nee auferre possumus quicquam.
Cod. Clarom. (d2): Nihil enim intulimus in hunc mundum, verum quoniam nee efferre (sic) possumus.
Cod. Boern. (g2, connected with St Gall): Nihil enim intulimus in hunc mundum, quod ('vel quoniam' added above the line) nee auferre aliquid poterimus.
Book of Armagh (A.D. 807): Nihil enim intulimus in hunc mundum, verum quia nee auferre quid possimus.
How are we to account for the presence of a very ancient form of OldLatin text in a West-Saxon charter of the seventh century? The liturgical books of the Church even at the present day retain Old-Latin texts in certain passages which have never been assimilated to the Vulgate. But I am not able to find that this particular verse has been anywhere so preserved in ordinary use. Others may perhaps be able to supply my defect of knowledge, and in that case our question would find a ready answer. Otherwise we must suppose the existence in Glastonbury or elsewhere in Wessex of a copy of the Pauline Epistles with a 'Celtic' or a 'mixed ' text, which presented the verse in this ancient form. I can but put the matter forward tentatively, in the hope that it may receive attention in the proper quarter. We may however regard the presence of this primitive reading in our two charters as prima facie evidence of an early date.
The interest of our proem is not yet exhausted. The same text and the same moral drawn from it appear in many other charters, though never again with the same simplicity of form. Thus the Pagham charter (B. C. S. 50), notorious for its unique phrase trimoda necessitas, has been fully discussed by Mr. W. H. Stevenson in the English Historical Review for Oct. 1914 (xxix. 689 ff.). It is written in what appears to be a Canterbury hand of the end of the tenth century; but it purports to be a grant made by Caedwalla, king of Wessex, to Bishop Wilfrid in a. d. 680. It begins thus:
✠ In nomine salvatoris nostri Jhesu Christi.
Nihil intulimus in hunc mundum, verum nee auferre quid poterimus: idcirco terrenis et caducis aeterna et caelestia supernae patriae premia mercanda sunt.
This may very well be an amplification of our proem, and derived ultimately from a genuine seventh-century Wessex charter. We have seen that 'poterimus' occurs in the 'Celtic' Cod. Boernerianus. It offers an unfavorable contrast to the terseness of the clause: ' ideo terrenis caelestia et caducis aeterna mercanda sunt.'
Exactly the same opening is found in B. C. S. 64 (except that it has 'possumus', not 'poterimus'), another grant by Caedwalla to Wilfrid, A.D. 683. This is a Chichester charter, drawn up in much the same language as the Pagham charter, but with an obviously impossible signature. Two Worcester charters (B. C. S. 187, 218) have the same amplification of the moral, but in a corrupted form: the former of these was supposed to be a contemporary document of a. d. 759, but it is now regarded as of later date (see Stevenson, loc. tit., p. 695 n.); the latter also claims to be of the middle of the eighth century, but we have only a copy of it in Hemming's chartulary.
Malmesbury has a special form of the text and its moral. Thus B. C. S. 58 (dated 681) begins thus:
In nomine domini dei nostri Jhesu Christi salvatoris.
Nichil intulimus (ut apostolicum confirmat oraculum) in hunc mundum, nee auferre quid possumus: iccirco terrenis ac caducis aeterna ac mansura mercanda sunt.
We note here the omission of 'verum'. B. C. S. 59 (680 for 681) has the same form, save that it does not omit verum'. Variations of the same form, with 'verum' omitted, are found in B. C. S. 70 and 279. The Malmesbury charters are for the most part quite untrustworthy.
There are three Worcester charters in which 'verum' is omitted (B. C. S. 164, 216, 701); but the omission does not seem to occur anywhere else. In these three charters the text is not followed by the moral: and this is the case with those to which we now go on to refer.
Abingdon, another home of forgery, has a peculiar form of the text. It prefixes Job i. 21: 'Nudus egressus sum,' &c, and for our text it reads:
… verum nee ab eo auferre quid poterimus.
This is found in B. C. S. 680 (Athelstan); 1058, 1080, 1169, 1171, 1172 (all Edgar).
Winchester occasionally has this form (B. C. S. 1114, 1149, 1230); but it occurs nowhere else. We have noticed that 'poterimus' is found in Cod. Boern.
We may add two isolated forms: B. C. S. 182, which has 'sed' for 'verum', and B. C. S. 206, which has 'veruntamen' (cf. Cod. Fuldensis).
The perusal and classification of these various forms may perhaps be of service to students of the chartularies in which they occur. I shall not venture to comment on them further than to say that we turn back with relief to the simple form in our two Glastonbury charters, confirmed in our belief that here more surely than in any of the rest we have the language of the seventh century.
- The 'two islands' are perhaps regarded as a part of Ferramere. This would explain the calculation of Ferramere as n hides in what seems to be a confirmation of this grant by Bishop Haeddi under K. Centwine in the charter already discussed.
- In G. R.3 (p. 29) the passage appears in a shorter form: 'Anno dominicae incarnationis sexcentesimo septuagesimo, Ceonwalh, regni sui vicesimo nono, dedit Bertwaldo Glastoniensi abbati Ferramere, duas hidas, archiepiscopo Theodoro interveniente.'
- I have to thank my friend Professor Souter for the correct texts of Pelagius and 'Ambrosiaster'.
- We might have supposed that the Vulgate, introduced by St Augustine and his companions, would have reigned without a rival in the Anglo-Saxon Church. But this was so far from being the ease that we find almost at once the phenomenon of 'mixed' texts: that is, either copies fundamentally Vulgate but with a large admixture of 'Celtic' (Irish) readings, or copies fundamentally 'Celtic' but corrected largely from the Vulgate. Professor Souter writes to me: 'With regard to mixed Celtic texts, I hold the view that they are Old-Latin revised here and there from Vulgate, and not Vulgate into which Old-Latin readings have been put by substitution.'
- The addition of 'quid' in B. C. S. 25 is perhaps due to familiarity with the Vulgate on the part of a copyist.