Monday, January 30, 2017

WoM: Welsh words in Anglo-Norman

Last month, we discussed the presence of loanwords from Irish in Anglo-Norman, and this month, we would like to look further into the linguistic contact between Anglo-Norman and Celtic languages – this time focusing on Welsh. While there has been considerable research into the influence of (Anglo-)French on the Welsh language, particularly in the literary sphere, linguistic contact between the two languages in the administrative and judicial spheres remains relatively poorly studied[1]

Among the materials and sources used for the compilation of the AND we find two editions that bring together documents written in Wales by Welshmen: the Calendar of Ancient Correspondence Concerning Wales, and the Calendar of Ancient Petitions Relating to Wales. The former contains mainly records of correspondence between the English royal court and nobles in Wales, while the latter provides evidence of petitions from individuals throughout Wales to the English king. There are limitations to using these two works - they are primarily calendars, intended to discuss the contents of the document, and often do not include the text. Although hardly indicative of there ever being widespread levels of comprehension of Anglo-Norman in Wales, this documentation does suggest the presence of scribes, notaries or other public servants, in all parts of Wales, with a high level of linguistic competence in Welsh, Anglo-Norman and Latin.

(Aberystywth Castle)

Trotter’s examination of the above mentioned texts, supplemented with his own transcriptions of a limited number of documents in Welsh archives, unearthed twenty Welsh words used in Anglo-Norman context. Of these, nine are currently included as entries in the AND: amobres, commot, frith 3, havoterie, hildenraeth, keveretz, kymorthas, merchet 1 and obeduz. As with Irish, all of the loanwords occur in administrative documentation, and for Welsh, a great number of the terms refer to legal concepts.[2] There is currently no evidence of Welsh words being used in Anglo-Norman literary texts, though it is known that a number of Anglo-Norman works were composed or copied in Wales.

Some of the borrowing of Welsh legal terms seems to be driven by the need to refer to Welsh law - one document refers explicitly to the law of Howel the Good, and in the entry for keveretz we find:

aprés la conqueste de Galis lor graunta lor leis et lor usages q'il aveient [...] la conqueste [...] keveretz Howel les queiles eux et lor auncestres ount eou et usé RLiR 58 482 (National Archives SC 8/146/7288)
['after the conquest of Wales, he granted them their laws and their usages as they had [before] the conquest [...] the cyfraith Hywel which they and their ancestors had and used']

The word keveretz is intriguing, as it renders the Welsh cyfraith, ‘law, legislation’[3], but clearly reformulates it, not only adjusting the spelling, but even producing an Anglo-Norman plural.
The documents also refer to amobres, a term for 'marriage fee' from the Welsh amobrand obeduz (from Welsh ebediw) referring to 'the rendering of a live beast to a lord at the death of a tenant'. A commot was a territorial and administrative division in Wales (Welsh cwmwd) and is attested in three different source texts in Anglo-Norman. The word was, from the end of the 12th century, present in Latin under the form commotus (DMLBS 397a) and from the 13th century, also present in English (see OED commot, n 37273) so it seems likely that the Anglo-Norman word is not a direct borrowing from Welsh.

Hywel Dda , Latin translation, British Library Harley 1792 f.3

Other Welsh words refer to the specific types of taxes owed by the local population: hildenraeth,[4] that is, a tax paid in oats, is likely derived from the Welsh hildaf  'to produce (a certain amount of crop)' and treth meaning 'tax' while kymorthas, from the Welsh cymorth, refers to a tribute of cattle owed to a lord, though the sole use of it in the Anglo-Norman citation below is ambiguous:

Item, qe nulls westours, et rymours, mynstrales, ou vacabundes, ne soient sustenuz en Gales, pur faire kymorthas ou quyllages sur le commune people, les queux par lour divinaciones, messonges et excitacions, sont concause de la insurrection et rebellion q’or est en Gales Rot Parl1 iii 508
[‘Item, that no wastrels and rhymers, minstrels or vagabonds, be sustained in Wales to receive tribute or money from the common people, who, by their divinations, lies and incitations are the root cause of the insurrection and rebellion that is currently in Wales’]

Legal terms are not the only type of borrowing from Welsh - a number of agricultural terms used by the local populations are also found in these texts. Frith is the equivalent of the Welsh ffridd, meaning 'mountain pasture', though the term ultimately derives from Old English (see OED frith,n.2). Havoterie 'summer meadow' derives from the Welsh hafod or 'summer residence' or 'upland farm'.

A nostre seignur le Roy et a son consail monstrent ses povre gentz bondes de sa havoterie en le counté de Meyronnyth qe [...] RLiR 58 482 (PRO SC 8/258/12874)
[To our lord the King and to his council we show the poor bondsmen of his summer meadow in the county of Merioneth that [...] ]

Welsh borrowings slated to be added to the AND in the course of its present revision to the Second Edition, include the terms raglot, raglour and ragelotie. These all refer to the Welsh rhaglaw, that is, ‘vicegerent, viceroy, deputy’. The term raglot is equally attested in Latin from 1304 (DMBLS 2650a raglotus ‘ragler, chief officer of commot’ ) while raglour was equally used in Middle English, though the earliest citation of the word in the MED and OED is in an Anglo-Norman text (MED raglore n.; OED raglour n. 157484). The DMLBS (2649c) attests to the use of the form raglarius ‘ragler, chief officer of commot’ from 1485. Raglotia ‘office of ragler, raglership’ is attested from 1314 while raglaria is attested from 1334. It seems likely that these terms have entered Anglo-Norman through Latin.

La ragelotie de Cruthyn  RLiR 58 483 (PRO SC 8/124/6154)
[The rhaglawry of Cruthyn’]

Ultimately, there are few identified borrowings from Welsh and Irish into Anglo-Norman, and those that are present are largely as a result of the borrowing of technical vocabulary in administrative documents. This may suggest a lower degree of contact between Welsh and Anglo-Norman as well as between Irish and Anglo-Norman than that between Middle English and Anglo-Norman. However, as Trotter emphasizes in his article, it is most certainly the case that Anglo-Norman documents from these areas remain less studied, and it may be that further investigations into the archival holdings in contacts zones in Ireland, Wales and Scotland may yield further evidence of language contact between the Celtic languages and Anglo-Norman.


[1] We are indebted to the work done by Prof. David Trotter on the topic, particularly the article ‘L’anglo-français au Pays de Galles: une enquête préliminaire’, Revue de Linguistique romane, 58 (1994), 461-88. Trotter provides a summary of studies on the (Anglo-)French influence on Welsh on p. 462, note 2 of his article.
[2] That’s not to suggest that the writing of Anglo-Norman by Welsh individuals was restricted to the production of administrative material. As Trotter notes (p.461), Hue de Roteland, who wrote Ipomedon [Ipom BFR] and Protheslaus [Proth ants] was from Rhuddlan (Dyfed) and a Simon of Carmarthen composed Le Chemin de Penitence [Penit].
[3] We use here the headwords and definitions of the Geriadur Prifysgol Cyrmu (
[4] It should be noted that this spelling has been rejected by the AND. The term should read hildevraeth.

Friday, December 30, 2016

WoM: Kerne and the Celtic languages

Over the last few months, our blog posts have focused on loan words in Anglo-Norman - from Greek, from Italian, from Mongolian ... This month and the next, we are going to have a look at some Anglo-Norman words borrowed from Celtic languages.

Medieval Britain was a multilingual environment, and it is clear that there was a high level of contact between Anglo-Norman, Middle English and Medieval Latin, resulting in a high level of loan words between the languages. But these were not the only languages used at this period in the Anglo-Norman regnum, which also included Ireland, Wales and Scotland. It is perhaps surprising then that relatively few words in the AND are tagged as deriving from either Welsh or Irish and no borrowings seem to have come from Scots Gaelic. This is a phenomenon that bears a closer look, suggestive of a very different contact situation in the Celtic countries than in England. Can the pattern and frequency of borrowings offer insights into the use of Anglo-Norman in Ireland, Wales and Scotland as well as offer clues about the level of contact between speakers of Anglo-Norman and the local languages? There appears to have been little work done on this perspective, to our knowledge, though we would be very happy to hear of work done on this topic![1]

There are a number of Anglo-Norman texts which were composed in Ireland - these are sometimes referred to as Hiberno-Norman works. These include La Geste des Engleis en Yrlande (Dermot2), composed around 1225, as well as numerous administrative texts extant from the period of Anglo-Norman rule, which can be found in our List of Texts under the following sigla: Affairs of Ireland, Chart St Mary's, Ireland, Irish Docs, Stats and Ords Ireland and Windsor. Despite the relatively large number of works produced in the area, there are currently only six words in the AND tagged as Irish: betagh, cro, jacoine, kerne, kernemen and grawe.

BL Arundel 14 f. 27v.; 'partially damaged map of islands including Ireland (labelled 'Hybernia') and 'Britannia' in Giraldus Cambrensis's Topographia Hiberniae' from the BL Catalogue of Illuminated Manuscripts

Betagh, may be from the Irish bétach, an adjective meaning 'doughty, valorous' but also 'violent, wanton', used substantively in the AND. It occurs once in a document from Waterford, written around 1300:

si un baron ou un chevaler ou autre gentishomme eit neifs, sicome maniere de betagh, e aventure aveigne q'il sont nees de sur la terre le avant dist baron ou chevaler ou autre frankhomme, e il ne puet pas estre demené a droiture, si com son pere estoit avant lui Bor Cust ii 89

['if a baron or knight or other gentleman has villeins, after the fashion of the betagh, and it happens that they are born on the land of the said baron or knight or other freeman, and he may not be held to right as his father was before him']

The editor of Bor Cust glosses the term as 'Irish villein' though it doesn't seem that the term had much currency in Irish.

Jascoine refers to the mythological fish encountered by Saint Brendan on his journey. The name of this fish, iasconius, may derive from the Irish íasc . The DMBLS suggests this possible etymology for the term as it appears in the Navigatio S. Brendani. The Anglo-Norman term no doubt derives from the Latin and not directly from the Irish term.

Grawe seems to refer to a type of cup or goblet, though the etymon remains unclear. It has been cautiously identified as Irish, due to the context, but may in fact be something else:

iii. coups e iii. grauntz hanaps qe hom appele ‘grawes’ d’argent  Affairs of Ireland 134

['3 cups and 3 big goblets of silver which one calls 'grawes']

Kerne and kernemen, from the Irish ceithern, refer to a type of lightly-armed Irish foot soldier. These terms seem to have been more widely used than the previous ones, and we find them used in a number of Irish administrative documents as well as in a letter written by Richard II from Waterford.

ou estoient de eux tuéz .clxij hommes armés et les kernes armez Lett & Pet 347.11

['there where .clxij. of their armed men and armed Irish foot-soldiers were killed']

Kerne is known in Middle English as well (MED kerne n. and OED kern n.1) so it is possible that the word entered Anglo-Norman though English rather than Irish. The earliest citation of the word is in a Latin text of 1297, which the use of it in Anglo-Norman is attested from 1316. The earliest use in an English document dates from 1423. The compound kernemen certainly suggests an English influence, though this compound does not seem to have been recorded in English.

Cro is a legal term, defined as 'fine for homicide' and appears in Irish as cro. It appears in a single, perplexing citation:

Item le cro et le galnys et le enach uniuscuiusque hominis sunt pares APS 664.

The citation, using Latin as the matrix language, but with the Anglo-Norman definite article to introduce the three terms from the other vernaculars, provides three synonyms for the same concept of 'wergild'. Firstly, cro from Irish; secondly, galnys, the Welsh galanas, an equivalent concept; and, thirdly, enach.

The last term could be from the Irish enech (found in the expression lóg n-enech 'honour-price') but also possibly a Scots Gaelic term, as it is found in a document from the Acts of the Parliament of Scotland and most likely a synonym for the other two terms. This is the only term identified as Scots Gaelic in the AND.

Ultimately, there are few identified borrowings from Irish into Anglo-Norman. This may suggest a must have been lower level of contact between the two languages than between Middle English and Anglo-Norman. However, it may also be the case that documents from these areas are must less studied, and that further investigations into the archival holdings in contacts zones in Ireland, may yield further evidence of language contact. We'll compare these results with Welsh in the next post.


[1] We'd like to thank the eDIL for their assistance, and for bringing the article H. Risk, 'French Loan-words in Irish', Études Celtiques 12 (1970/71) 585-655 & 14 (1974) 67-69 to our attention. Any mistakes in this blog are our own!

Wednesday, November 30, 2016

WoM: Greek lexis in Anglo-Norman

The alpha, but not quite yet the omega, of Greek lexis in Anglo-Norman.

As primarily a Romance language, Anglo-Norman more often than not traces the origins of its lexis back to Latin. As such, determiner comes from determinare, leun2 from leo and oreison from oratio – three entirely random but straightforward examples of how this type of development is so integral to the formation of a romance language that the AND will not highlight these words as Latin in origin. Evidently, Latin did not have exclusive rights to the formation of Anglo-Norman vocabulary – as our blogposts of the last few months have already testified, with examples from Mongolian (or not), Persian and Italian. Indeed, Anglo-Norman in its very nature is, to some extent, defined by an influx of Germanic, and specifically Anglo-Saxon/English, elements. For this month’s post we will take a look at the role of the second Classical language that contributed so much to the pan-European vocabulary: Greek. And what was its effect on the lexis of Anglo-Norman?

Medieval or Byzantine Greek, a term used for the language as it was used between approximately the sixth and mid-fifteenth century, was a mixture of the original Classical Greek and subsequent Koine Greek (the form of Greek which developed as the common Hellenistic and international dialect, and which was used, for example, in the Septuagint translation of the Bible and in the New Testament).

(Codex Skylitzes Matritensis, Bibliteca Nacional de Madrid, Vitr. 26-2, fol. 34v)

As before, using the ‘language tags’ that appear in the Anglo-Norman Dictionary entries, it now takes little or no effort to retrieve those words already considered to be of Greek origin. But with only 17 entries of the entire AND thus revealing themselves as ‘Greek’, the results seem at first look somewhat disappointing. Let us first take a look at those 17 entries and the contexts in which these loanwords appear.

Most of these words (7) are attested in medicinal texts: leucos, the Greek for white (λευκός), turns up in an explanation of leucoflemancie, a type of dropsy (λευκοφλεγματία):


si est dite leucoflemancia por ce qu'ele est faite de blanche fleume; leucos et blanc si est trestot une chose A-N Med i 230

(‘and so it is called ‘leuciflemancia’ because it consists of white phlegm; ‘leucos’ and white are one and the same thing’)


Similarly we have the words melangiron for ‘black jaundice’ (related to μελάγχιμος in the sense of ‘black, dark’) and inopos (from οἰνωπός, ‘red as wine’) to refer to a certain type of reddish urine. All of these appear mid-thirteenth-century translation of Johannes Platearius’s Practica Brevis – a Latin text but including Greek elements. The term kili vena occurs several times in that same text with reference to the main vein carrying blood from the lower body into the heart. The first element in this Latinized collocation derives from κοῖλος – the Greek word for ‘hollow’. In the Anglo-Norman Euperiston (an early-fourteenth-century text believed to have been translated from Latin as well) a discussion of anorexia explains orexis as meaning ‘appetite’:


Fastidium est apelé de Galien anorexia; si est dit de ‘a’, ke est a dire 'sanz', e ‘orexis’, ke est a dire 'appetit', sicum 'sanz appetit' A-N Med ii 168.133

(‘fastidium’ is called ‘anorexia’ by Galienus; this is formed by ‘a’, that is ‘without’, and ‘orexis’, that is ‘appetite’, and thus ‘without appetite’)


Finally, at least two separate collections of Anglo-Norman medicinal recipes use to term ana with the meaning of ‘in the same amount’ (Greek ἀνά, cf. DMLBS ana1):


Pernet les freides herbes si com jubarbe e teittesoriz e teles choses ana de chescune Five Med MSS 107.S185

(‘take cold herbs, such as houseleek and stonecrop, in equal quantity’)


It is apparent that in these cases, Greek terminology is retained from an original source text, and more often than not the word remains a ‘foreign’ borrowing in Anglo-Norman, with an explicit awareness of its Greek nature and origin.

(BL, Add MS 24371, fol. 15v, John Chrysostom’s 72nd Homily on the Gospel of Matthew)

The same can be said for Anglo-Norman words that derive from Greek mythology or history. The AND currently includes catoplepa (a Latinized from (as used in Pliny) of Greek κατω-βλέπων, literally ‘down-looker’) which refers to an African animal, usually identified as either a buffalo or an antelope; lethes (Λήθη, one of the five rivers of Hades whose waters induce forgetfulness) meaning ‘amnesia’; Melos (Μήλος, a Greek island north of Crete); and omega (Ωμέγα, the last letter of the Greek alphabet). The Anglo-Norman translation of Vegetius’ De re militari mentions the use of a monoxille, a type of boat made from a single piece of wood (μονόξυλος), by the Roman military.

The AND also picked up Greek terminology from botany and philosophy, but again the results are surprisingly meagre. Firstly, yperichon is the Greek name for St John’s wort (ὑπερικόν) and appears as an ingredient in thirteenth- and fourteenth-century medical or dietary treatises in Anglo-Norman.[1] Secondly, yle, from Greek ὕλη (‘wood, timber, material’ but since Aristotle used to refer to ‘matter’; modern English hyle), is discussed and defined in the early thirteenth-century text, La Petite Philosophie:

Ore escutez des element, Ço est de yle les liemenz. Tant dit yle cum fet matire, Dunt tute rens pernent afeire; Yle est matire divine, Dunt tutes riens pernent orine Pet Phil 312, 313 and 315.
(‘Now hear about the elements, that is the attachments of ‘hyle’. ‘Hyle’ means as much as ‘matter from which everything takes form’; ‘Hyle’ is divine matter, from which everything originates’)

In a religious context, the Anglo-Norman version of Jerome’s Letter to Paulinus retains the term ogdoad (‘a group of eight divine beings’) with references to Egyptian deities, which derives from Greek ὀγδοάς (‘eight’).

One final word currently tagged as Greek in the AND is diadocupo, apparently a type of oven, which appears in a fourteenth-century Anglo-Norman collection of alchemical material and which, aside from its Greek-looking form, has not yet been identified by the editor of the text or the AND editorial team.

Clearly, the influence of Greek cannot have been as immense as that of Latin: Latin was, after all, both the source and a medieval living language, in constant contact with the Anglo-Norman world. In contrast, Greek must have been either the language of (major) historical texts or the language of communication with the distant Byzantine Empire – with which direct contact must have been minimal.[2] Still, this does not explain the overall paucity of Greek lexis as appears from our initial examination.

One reason, it should be noted, must be the very nature of the AND language tag. This tag was not intended to indicate etymology, but rather to highlight the origin of words that still seem to be loanwords or borrowings when used by Anglo-Norman scribe or authors. In many cases those two aspects will overlap, but it means that when a word is fully naturalized in Anglo-Norman, it normally does not carry a language tag. This will mainly (and understandably) affect Latin etymologies: for example, pere1 is not tagged Latin despite its origin in pater, simply because it no longer registers as a Latin loanword in Anglo-Norman, in the way, for example loquendes (‘points for discussion’) probably still did (hence the ‘Latin’ tag). The same will often be the case for Greek, and that is why an entry like apocalipse (from ἀποκάλυψις) does not currently have a Greek tag.

 (The Prognosticon of Hippocrates, Harley 6295, fol. 98r)

However, such a distinction (of etymology vs. loanword status) is often debatable, as it may be very difficult to gauge the extent to which a word is naturalized or not in Anglo-Norman, as it is for any medieval language. As such, it may be argued that some entries currently in the AND should be tagged as Greek. For example, the abovementioned ogdoadis is tagged, while ebdoadis, a similar term appearing two words earlier in the same source text, has not. Also, how about the many other Greek mythical beasts, such as gorgone, monosceros and fenix, currently included in the AND without any language tag? It is clear that a more concentrated re-consideration of the AND language tagging system could reveal a greater influx of Greek lexis.
To give but one example, a passage in the Anglo-Norman version of Mandeville’s Travels provides the names of all Greek letters of the alphabet, but so far, only omega has been given its own entry:

Si vous voilez savoir de lour A B C quelles letttres [sic] ils ount, ici les poez veer ovesqes les nouns qe ils les appellent: alpha, betha, gama, delta, ebrevis, elonge, epilmon, zetha, hetha, iota, kapda, lapda, or, ni, exi, obrevis, pi, cophe, ro, summa, thau, vi, fy, chi, psi, othomega, diacosin Mandeville 112

Fortunately, the online AND is a constant work in progress and its nature allows these omissions to be rectified in our next batch-update.

The problem also lies in a more general linguistic conundrum: very often Greek lexis reaches Anglo-Norman filtered through the medium of Latin. Particularly in Hellenistic times, Latin was prone to borrow Greek vocabulary extensively and subsequently Latinize it: for example, the abovementioned yperichon for St John’s wort (from Greek ὑπερικόν) has its Latin counterpart in hypericum (DMLBS 1192a). And randomly opening the DMLBS, we see phtisicus from φθισικόϛ (DMLBS 2270b, ‘one who suffers from consumption’, cf. AND tisik, while the new edition of P, to be published in February 2017, will contain the new entry ptisic), plectrum from πληκτρον (DMLBS 2316a, ‘instrument with which one plucks the strings of a harp or lyre’, cf. AND pleitrun), and pasta from παστά (DMLBS 2138b, ‘dough, paste’, cf. AND paste).
When such words turn up in Anglo-Norman, should the source be considered Latin or Greek? To some extent, the historical context may indicate a direct link with Greece or Greek writings: Anglo-Norman ju d’Olimpiades refers to a Greek event, even though it uses the Latin ‘i/y’ spelling instead of the Greek ‘u’ (Ὀλυμπιάς). Similarly, the abovementioned monoxille – the boat made form one piece of wood – features in a Latin text on the Roman military.

(BL, Add. MS 39594, fol. 1r)

On the whole, spelling can offer some insight into whether such a lexeme may have entered Anglo-Norman in its Greek or Latinized form (for example, the use of Greek ‘k’ versus Latin ‘c’, or nominal endings ‘-os’ and ‘-on’ versus ‘-us’ and ‘-um’), but on the whole, in a language where nominal endings are usually dropped and where ‘c’ spelling is generally preferred to ‘k’ spelling, such a distinction can rarely be made.

In conclusion, we must assume that, just as in the DMLBS, a great deal of Greek vocabulary is ‘hidden’ in AND entries that are currently tagged solely as ‘Latin’ or not tagged at all. Therefore, in this case an overview of the entries tagged as Greek in the AND constitutes only the tip of the iceberg of Greek influence on the Anglo-Norman lexis.


[1]  In Trevet’s Chronicle the word clearly retains its Greek nature:
la racine del herbe q’est en Gru appellé 'yperycon', qe nous apelloms 'herbe percee' ou 'herbe Johan' TRIV 48.76
(‘the root of the herb which in Greek is called ‘ypericon’, which we call ‘herbe percee’ or ‘herbe Johan’)
[2]  For example, the visit of Manual II Paleologus, Byzantine Emperor, to England (and his stay at the palace of Henry IV) in 1400 was a highly exceptional moment, as D.M. Nicol notes in, ‘A Byzantine Emperor in England: Manuel II's visit to London in 1400-1401’, University of Birmingham Historical Journal 12 (1970), pp. 204-25.

Friday, October 21, 2016

Word of the month: Some Italianisms in the Port Books of Southampton

A map of medieval Southampton based on the Terrier of 1454:

My previous AND blog [July 2016] on luxury fabric colours introduced the study of Anglo-Norman / Italian language contact, an area which has been largely overlooked by academics so far.[1] Tuscans, Genoese and Venetians played a crucial role in the economy of late medieval England and Italian merchants and bankers could be found in many social niches. They dominated the textile and wool markets, they were the main importers of sugar and spices, they acted as personal money-lenders to the King and ran the Royal Mints, they worked closely with the London Guilds (such as the Worshipful Company of Grocers) and they were the undisputed European masters of shipping. One of the aims of my recently submitted PhD thesis (Money Talks: Anglo-Norman, English and Italian language contact in medieval merchant documents, c1200-c1450) is to uncover probable Italian borrowings in the AND2 which have, so far, not been identified. Examples of loanwords are as wide-ranging as the Italians’ influence and include silks and brocades (AND damask / baldekin / taffata), dyes (AND cramoisé), financial terms (AND tare1), boat names (AND carrak / tarette), wine types (AND vernage), a high-quality Indian ginger (AND belendin) and a verb for ‘sifting the refuse from spices prior to sale’ (AND garbeler).

This month’s blog, however, looks at some of the handful of entries in the AND2 which are already labelled as ‘Italian’. All three are found in two sets of Anglo-Norman records tracking imports to and exports from the busy port of Southampton, a major destination for Genoese carracks and Venetian galleys in the 1400s. Rather than sailing in and out of the harbour with no interaction with locals, we have evidence that Italians became very much part of the life of the town. For instance, there are surviving records of 134 Italians  - listed as Genoese (78) / Venetians (20) / Florentines (15) / Italians (12) / Lombards (9) - who were residents of Southampton between 1431 and 1472.[2] The diary of Luca di Maso degli Albizi, captain of a state galley from Florence, details his stay in the Hampshire town in the winter of 1429-30 where he lodged with a wealthy ship owner, William Soper, and was wined and dined by local dignitaries.[3] Italians even became civic officials themselves: the Venetian, Gabriel Corbizzi, was Port Steward in the 1440s, overhauling the office accounting along Italian lines.[4] More impressively, the Florentine Cristoforo Ambruogi (Christopher Ambrose) was twice elected mayor of Southampton in 1486 and 1497, a rare privilege for an ‘alien’ in England at the time.[5] Finally, Stewards’ Books from 1487-88 and 1492-93, written in Middle English, reveal that Southampton townsmen and Venetian galley crew worked together on several occasions to fell timber in the New Forest and build derricks (or scaffolding) on the dockside to unload cargo.[6] Italian loanwords also feature in these sources, e.g. maregon (‘ship’s carpenter’) < Venetian marangóne. In all these situations, it is fascinating to imagine how the English and Italians involved communicated and how loanwords moved from one language to another.

The route followed by the merchant galleys from Venice to Southampton:

1. AN fangot (‘bundle of cloth’) < Ital. fangotto
 liij bales xxi fangot , contenu vij c x draps ij verges  (Port Bks 50) (att. 1428) (AND fagot, no. 2)
The Port Books of 1427-30 were edited back in 1913 by Paul Studer and he was the first scholar to recognise an Italianism in an insular French text: “As the word is of Italian origin, cf. Ital. fangotto, it may well have been introduced into England by the Genoese settled in Southampton”. He also notes enthusiastically that “the word must have been quite familiar to Southamptonians of A.D. 1428, seeing that it was commonly applied to bundles of exported cloth (!)”[7]

The loanword fangot is used eleven times in Port Book entries between 1428 and 1430 and reappears as fangottis in a Latin Port Book from Southampton from 1440:  

pro xxv pakikis, vij balettis et xiij fangottis panni / pro iij fangottis panni continentibus j pannum et xviij vergas (Port Bk Southampt. 72 / 73) (DMLBS fagotus).

In 1474, we also find fanget of cloth in a Middle English account, from the Guildhall in London:
For sealyng of a litill Fanget of cloth (Let. Bk. Lond. L, Gldh, 118) (MED fangot)

Fangotto is a lesser known variant of Italian fagotto (itself a Gallicism from OF fagot ‘bundle of sticks / firewood’): TLIO fagotto (att. 1348).[8] The word developed a new meaning of a ‘bundle of cloth’ but this specialised use is not found in France, only in Italy and England. It was still being used in this way in northern Italy in the nineteenth century, according to Giuseppe Olivieri’s Dictionary of Genoese sub fangotto. The borrowed fangot is unusual as, unlike many Italian loanwords in medieval England, it remained in use by English drapers for over three hundred years. We find it in Edward Hatton’s The Merchant’s magazine or Trade-man’s treasury (first published in 1695 with eight subsequent editions) in a chapter helpfully called ‘A Merchant Or Trader's Dictionary, Explaining the Most Difficult Terms Used in Trade’ :

Fangot: an uncertain quantity, as of Raw Silk, 1 to 2 ½ C. Grogram and Mohair Yarn 11/2 C. to 2½ C.

We can more examples in early modern English the OED2 entry sub fangot, e.g.

one Fangot of White Cyprus Silk (London Gaz. No. 841/4) (1673)
Fangotts of Italian raw silk (London Gaz. No. 4472/4.4) (1708)

2. AN sport (basket) < Ital. sporta

vij sport de resins (Port Bks 43) (att. 1428) / i sport de suchre pot, val. xxs. (Local Port Bk 66) (att. 1436)  (AND sport)

Another example of a commercial loanword from Italian in the Port Books of 1427-30 is sport or ‘basket’. We also find the borrowing a few years later in the Local Port Book of 1435-36, another set of accounts written by the same man: the Southampton Water Bailiff, Robert Florys. He uses sport over twenty times in total and always with the same five commodities: almonds, raisins, figs, soap and suchre pot (‘pot-sugar’). Interestingly, its use is not confined to the ‘Alien Book’ (the section in the Local Book dedicated to trade with Italians) but also the ‘Common Book’ which deals with everything else. This suggests that, for Robert Florys at least, sport was an everyday part of his business vocabulary.

We also find rarer examples of the diminutive, sportin, which appears to mean ‘small basket’ or ‘half a sport’ in both Florys’ Port Books:

iiij sport ij sportin de almand (Port Bks 49) (att. 1428) / viii sportin de resin (Local Port Bk 108) (att. 1436) (AND sportin)

Records of medieval Italian sporta are easy to find. If we look at the OVI corpus, we find over fifty examples from 1318 onwards including this from the Florentine Pegolotti’s merchant handbook where he states that pepper is sold in such baskets:

Olio in giarre. A sporta si vende: Pepe. A peso si vende: Indaco. (Pegolotti Practica 70.16) (c1335-1343) (OVI sporta)

Sporta even remains in modern colloquial Italian but mainly in fixed expressions such as un sacco e una sporta (‘a large amount’) and dirne un sacco e una sporta a qualcuno (‘to give someone a telling off’). An etymon for sportin is, surprisingly, less easy to find. The obvious candidate is sportino / a but there is no medieval record of the term. Could the name of a round Florentine fruited bread known as the sportina di Pasqua, although not recorded until the 1800s, hint at the earlier existence of the word?

Sportina di Pasqua:

3. AN cotegnate (‘quince marmalade or paste’) < Ital. cotognato

j barel de cotegnate   (Local Port Bk 90) (att. 1436) (AND cotegnate)

Unlike fangot and sporta, this Italian borrowing  is a ‘hapax’ and appears only once in the Anglo-Norman record, with one entry referring to a 100lb barrel of quince marmalade, worth 13 shillings and 4 pence, arriving into Southampton on the 4th of January on a Venetian ship.[9]

La cotognata:

La cotognata is still eaten in Italian today: a traditional confection served in small squares and equivalent to Spanish membrillo. Its first appearance in an Italian text actually records its import into London in 1306 by the Gallerani of Siena, alongside other sweet treats like candied nuts and ginger confit (OVI cotognato). We also know that English cooks prepared their own version of quince preserve with ‘native’ names such as connates and quynade. These terms feature in Middle English recipes from a1399 and a1450 and appear to be compound words made up of variants of ‘quince’ (AND coign1 / OED2 coyn), followed by an –ate / -ade suffix:

Connates. Take Connes and pare hem. (Form Cury (Add 5016) p. 18) (att. a1399) (MED connates)

Quynade: Take Quynces & pare hem clene [etc.]. (Two 15th-cent. Cookery-bks. (1888) 27) (att. a1450) (OED3 quinade)

The later name codiniac found in English in 1539 (OED2 codiniac), is clearly borrowed from the Continental French variants coudoignac / coudougnac, first recorded in France in the 1380s and themselves influenced by Occitan codonat / quodonat, according to the FEW (II-2, 1606a: cydonem).

However, given the circumstances of this loanword’s use in a Southampton document directly linked to imports from Venice in 1436, Italian cotognato seems the obvious choice for our lone example of cotegnate. Overall, what is interesting about the Southampton material is that in all three of our loanword examples, the writer could have used a range of pre-existing Anglo-Norman or Middle English terms: e.g. bundelle / fardel for ‘bundle’, couple / frael / paner for ‘basket’, connates / quinade for ‘quince paste’. But instead, Robert Florys chose (consciously or not, we will never know) to borrow a term from Italian instead. [MST]

[1] David Trotter discusses this issue in two articles from 2011: ‘Death, taxes and property: some code-switching evidence from Dover, Southampton and York’, in Code-Switching in Early English, ed. by Herbert Schendl and Laura Wright (Berlin: De Gruyter), pp. 155-89  /  ‘Italian merchants in London and Paris: evidence of language contact in the Gallerani accounts, 1305-08, in Le changement linguistique en français: études en homage au professeur R. Anthony Lodge, ed. by Dominique  Lagorgette and Tim Pooley, (Chambéry: Presses de l’Université de Savoie), pp. 209-26. I have also presented various new sources of loanword evidence: (2012) ‘Mercantile multilingualism: two examples of Anglo-Norman and Italian contact in the fourteenth century’, in Present and future research in Anglo- Norman: Aberystwyth Colloquium, July 2011, ed. by David Trotter (The Anglo-Norman Online Hub), pp. 91-99 / (forthcoming) ‘Early Anglo-Italian contact: new loanword evidence from two mercantile sources, 1440-1451’ in Merchants of Innovation: The Languages of Traders, ed. by Esther-Miriam Wagner, Bettina Beinhoff and Ben Outhwaite (De Gruyter Mouton).
[2] This data is available on the England’s Immigrants 1330-1550 website, a project based at the University of York and directed by Prof. Mark Ormrod.
[3] This Tuscan source is transcribed in full on pp.207-80 of Michael Mallett’s The Florentine Galleys in the Fifteenth Century (London: Oxford) from 1967.
[4] See The Fifteenth-Century Stewards’ Books of Southampton, a PhD thesis by Anne Thick from the University of Southampton (1995) and also T. B. James’ (2015) ‘The Town of Southampton and its Foreign Trade 1430-1540’ in English Inland Trade, ed. by Michael Hicks (Oxford: Oxbow Books), pp.11-24 (p.13)
[5] See Alwyn Ruddock (1951): Italian merchants and shipping in Southampton, 1270-1600 (Southampton: University College), pp.185-86.
[6] See Alwyn Ruddock (1944): ‘The Method of Handling the Cargoes of Mediaeval Merchant Galleys’, in Bulletin of the Institute of Historical Research, XIX: 140-48.
[7] Paul Studer (1913): The Port Books of Southampton or (Anglo-French) accounts of Robert Florys, Water-Bailiff and Receiver of Petty-Customs, A.D. 1427-1430 (Southampton: Southampton Record Society),  p.50.
[8] The Tesoro della Lingua Italiana delle Origini (TLIO) is available online at but hyperlinks cannot be made to individual entries.
[9] See Brian Foster (1963): The local port book of Southampton for 1435-36 (Southampton: The University Press), p. 91.