Search This Blog


Monday, November 17, 2014

Word of the Month: Gagging, queasy and squeamish

Writing a dictionary is a never-ending process. As editors we are constantly moving forward through the dictionary, revising the original entries and adding new ones, currently for the letters P and Q. However, new research, new editions and new perspectives can cause us to return to earlier, completed entries and rethink them. As last month’s entry showed, the addition of semantic tags prompted some rewriting of entries so that overlapping semantic fields can be made more evident and more easily searched through the tagging system. A similar rewriting process has been inspired by incorporation of links to other dictionaries, which can highlight words that don’t belong to the etymon once believed.

The original entry for gagé (not present in the first print edition, but added in the online second edition, in 2005) defined estre gagé de as ‘to be inclined to refuse to’. This was illustrated by a single citation:

E pur enducer le beyre, metés un poy de sucre ou licoris pur se que plusures sunt gagé de beyre amer choses Receptaria 126.431

(‘To sweeten the drink, add a bit of sugar or licorice, as many are *gagged* by drinking bitter things’)

No definition for the term is offered by the editor of the text.[1] 

While the word appears to be a derivation of gage1 (i.e. ‘pledge’, see FEW *waddi 17,443a), the sense of the word does not resemble any known French verb or adjective of this spelling or derived from this etymon. This posed a problem when we attempted to add links from gagé to the other dictionaries. Where did this word come from?

It did seem to look a lot like the English verb ‘to gag’. Was it possible that this was a Middle English borrowing? The OED sub gag v.1 notes that word is first attested in 1440, with the sense of ‘to strangle, suffocate’ but does not appear to have been used again in that sense. The sense of ‘to stop up the mouth with a gag’ appears in the sixteenth century while the (in this case) the most relevant sense of ‘to choke, to retch’ doesn’t appear until the early eighteenth century.

The MED sub gaggen v. notes two earlier uses of the word in English, in 1269 and 1327, albeit as surnames, with no clear meaning:  are Thom Gagge and Robertus Gaggemon ‘stranglers’? However, these instances suggest that the word must have been in use earlier than the OED can attest and thus could have been borrowed from English into Anglo-Norman. It does appear that the term was also borrowed into Medieval Latin – the DMLBS attests to its use in the sense of ‘to force (someone’s) head back’ (sub gaggare 1044b) a single time in the early fourteenth century.

The etymology of the word remains enigmatic. The OED suggests it is onomatopoeic, while the MED points to the Old Irish form gag-háls, ‘with the neck drawn back’. 

Interestingly, Anglo-Norman appears to have another word that may come from the same etymon. At the beginning of the fourteenth century, Nicolas Bozon used gageous – a possibly related adjective – twice in his Moralized Tales (Contes moralisés):

si il poy mange e beyt poy, lors est gageous ou escoymous (var. gayons (l. gayous) e escoymuse)   BOZ Cont 158

(‘if he eats and drinks little, then he is queasy or squeamish’)

Et coment qe seit orde beste, uncore est gageouse en tant qe quant le gopil [...] avera fet sa vilenye en le entree, jamés illeoqes ne vendrent  BOZ Cont 179[2] 

(‘and although it (=the badger) is a dirty beast, still it is squeamish – so much so that when the fox, [...] defecates in the entryway (=of its burrow), they will never return there’)

London, BL, Harley 4751 fol. 30r; Shockingly few images of badgers in medieval manuscripts.

In both cases, the sense of the word seems to be that of ‘queasy, squeamish, physically unable to support swallowing or turned sick.’[3] Interestingly, although queasy is attested in English from about 1450 (MED quaisi(e adj 35516; OED queasy adj.), no related form is found in Anglo-Norman. The etymology of queasy is also uncertain, and may possibly be an Old Norse or Old French derivation.

Semantically, though not formally, gageous is linked to yet another English word of Anglo-Norman derivation, squeamish (OED queamish adj.; squeamish, adj. (adv. and n.) and squeamous adj.; MED queimous adj. and squaimous adj.).

Initially this word seems to be an insular creation, attested only in fourteenth-century Anglo-Norman. As the entry escoimus shows, it is found in Bozon (who is the first to use it, around 1300) and Henry of Lancaster’s Seyntz Medicines as well as two other fourteenth-century texts (Respit, a collection of proverbs and Courtoisie, a manual about fine manners). The entry in T-L (escomos 3,954) and the DMF (escoimous) are illustrated only with these same citations, that is, there are no Continental examples of the use of this word.

But is the word really an Anglo-Norman creation or once more a borrowing from Middle English? The OED and MED suggest that the word came into English from Anglo-Norman. Indeed, the spelling of the word certainly implies an Anglo-Norman provenance.

Its presence in Bozon’s text, and its use next to the ‘anglonormannified’ version of another Middle English word could be seen as suggestive. Bozon included a number of Middle English phrases and proverbs in his work, especially the Contes, and it appears that the author was competent in both languages. Was he relying on his Middle English vocabulary to introduce these Anglo-Norman words?

Within a year or two of composing his text, we find in the Middle English Handlyng Synne (1303):

Anouþer vyce ys ʒyt to graunte, Þat rychë men mochyl haunte, Þat many one are so daungerous, And oute of mesure esquaymous,  Þat  hys kokë may no day Greyþë hym hys mete to pay (ll. 7245-7250)

 (Another vice is yet to be acknowledged, That rich men practice muchly, That many a one are so dangerous, And out of measure squeamish, That their cook may no day, Prepare his meat to please him.)

This poor soul in Rome, Biblioteca Casanatense 4182, fol 192, seems out of measure squeamish.

Unfortunately, the (later) Anglo-Norman version of this text doesn't translate these lines verbatim.

The appearance of the word at nearly identical times makes it difficult to determine precisely how the word entered both languages. The etymology of escoimous remains unclear. There doesn’t appear to be any related words in (Medieval) Latin.[4] The FEW suggests an etymon in the Old Gascon sceamu (‘shame’ 17,22b) and the REW suggests a related etymon in the Old Norse skömm (‘shame’ 8005b) though the semantic development from this word is problematic.

The etymon skömm, gives us in English the word shame n. as well as other derivations. Two medieval derivations look very similar to our escoimous shamevous adj. ‘shameful, disgraceful’ and shamously adv. ‘shamefully’ (for which the OED suggests an unattested shamous adj.). However neither of these terms are attested before the middle of the fifteenth century.

Another possible source for the word is the one that gave use qualm n.3 which is very similar both formally and semantically (‘a sudden feeling or fit of faintness or sickness’). The etymology for this word is equally uncertain, but appears to be of Germanic origin, with cognates found in Danish and Swedish. However, the word is unattested prior to the early sixteenth century.

In the end, the word remains obscure. Skömm remains the most likely etymon of the word but it is difficult to know if escoimous  was an early adjectival derivation in Middle English (e.g. scamous) which was borrowed into Anglo-Norman by Bozon or conversely, if it was a Norse word used in Normandy (though otherwise unattested in Norman) which Bozon then picked up. Or perhaps it was created independently on both sides of the channel.

In the end there is no neat solution to the puzzle (as yet! Suggestions are always welcome!). It does underline some of the complexities of interlinking dictionaries – finding the cognate references is not always straightforward! Luckily, the editors can now add commentaries in entries where they can explain some of the challenges they faced in creating the entry.


[1] Elizabeth Valentine, in An Edition of the Anglo-Norman Content of Five Medical Manuscripts of the Fourteenth and Fifteenth centuries, Ph.D. thesis, University of Exeter, 1990, also edits this text and offers the translation ‘nauseated’.
[2] The editor offers the definition ‘dégouté, délicat’.
[3] The DMF attests to one use of the word in Middle French (gageux) but believes that the terms is related to *waddi and defines it as ‘engagé, obligé’. It would equally be possible to define it as ‘sickened’: Plus que vous deux ne suy hardis ne corageux: Se premiers je vouoye, trop serroie oultrageux. Pour tant ne le feray, de ce suy trop gageulx : A men honneur seroye laidement damageulx. (Ysaÿe Triste G., p.1400, 282).
[4] Medieval Latin has the very close squameus (DMLBS 3173c); however, this is defined as ‘covered in scales, scaly’. See English squamous adj.

Thursday, October 23, 2014

Word of the Month: Anglo-Norman chess terminology (and how to find it)

As an online resource, the AND is constantly  evolving. Not only are we in the process of producing a much-expanded new edition of the dictionary itself ( headwords starting with P/Q are our current focus), we also continue to expand the already revised A-O section, adding citations, senses and even entries whenever we come across new material. A word you looked up and couldn’t find yesterday might be in there today!

In addition to this work, we have been designing new features and search facilities for the site. For example, over the last year and a half, we have been providing each and every AND entry with cross-references to other relevant dictionaries of French, English and Latin (so far this is available for entries beginning with G-O – that is about 7500 entries).[1] At the same time, we are also developing a new and searchable semantic tagging system. What this is and how it would serve the user, this blog-post will briefly explain and illustrate.

The so-called ‘semantic tag’ has always been a (minor) feature of the dictionary and appears in the English definitions: plant-names, for example, have been preceded by ‘(bot.)’ and bird-names by ‘(orn.)’ – merely to indicate that, for example, dove was an ornithological term whereas dove’s foot belonged to the realm of botany (sub colum).

 Example of an AND entry with current semantic tags (in red).

The current project will considerably expand the scope of these ‘semantic tags’, improve their reliability, and (ultimately) make them searchable.

By providing such a search facility, it will become possible for any user to bring together  sub-groups of Anglo-Norman vocabulary used in a particular field of medieval society, such as  shipping, heraldry, law, music, games, or medicine (to name but a few), which we intend to expand with clothing, minerals, kinship, horse-riding, alchemy, weaponry, emotions, and several more. Scholars of medieval  clothing terms or legal terminology will have their relevant source material made available with one simple search.

To give just one provisional example of what sort of data we could possibly gather through such a search, we did an ad hoc search of the underlying data for ‘chess’ – a ‘semantic tag’ which has already been used fairly consistently in the AND. We found that the tag is present in 73 entries on the AND (A-Z).

Oxford, Bodleian MS 264, fol.60r

The game of chess has its origins in India and was originally introduced to the Western world by the Arabs as Shat-ranj (from Sanskrit chaturanga: ‘four members [of an army]’). The Anglo-Norman word for ‘chess’, eschec (cf. modern French échec(s) ) derives from the late Latin word for chess scacca[2]. In English the word produced both ‘checkers’ and ‘chess’ – two different games employing a similar sort of board.[3]

Two adversers (‘opponents’), one blanc (‘white’) and the other neir (‘black’), play a ju (‘game’) or bataille (‘battle’). They have their gent (‘people’) or mesnee (‘household’) of homes (‘men’, i.e. pieces) asseé (‘set up’).

On the board, the squares are called places or points (except the corner-square, which is called an angle), and the rows or lines are railles.

Although they did not always move in ways we are familiar with nowadays, the names of pieces are more or less what we expect them to be: for king we find rei, queen is reine, knight is chevaler , roc rook, and peon is pawn (also the word for walker or someone on foot).

Bonus Socius's chess treatise from the late fourteenth century, MS Ludwig XV 15, fol. 97

A pawn is also referred to (only in Anglo-Norman) as pet (which must be related to in the sense of foot, rather than the other entry pet – which means ‘fart’) and curliu (‘herald, messenger’). The latter use is considered problematic in the dictionary.[4] The pawns were considered masle (‘male’) or femele (‘female’).[5] Presumably this attribution of gender must have distinguished pawns that move on the same colour as the king from those that move on the colour of the queen, but further research on these senses is still needed.  

In medieval chess, the queen was also called ferce (from the Persian word ferzen or Arabic firzan/ferz , meaning ‘wise man’ or ‘counsellor’) and functioned as a piece that was only able to move one step diagonally.[6] The Anglo-Norman examples show that the term was also used to refer to a pawn that reached the eighth square and was consequently promoted to a queen.  Finally, the bishop is still known as alfin, from al-fil the Arabic word for ‘elephant’ – a piece that was able to move two squares diagonally and jump over other pieces.[7] It is also called cornu (‘horned’ or ‘the horned one’), possibly indicating the elephant’s tusks.[8]

The verbs used for playing the game are semover (‘to move’) or (se) muer (normally ‘to change’, but in some instances also ‘to move’). Pieces that move up the board are said to be munter (‘to climb, ascend’), and if they are trying to take another piece chacer (‘to chase’), mener (‘to lead’) or enangler (‘to corner’). Embracer (‘to embrace’) is used when one piece covers or protects another one. Taking a piece is haper (‘to grab’) or gainer (‘to gain’) – and hence the expression qui done, gaine was used to describe a strategy which involved sacrificing one of one’s own pieces in order to take one of the opponent’s. When the king was descovert (‘uncovered’) possibly by lack of an effective defense, defension or garde, it was said to be in meschief (‘trouble, difficulty’). Subsequently, placing him in check is faire or dire eschec or, when checkmate, mat, eschec mat or eschec plener (‘full check’). In the later case, also mater (‘to checkmate’) was used or even juger (‘to pronounce judgement’). The result was a matement, matesun or (in analogy with the aforementioned verb) jugement. For a game ending in a stalemate (which, in medieval chess was not a draw, but resulted in a win for the player delivering the stalemate), the verb used was estaler.

Große Heidelberger Liederhandschrift, Cod. Pal. germ. 848, f. 13r

Two thirteenth-century Anglo-Norman treatises on chess have been preserved,[9] whose main concern is the explanation of jupartis: prescribed chess moves and strategies. Juparti became the English word jeopardy (‘peril, danger’)[10] but was originally a chess term (also in English until the fifteenth century) and stands for ju parti: a ‘divided game’ or a game which is as yet undecided.

This is nothing but a cursory overview of all of the material that is currently labelled ‘chess’ in the AND. Although it already gives an idea of the range of terminology, it is also incomplete (I already mentioned the specific use of femele for pawns, and there are no words at present for, for example, ‘to lose’, ‘position’, ‘opening’, ‘to jump over’, or even the board itself). Therefore, before we can open this search-facility to the public, further work needs to be done (and is currently being done) to improve the robustness of our ‘semantic tag’ system.

The Lewis Chessmen, twelfth-century set of chess pieces, 
from Scandinavia, but found on the Isle of Lewis


[1] Each entry links to two leading etymological dictionaries of French (the Französisches Etymologisches Wörterbuch (FEW) and the Dictionnaire Étymologique de l’Ancien Français (DEAF); two medieval French dictionaries (Godefroy and Tobler-Lommatzsch), the online Middle French dictionary (the Dictionnaire du Moyen Français or DMF), a Modern French dictionary (the Trésor de la Langue Française (TLF); the Oxford English Dictionary (OED); the online Middle English Dictionary (MED); and the Dictionary of Medieval Latin from British Sources (DMLBS). Further relevant dictionaries may be added in the future.
[2] See scacca DMLBS 2948a, where the etymology is explained as originating from shah, the Arabic word for ‘king’.
[3] See MED ches n. and OED  (only from the early eighteenth-century) checker n.2.
[4] The definition is based upon the context, and no other instances of it have been found elsewhere: ‘Encontre reison eust mespris Que ad curlus reine eust assis’ (He’d have acted against reason if he had placed the queen by the pawn) Eschez ANTS 24.110.
[5] This sense is currently missing from the AND entry for femele, but the citation under masle provides a usable attestation of this. This just as an example not only of how the current semantic tagging is not yet consistent and therefore somewhat unreliable, but also of how a closer look at semantic subgroups may help editors of the AND to identify gaps or omissions.
[6] See also OED fers n. and MED fers n. It is only in the fifteenth century that the queen gets her modern moves.
[7] See also MED aufin n. The piece was referred to as bishop from the twelfth-century in Courier Chess, but according to the OED, this sense appears only from the sixteenth century onwards in England.
[8]Al neofime (=move) vint avant li cornuz, Si li mostre ses corns aguz’ (At the ninth move, the horned one moves forward, and thus shows him his sharp horns) Eschez ANTS 11.111.
[9] Both are made available in ‘plain text’ editions in Tony Hunt, Les Gius Partiz des Eschez: Two Anglo-Norman Chess Treatises, ANTS Plain Texts Series 3, London, 1985.
[10] See OED jeopardy n. and MED juparti(e n. The same word also appears in Medieval Latin, clearly derived from the vernacular, as  jupartia (DMLBS 1514b), and Continental French (cf. DMF jeu (jeuparti) and FEW jocus 5,42b).

Monday, September 15, 2014

Word of the month: Nice! An Anglo-Norman insult.

English speakers may be surprised to learn that the etymology of nice is not very nice at all and that its semantic development is unparalleled in the Romance languages. This word, which style guides recommend that you avoid as it both ubiquitous and nearly devoid of all meaning, has a most complicated semantic evolution.

The word nice is attested quite early in French – ca 1160 and has its roots in the Latin nescius, an adjective meaning ‘ignorant, unknowing’.[1] The word was used in French (and other Romance languages) in Middle English (c. 1400) to disparage people, actions and sayings as silly or foolish. This is the meaning the word retained in the Romance languages, though in French the word is rather uncommon today though you may find it in some older texts to refer to someone as simple or naive, such as those the TLF cites: Un brave homme, un peu nice, appelé Monthyon (Pommier,Colères,1844, p.66)

The semantic development of the word nice in English is a rather complicated affair as its entry in the OED explains as it breaks down the fourteen (!) senses the word has had over the centuries. From the original meaning of ‘foolish, silly, simple’, the word went on to be used with a wide variety of senses: slothful, effeminate, shy, wanton, meticulous. The current sense of nice as ‘agreeable, pleasant’ doesn’t show up until the mid-eighteenth century and has now overtaken all the earlier meanings.

In Anglo-Norman, the word nice only seems to have been in its Latinate meaning of ‘foolish, ignorant’, so it doesn’t seem that the development in English is due to Anglo-Norman and it doesn’t appear that the Middle English use of the word particularly affected the use in Anglo-Norman. A more Latinate form is also attested in Anglo-Norman in the entry nescient, with the same meaning as nice, a form that is equally attested in English (nescient), though at a much later date (c. 1500).

Carrow Psalter, Fool with bladder on stick eating cusped loaf, Walters Manuscript W.34, fol. 113r detail

A similar separation of the languages appears in their respective uses of another word to refer to a simple or ignorant person, naif.[2] Naif is derived from the same Latin word nativus that would give us native in English, and nearly all the citations for naif demonstrate this sense of naturalness or nativeness.

English develops several forms of the same word – naif, naïf, or naive – but none are attested prior to the 16th century. They all enter English with the sense of ‘native inhabitant or bondsman’, meanings much closer to their Anglo-Norman counterpart. In English, naive does not develop the current meaning of ‘unsophisticated, credulous’ until some time in the seventeenth century.

One citation from Gower, a medieval writer of Latin, English and Anglo-Norman works, suggests an interpretation for naif closer to the moden usage of ‘foolish, naive’:

Plus nyve que le prisonner Qui tout jour voit l'uiss desfermé Dont il pourroit en saulf aler, Mais ne se voet desprisonner, Tanq’il au gibet soit mené  GOWER Mirour 5695

trans: More foolish than the prisoner who all day sees an unlocked door, Out of which he could go safely, But does not wish to free himself Until he is led to the gallows.

Surprisingly, natif does not seem to be well attested in Anglo-Norman, despite the fact that the form is found in Old French from the twelfth century.[3] In Anglo-Norman, natif is currently found only in a few late glosses of Nequam’s De Nominus Ustensilium, as a gloss to Latin nativam, alongside the more usual naif, as well as in some late parliamentary rolls with the sense of ‘native inhabitant’ It is the term naif which is used most frequently in Anglo-Norman to refer to the feudal state of bondsman but it is attested a few times from the 13th century, often modified by fol, to mean a foolish person.

King Solomon instructing his son, Bible historiale, Clairefontaine and Paris (1411)

So if calling someone ‘nice’ means you are actually calling that person an idiot, how can you refer to someone clever? In medieval French the term generally used for this was cointe, which could be used to call someone clever, or quick-witted or skillful.[4] It was a fine line however, as the same term was used to call someone crafty or devious – you could be clever, but not too clever! The term is derived from the Latin adjective cognitus, meaning ‘wise’ or ‘clever’ and continues to have a positive sense in Modern French (‘joli, agréable), though it is considered an archaic term.

Cointe was borrowed into Middle English, but you might recognize it under the more familiar form of quaint. When it first entered English, it was used in similar senses to the use in Anglo-Norman, that is, to characterize things that were cunning or clever or skilful. However, it certainly no longer has this sense, but generally connotes something that is pleasingly old-fashioned, a meaning the word would acquire some time in the mid-eighteenth century.

So be careful when addressing a medieval re-enactor! Your 'Nice job!' may not be the compliment you intend!


[1] FEW: nescius 5,494a; Gdf: nice 5,494a; TL: nice 6,661; DMF: nice; TLF: nice; DMBLS: nescius 1909c
[2] FEW: nativus 7,44a; Gdf: naif 1 5,464a; GdfC: naif 10,190a; TL: naïf 6,479; DMF: naïf; TLF: naïf; DMLBS: nativus 1889b 
[3] FEW: nativus 7,45a; Gdf: natif 5,474a; GdfC: natif 10,192a; TL: natif 6,521; DMF:natif; TLF: natif; DMLBS: nativus 1889c
[4] FEW: cognitus 2/i, 843b; Gdf: cointe 1, 2,173c; TL: cointe, 2,254; DMF: cointe TLF: cointe

Friday, August 22, 2014

Word of the month: 'Outremer'

Outre-mer (see TLF) is a French term that can be used to refer to faraway countries, be it in Africa, the Orient or America. It is a direct translation of the Latin ultra mare, literally ‘across the sea’, which in its adjectival form ultramarinus (cf. DMLBS 3545a), also produced the English word ultramarine: the blue pigment derived from the mineral lapis lazuli which, in medieval times, was imported from Asia by sea. Within a medieval context, Outremer also became a word used to refer to the Crusaders’ Holy Land and more specifically to the French settlements in the conquered territories of the Near or Middle East: the lands ‘across the sea’. It is mainly with that latter sense that outremer was used in English (OED Outremer n.), albeit only from the first half of the nineteenth century, when it makes its first attested appearance, rather surprisingly, in the Longfellow’s travel book Outre-Mer: A Pilgrimage Beyond the Sea:

‘I, too, in a certain sense, have been a pilgrim of Outre-Mer; for to my youthful imagination the old world was a kind of Holy Land (p.7)’

Louis IX sailing off on his second crusade,
from the Chroniques de France ou de St Denis, 1332-50 (Royal 16 G. vi, f. 437v)

Dr. Laura Morreale of Fordham University currently runs a project under the title ‘The French of Outremer’, which uses the term (with the necessary caveats) with precisely that sense, to bring together studies of the French language as it must have circulated in Crusaders’ settlements.

The word or phrase outre mer – sometimes with a definite article outre le/la mer – was quite common in medieval French, and also in Anglo-Norman. For example, in the early thirteenth-century romance of Gui de Warewic, the word appears in Anglo-Norman with reference specifically to the ‘land of the Saracens’:

‘Les Sarazins de ultre mer En Romanie venu esteient’ (l. 4650)
(The Saracens from ‘outre mer’ had arrived in Romania)

Another example can be found in Brevia Placitata, a fourteenth-century collection of legal texts:

‘s’en ala outremer en pelerinage e lessa le maner saunz garde’ (p. 184) 
(he went on pelgrimage ‘outre mer’ and left the manor without a ward)

Evidently, the sense of ‘Middle East’ or ‘Holy Land’ in these two examples is only circumstantial  and the phrase outre mer itself barely has more significance than ‘across the sea’.

In a thirteenth-century medical text, one of the ingredients of a medical preparation is urtie de outremer (i.e. ‘nettle from outremer’); see AND2 sub urtie. No further indication is given of the precise nature of this plant, but similar recipes suggests that this may refer to the Greek Nettle (see OED Greek a.) – the nettle from across the Mediterranean Sea, though not quite as far as the Holy Land.

Jehan de Mandevilles' Le Livre des merveilles (Paris, BN fr. 2810, fol.188v)

It turns out, however, that most Anglo-Norman attestations use the word without specific reference to any country or area and because of the geography of England, that is, surrounded by the sea, the term could simply mean ‘abroad’. For example, in the early-fifteenth-century Liber Albus, a compilation of earlier Guildhall records, we find:

‘Des avoirs qe veignent d'outre meer: ciere, argoil, quivere, estein [...]’ (p. 231) 
(Goods that come from ‘outre mer’: wax, argol, copper, tin [...])

which are not necessarily the most exotic commodities.

Another citation, from a case account taken from the Exchequer Chamber, shows a usage of the term which is most likely deliberately unspecific:

‘si un apport bienz de ouster le mer en Engleterre par cause de merchandiser et les jett sur le terre nient customés [...]’ (Exchequer Chamber ii 34.14)
(if somebody brings goods into England from ‘outre mer’ with a view to selling them, and brings them on land without paying customs [...])

In some instances, the context makes it clear that outre mer does not go any further than across the Channel, for example in the fourteenth-century Anonimalle Chronicle:

‘En cel temps le roi ové simple compaignie des gentz passa outra mier au roi de Fraunce’ (p. 142)
(At that time, the king together with a simple train of people crossed ‘outre mer’ to the king of France)

Similarly, towards the end of the fourteenth century, Richard II wrote to Maud, countess of Oxford:

‘[...] considerantz les [...] disaises que nostre bien amé W[auter] H., nadgairs [...] cook a [...] nostre cousin le Duc d'Irlande vostre filz [...], avoit pur le temps q'il estoit demorant en le service de nostre dit cousin es parties outre la meer’ (Lett & Pet p. 64)
([...] taking into consideration [...] the inconveniences which our beloved Walter H., former cook of our cousin the Duke of Ireland your son, had during the time when he was staying in the service of our said cousin in those regions ‘outremer’)

In this instance, outremer is Ireland.

World map (BL Add. MS 28681, f.9)

It seems that from quite early on outremer also became a legal term in Anglo-Norman, taken over in seventeenth-century law English as oulter-le-mer n.,  which functioned as a type of essoin, i.e. an excuse for non-appearance in court:

‘Purceo qe mulz de genz se font fausement assoigner de utre meer, la ou il furent en Engletere le jour de la somonse [...]’ (Stats i 37 )
(As is the case that many people have themselves incorrectly essoined of ‘outremer’, while they were in England on the day of the summons [...])

‘[...] ke essoigne de utremer ne soit aluee en nul manere de plai jeté pur celi ke soit truvé a sumunse’ (Winchester 46.11)
([...] that the essoin of ‘outremer’ would not be allowed in any way in the case of a plea put forward for someone who is found present for the summons)

Although the same ambiguity may be at play in these examples, the essoin is ultimately one of not being in the country at the time of a court case.

Altogether (and perhaps not surprisingly), in Anglo-Norman (and the same can be demonstrated for Continental French, cf. DMF outre-mer) the mer in the expression outre (la) mer seems to have referred to any major expanse of water, from those that are the immediate borders of the country to the Mediterranean Sea and possibly beyond. Ultimately, outremer was anywhere but England.