[Image: ]


  • Feature of the month: No.7 December 2009

Feature of the month: No.7 December 2009

Making sense of an inaccurate copy: the case of Alexander III's brieve of protection for Dunfermline Abbey, 1271 1

Professor Dauvit Broun, PoMS Principal Investigator


Only a small proportion of the documents used as sources of information in the database actually survive as 'originals'—that is, the very sheet of parchment on which the document was originally written and which was sealed and kept for posterity. Most survive only as copies: for example, it was common for major churches at some stage have copied most (but not all) their documents into a manuscript-book (or 'cartulary'). Medieval copies are, on the whole, fairly accurate (except that the earliest generations of copyists in the thirteenth century tended to omit the lists of witnesses at the end of a charter). A few documents, however, only survive in copies made in the seventeenth and eighteenth centuries, when charters were no longer part of a copyist's staple diet. It is also clear that medieval script and handwriting posed a serious challenge for clerks to decipher. As a result, some of these copies can be nonsensical in places. What is a historian to do about this? An additional challenge for these copyists was that charters of the twelfth and thirteenth centuries were almost always in Latin. They did not all find this easy to handle. Occasionally the copyist made such a hash of the text that you have to wonder if they could read Latin with any confidence.

The example (below) is one of the more elaborate brieves of protection in the name of Alexander III (1249–86). It was a type of document which could be obtained for a fee from the royal writing office, and had a standard core text that was common to all brieves of protection. It is likely that there were some particular threats or problems which led someone (or some institution, like an abbey) to seek the brieve of protection in the first place. If the text is fairly standard, however, then the nature of this difficulty can only be guessed. In this instance the text goes on at some length to make sure that those who are in debt to Dunfermline Abbey are made to pay off what they owe. Presumably Dunfermline Abbey was finding it difficult to collect its debts, and sought this brieve in the hope that it would make royal officials more ready to help them. But this did not mean that Dunfermline was being offered a free hand. Care was taken to explain that the royal officials should decide for themselves whether a claim against a potential debtor was justified, and assess what was reasonably owed to Dunfermline.

The Manuscript

The brieve survives in manuscript only in a copy of some old documents relating to Dunfermline Abbey that the copyist himself tells us were found in a long chest at the foot of the old timber cupboard in the gallery of Pinkie Castle (Musselburgh). Once the copy had been made, the documents—which included original charters of the twelfth and thirteenth centuries—were delivered to royal officials on 12 February 1624 or 1625. They have never been seen since. Presumably the copy was made shortly beforehand. It survives today among the muniments of Fyvie Castle as item 295 (in the bundle 289-297), and consists of a gathering of 6 folios. Alexander III's brieve of protection is the fourteenth document, and appears at the top of fo.4v.

The copyist struggled to read some of these medieval documents. In one case—a letter of Robert I of 17 September 1317—he gave up and left blank spaces. 2 Fortunately most of the charters survive in Edinburgh, National Library of Scotland Advocates' MS 34.1.3a, the medieval cartulary created by the monks of Dunfermline themselves not long after 4 February 1254, to which they added further documents throughout the medieval period. As far as these lucky ones are concerned we can safely ignore the inaccurate copy made in (or shortly before) 1624/5 and read the text in the much better copies in the medieval cartulary. There are, though, a few which were not included in the original cartulary, or were not added to it later. Alexander III's brieve of protection is one of them.

Unpublished Edition by J. Maitland Thomson

Alexander III's brieve of protection is interesting not only because it only survives as an inaccurate copy. It also lets us gain an insight into the working habits of the greatest scholar of medieval Scottish charters in the late nineteenth and early twentieth century: J. Maitland Thomson (1847–1923), who was Curator of the Historical Department at the Scottish Record Office (now the National Archives of Scotland). 3 He came across the copy in the muniments of Fyvie Castle, and made a typescript which survives among his unpublished papers (Edinburgh, National Archives of Scotland, GD 212/1/126 item 6). 4 Maitland Thomson—unlike the copyist in (about) 1624—knew medieval charters intimately, and was (we may guess) fully literate in Latin. In his typescript of the document he corrected the text—but he did not show what he had changed. As we will see, the process of correcting an inaccurate copy is not always straightforward. There is always a risk that this can be taken too far. If every change is not noted carefully, a 'correction' could result inadvertently in the loss of an important detail in what was actually said when the document was first drawn up and sealed.

In what follows I will focus on two aspects of the text that has been bequeathed to us. The first are obvious errors in the Latin which are unlikely to have been committed by the scribe who originally wrote the brieve in 1271. These are almost certainly mistakes by the clerk who copied the brieve into the manuscript that survives from (about) 1624. The second are words that Maitland Thomson changed because (presumably) he regarded them, too, as copying mistakes. In these instances, however, there is (at the very least) some basis for arguing that the copy could be representing the original text of 1271 accurately.

Edited Latin Text

In the edition of the text of the brieve (below) the use of bold signifies words which seem to have been miscopied in a way that changed the meaning (sometimes making the sentence impossible to understand). The word in bold is the 'correct' version; the scribe's mistaken rendering of each of these is discussed below. There are also a few unusual spellings, but these have been retained, as they do not affect the meaning.

This brings us to the way the text was treated by Maitland Thomson. He, not unnaturally, corrected obvious errors: this includes all the words in bold which the copyist tripped up on. But he also made other 'corrections' which may not have been warranted (even though the Latin was improved by his interventions). In these cases the forms and spellings that are found in the manuscript have been retained, and each word starred for ease of reference. The same procedure is followed in the translation that follows: bold is used to indicate where the translation follows the Latin words in bold, and starring shows where Maitland Thomson made a change, but where the translation sticks to the text as it appears in the copy of (about) 1624.

Alexander Dei gratia rex Scotorum omnibus probis hominibus totius terre sue salutem. Sciatis nos abatem et conventum de Dumfermleine terras suas homines suos et uniuersas eorundem possessiones et omnia bona sua mobilia et immobilia tam ecclesiastica quam mundana sub firma pace et protectione nostra iuste suscepisse. Quare firmiter prohibemus ne quis eis malum molestiam iniuriam seu gravamen aliquod inferre presumat iniuste super nostram plenarium forisfacturam. Concessimus eisdem abati et conventui ut nullus namos suos uel hominum suorum capiatt pro alicuius debito plegiagio uel forisfacto, nisi pro eorundem proprio debito pleggiagio uel forrisfacto saluis burgis nostris: firmiter *inhibentes* ne quis eos contra hanc concessionem nostram vexare presumat iniuste super nostram plenariam forisfacturam. Mandamus insuper et firmiter precipimus iusticiariis uicecomitibus prepositis et eorum balliuis ad quos presentes littere *pertenerint* ut omnes illos in eorum balliis seu burgeis qui debita debent eisdem abbati et conuentui ad eadem debita eis uel certo eorum attornato latori presentium iuste et sine dilatione reddenda sed in quantum ipsi uel dictus eorum actornatus dicta debita sibi deberi ab eisdem rationabiliter probare *poterint* uel poterit coram eis prout iustum fuerit compellant ita ne pro eorum defectu amplius inde iustam querimoniam audiamus. *Testatur Johannes* Cummynge *presente* Willelmo de Sancto Cleir et Johanne de Fentoun apud Newbottill. Tertio die Octobris anno regni nostri uicesimo tertio.


Alexander, by the grace of God king of Scots, to all worthy men of his whole land: greeting.

Let it be known that we have justly received the abbot and convent of Dunfermline, their lands, their men and their entire possessions, and all their goods, moveable and immoveable, ecclesiastical as well as worldly, under our firm peace and protection. For this reason, we firmly forbid that anyone should dare unjustly to inflict a wrong, an annoyance, an injury or some other trouble on them, on pain of our full penalty.

We have granted to the same abbot and convent that no-one may seize their goods, 5 or the goods of their men, for someone's debt, pledge or penalty, unless for their own debt, pledge or penalty, except for our burghs; firmly *forbidding* that anyone should dare to trouble them unjustly against this grant, on pain of our full penalty.

Moreover, we order and firmly command justiciars, sheriffs, provosts and their officers to whom the present document *would apply*—so that we do not afterwards hear further just complaint for their failure—to compel all those in their sphere of operation or burghs who owe debts to the same abbot and convent to repay the same debts to them or to their appointed representative, bearer of this document, justly and without delay—but only as far as they or their said representative *will have been able* reasonably to prove in your presence that the said debts are owed to them by those [from whom the debts are claimed], according to what is just.

*John* Comyn *bears witness, with* William of St Clair *present*, and John of Fenton. At Newbattle, on the third day of October in the twenty-second year of our reign [i.e., 1271].

Copying errors

Let us look at the words in bold. What did the copyist actually write down?

mundana: the manuscript reads 'undana', which is nonsense.

abati et conventui ('to the abbot and convent'): the manuscript reads 'abate et conventue'. Abate literally means 'by/with/from the abbot', which is clearly inappropriate; 'conventue' is not even a recognised form of the word conventus.

presentes ('the present'): the manuscript reads 'presentis', which literally means 'of the present', 'of the [thing] to hand'.

balliis: 'bailiwick', 'sphere of operation': the manuscript reads 'balliuis', meaning 'bailies', 'officers'.

qui ('who'): the manuscript reads 'cui', meaning 'to whom'.

latori ('the bearer', connecting this grammatically with 'attorno', 'to the representative'): the manuscript reads 'latorum', meaning 'bearer', but without any grammatical link with 'attorno'.

quantum ('as far as'): the manuscript reads 'quod', meaning 'which', 'that', and also 'although', 'because', none of which make sense here.

audiamus ('we might hear', using the subjunctive form that would be expected in this clause): the manuscript reads 'audiemus', meaning 'we will hear'.

Before we wag a finger too briskly at the copyist for his negligence, it is important to recognise that medieval scribes themselves could be flexible in their spelling, and that this could even make words appear ungrammatical. 6 It is possible, therefore, that some of these 'mistakes' were in fact penned by the original scribe of the document in 1271. How are we to tell whether the copyist of the surviving manuscript has dutifully recorded the exact spelling as he found it, or whether he has made a mistake? There are some words, however, which must be miscopyings: the original scribe would hardly have written 'undana' for mundana, 'latorum' for latori, and 'quod' for quantum.

There are also some strange spellings where a letter has been doubled unnecessarily. Again, it is conceivable that this was the work of the original scribe, rather than a blunder by the copyist. 7 The examples are:

'capiatt' (for capiat)

'pleggiagio' (for plegiagio)

'forrisfacto' (for forisfacto)

Another unusual spelling is 'burgeis' (for burgis). 8

Unwarranted 'Corrections'

Finally, what about the starred words—those where I have not 'corrected' the manuscript-copy, but which Maitland Thomson thought were mistakes? The first is *inhibentes* ('forbidding'), which Maitland Thomson 'corrected' to read prohibentes, which would also be translated as 'forbidding'. Presumably he made this change because the verb prohibere is much more common in this context; inhibere has a wider range of meaning (e.g., 'to restrain', 'to curb', as well as 'to forbid'). There can be no doubt that inhibere is feasible here, though, because it is used in the same context in other brieves of Alexander III. 9 One is almost identical to our document (it is a long brieve of protection for Coldstream: it is H 1/8/43 in the database). Others are similar: one concerns the recovery of debts (H 1/8/3 in the database), and the other is a shorter brieve of protection (H 1/8/24). 10 These all survive today as the original piece of parchment that was sent out by the king's writing office.

In the case of *pertenerint* ('[they] would apply'), Maitland Thomson thought this should read peruenerint ('[they] would come to'), which perhaps makes better sense: instead of 'we order and firmly command justiciars, sheriffs, provosts and their officers to whom the present document *would apply*', the text would be translated, 'we order and firmly command justiciars, sheriffs, provosts and their officers to whom the present document would come to'. The problem, however, is that the manuscript reading ('pertenerint', 'would apply') is possible. It is a matter of judgement—or guesswork—whether the document originally read one way or the other. 11

On the face of it neither of these examples appears to be very significant for our understanding of what the document says. They do, however, shed light on the perils of correcting a text not because it is obviously wrong, but because the choice of words is unusual. Maitland Thomson's 'correction' of 'inhibentes' to prohibentes is unwarranted because the word was definitely used in this way by Alexander III's own scribes. This does not, as it happens, change the translation. But it does mean that, if we were to follow Maitland Thomson's 'correction', we would lose evidence about the formal prose that was used in the royal writing office—in particular, about how rigid this may or may not have been.

Standard Latin Grammar v. Interchangeable Medieval Forms

The desire to correct a text when all you have to work with is an inaccurate copy is particularly strong when the form of a word cannot be found in a standard Latin grammar book. This can be illustrated by *poterint*. At first glance you might think that this is 'they will be able', because the equivalent singular form, 'he/she/it will be able', is poterit, which appears in the same passage in our document. But if you look up a standard grammar book you will find that 'they will be able' is poterunt, not 'poterint'. This would require that the copyist in about 1624 has made the common mistake of misreading a group of minims. (A 'minim' is the single stroke of an 'i' which is repeated to produce 'n' and repeated again to write 'm'.) Maitland Thomson, however, decided that 'poterint' was a mistake for another form of the verb, potuerint, meaning 'they will have been able'. This would require that the copyist in about 1624 has simply omitted a 'u'. Maitland Thomson's solution makes better sense if you read the sentence as a whole, although only if 'poterit' is read as potuerit, 'he will have been able'. He did not, though, change 'poterit'! We will find out why soon.

If we were to change 'poterint' to either poterunt or potuerint, would we be tipping over into correcting the original scribe's Latin, rather than the copyist's mistakes? As with inhibentes, it pays to look at other examples of Alexander III's brieves of protection. Here the crucial comparison is the brieve for Balmerino (H 1/8/147 in the database), another rare example of a 'long' brieve of protection from Alexander III. It has almost identical wording with our brieve for Dunfermline—including 'poterint' (rather than potuerint). This makes it unlikely that *poterint* is a mistake by the copyist. It should, instead, be recognised as a possible medieval form of the verb 'to be able'. 12 But is it a variant of poterunt, 'they will be able', or of Maitland Thomson's potuerint, 'they will have been able'?

It will be recalled that potuerint reads better than poterunt in this sentence, but that this is only true if 'poterit' in the same sentence is thought of as a variant form of potuerit. Why, then, was Maitland Thomson prepared to change 'poterint' to potuerint by adding a 'u', but did not make the same change to 'poterit'? This was not a mistake: 'potuerint uel poterit' is found in another of the rare extant long brieves of protection (H 1/8/43 in the database). 13 This survives as the sheet of parchment that was actually produced by the royal clerk, so there is no possibility of interference by a later copyist. There can be no doubt that in this case a royal scribe wrote 'poterit' for potuerit. (The fact that Maitland Thomson did not change 'poterit' to potuerit even though he changed 'poterint' to potuerint shows, incidentally, that he was not simply imposing schoolboy Classical Latin onto our document, but was using his knowledge of medieval Latin to inform his decisions.)

In the case of 'poterint' and 'poterit', therefore, we discover that Scottish royal scribes treated these as interchangeable with the Classical 'correct' forms with a 'u', potuerit and potuerint. This is important, because an unwary modern reader who depended solely on their standard Latin grammar book would find that 'poterit' could mean only 'he/she/it will be able', when in fact it could, in a medieval context, mean 'they will have been able'. In the case of our document they might then have felt justified in treating 'poterint' as a mistake for the plural equivalent of poterit, and thus changed it to poterunt. The result would be only a slight difference in the translation, but it would read less fluently, and might lead a future scholar to raise questions about the competence with which royal documents were drafted—questions which, in this case, would be unjustified.

Neither Copying Error or Unwarranted 'Correction'

What happens if there is not such firm evidence to corroborate that a reading corrected by Maitland Thomson was, in fact, what was written by the original scribe in 1271? There are two examples of this dilemma: it is best to consider them together. We begin with *Testatur Johannes* ('John [Comyn] bears witness'). Maitland Thomson thought this should read Testibus Johanne ('as witnesses John [Comyn]': 'witnesses' is plural because the two other men mentioned are included). He also thought that *presente* ('with [William of St Clair] present') should read patre ('father', referring to John Comyn). Instead of reading the sentence as 'John Comyn bears witness, with William St Clair present, and John of Fenton', therefore, Maitland Thomson 'corrected' the manuscript so that it reads: 'As witnesses: John Comyn the father, William of St Clair and John of Fenton'. It is very easy to see why he should have done this. The manuscript reading is extremely unusual, whereas the 'corrected' version is in a standard form. Also, changing 'presente' to patre makes sense in two ways. First, John Comyn, Lord of Badenoch and Justiciar of Galloway, had an adult son, also called John, so there would be some point in calling the elder John Comyn 'the father'. (There is another charter of Alexander III in which John Comyn is called 'the father', a charter for Robert Bruce, 28 March 1270. ) 14 The second reason is that 'presente' is ungrammatical. It is singular; but if it refers to both William of St Clair and John of Fenton, then it should be plural (presentibus). I have translated it as if it refers only to William of St Clair, but this is unsatisfactory because it leaves John of Fenton grammatically high and dry. The fact that 'presente' is ungrammatical therefore tips the balance of probability in favour of Maitland Thomson's corrections—not only of 'presente' to patre, but the rest of this sentence, too. The one correction leads inevitably to the others.

There is, nonetheless, a nagging doubt that we might be erasing precious evidence about the nature of witnessing. Charters tend to be formulaic in their language and constructions. This in turn can make it difficult to see what is going on beneath the surface. In this case, what did it mean to be a witness? Did it mean that you were present when the document was drawn up, or were you deemed to be taking on a particular commitment to verify the document's content? Could witnessing simply signify that you had been informed about the action recorded in the charter? Any deviation from the standard form could shed rare light on such questions. If the text originally said that John Comyn bore witness, and that William of St Clair (and, presumably, John of Fenton) were present, then this would be remarkable evidence that a distinction could be made between being a 'witness' and merely being present. This distinction can be found in one of the Gaelic property records in the Book of Deer, but would be unprecedented in a Latin brieve. There can, therefore, be little doubt, especially given the inaccuracy of the only surviving copy, that Maitland Thomson was justified to correct this the way he did, and make it read like a regular witness list. An English example that could be relevant here is a number of brieves ('writs' in English parlance) of Henry I early in his reign (1100–35) in favour of the abbot of Abingdon. A few have an unusual feature. Instead of naming the witness(es) in one of the standard ways ('Teste X' or 'Testibus Y et Z'; 'As witness X' or 'As witness Y and Z') there are occasions where we are told that X witnessed 'through A' (per A). For example: 'Teste Roberto filio Hamonis per W. de la Rochelle' ('As witness: Robert son of Hamo through W. de la Rochelle'), and 'Teste Rogero Bigod per Aret falconarium' ('As witness Roger Bigod through Aret the falconer'). 15 It appears that Robert son of Hamo and Roger Bigod, both major figures in Henry I's realm, were not there when the documents were drawn up, and 'witnessed' through someone of much less significance, who presumably was present and informed Robert and Roger what had happened. It is therefore conceivable that something similar could have been intended if the brieve of Alexander III originally stated that John Comyn bore witness but that someone else was present. Unfortunately, because our text survives only in an inaccurate copy, it is difficult to be confident that this is the original reading. It is much easier to dismiss it as an error, and to 'correct' the text to read like a standard witness list.

An Editor's Dilemma

Although there are clear grounds for making such a decision in this case, it gives a flavour of how anyone editing an inaccurate manuscript faces a fundamental dilemma. If a reading is very unusual, the natural inclination is to assume that the copyist has made a mistake. But can we be sure? By playing safe in this way by making the text conform to what is common, do we risk smoothing the edges in our sources, and limiting the evidence to only what we have the imagination to accept as plausible—and therefore lose a rare opportunity to challenge our assumptions and shed a flickering light on something we might not otherwise see? That, at least, would be the danger inherent in making corrections without comment, as Maitland Thomson did in his unpublished typescript. In this day and age we would take care, even in our unpublished work, to record what the manuscript said in the apparatus (preferably as footnotes), and to discuss the more interesting 'mistakes'.

For many documents the process of establishing what the text said is fairly straightforward. If the only available copy seems to be inaccurate, however, then the text is less certainly the words as written by the scribe of the original charter. What, then, is a scholar going to publish in an edition? The text that a scholar makes available for everyone to read will, of course, be the result of a number of difficult decisions they will have had to make. As a result, the published text will depend not only on his or her standard linguistic and palaeographical skills, or on their familiarity with documents of a similar type—and particularly with the rarer features that can be found. It will also reflect to some extent their personality—their willingness to take charge of the text and correct its 'flaws', or to allow something less tidy to remain in full view. Which would you regard as the 'safer' option? Your answer might reveal what kind of editor you are—or might become!


Feature of the Month