Guillaume and William, Guest and West, guile and wile: The vagaries of language

I have been thinking about the vagaries of language today, realising that the certainties of today melt as we trace words back through the very fluid languages of the past.  It started with the wonderful graphic below appearing in my Twitter feed:

It graphically displayed what I had only intellectually grasped before: that languages are transmitted and mutate, but retain commonalities through transmission.  Phoenician gave rise to Greek, and Greek gave rise to both Latin and Norse Runes via Etruscan. 


Why does this matter?  Because there was no right way to spell things when few things were written down, and mostly then by foreign clerics.  In the Carmen the West in Westminster is written as Guest, because to the Franks Gu represented a W sound.  William was Guillaume.  Wile was guile.  It goes on.  There was no W in Latin.  Anglo-Saxon uses 'uu' and then W.  Lundenuuic became Londonwick.


Seeing the commonality helps understand the differences.  And it really all goes back to the Phoenicians who settled and exploited Britannia as the Cassiterides - tin islands - for centuries before Julius Caesar arrived on these shores.

The 1066 Surrender of London was at Westminster, not Berkhamsted

I have been thinking much of Berkhamsted, having presented last month to the Military History wing of the Dacorum U3A.  It was a full house, which was very encouraging as I was debuting a new lecture on the 1066 Siege and Surrender of London, and also launching the full colour and larger format new book: Carmen Widonis – The First History of the Norman Conquest.  The topographic maps look fantastic in colour!

Folks in that part of the world like to think that the surrender of London and the northern earls took place at Berkhamsted in Hertfordshire in 1066.  This sad confusion came about because of careless association of an entry in the Anglo-Saxon Chronicle, Manuscript D, which says King William came to Beorhhamstede where London surrendered.  Beorh means noble or grand, ham means home, and stede means settlement, so together they mean ‘noble home settlement’.  In 1066 the only royal palace settlement for King Edward the Confessor was near Westminster, where he had built a grand Gallic palace next to the refounded  and papally privileged Abbey of St Peter. Both buildings were constructed on the model across the Channel at Fecamp in Normandy, where the abbey was next to the ducal palace during Edward's youthful residence there.  Edward also took some inspiration from other Gallic abbeys and courts he visited during his 28 years of living in Gaul in exile. 

King Edward modelled his royal lifestyle on Gallic comforts and his royal administration on Gallic methods.  The palace had apartments for King Edward's retainers, mostly Gauls and Franks he had recruited to his service from abroad.  The Abbey had a library, archive, treasury and school for royal administration.  The clerics too were mostly Norman, Gallic or Frankish.  He left his young cousin the perfect set up for taking command of England's royal administration.

English historians don't like being reminded that King Edward was far more Gallic than Anglo-Saxon, and that he had reason to love Duke William better than the Anglo-Danish earls that surrounded him in England.  They conveniently forget that Edward, as senior royal at the Norman court, was a foster-father to Duke William from the death of his cousin Duke Robert the Magnificent in 1035 until his departure for England in 1041.  William of Poitiers styles King Edward as dominus or lord or master to the young duke, perhaps indicating that Edward had help teach the young duke such arts of royal administration as he himself had acquired.

The Carmen says that King William settled at ‘a royal palace of surpassing beauty’ at line 666 and that the rebels and Londoners went there ‘to the palace’ to surrender at line 743.  Manuscript E of the Anglo-Saxon Chronicle confirms that the place of surrender is Westmynstre, but that didn’t settle the matter for some Victorian historians.

Edward Augustus Freeman is justly famed for his mammoth effort in compiling The History of the Norman Conquest of England from 1867 to 1879, but Freeman was a man of his time and very carefree about making assumptions that suited his personal biases.  He was a racist, misogynist, imperialist, Protestant bigot, and that comes through in his work.   My own view is that everything he wrote should be reinterrogated by modern historians. Freeman despised Scots, Irish, Welsh, and Cornish equally.  All Danes were little better than pagan thugs.  The French were, of course, beneath contempt and damned to perdition by their Catholicism.  No wonder Freeman was so popular with English historians and schoolteachers for generations and remains one of their favourite authorities!  

Freeman’s careless speculation with geography is particularly irritating to me.  He got so much wrong by being ignorant of ancient place names and equating whatever name he came across with its 19^thcentury English homonym.  He ignored the obvious fact that successive invasions of Romans, Saxons, Frisians, Angles, Danes and Normans all had given different names for the same places.   For Freeman Peunessellum becomes Pevensey, despite Pevensey being founded in the 14^thcentury by a charter that says the old place was abandoned due to siltation of the port.  Hastinge portus becomes Hastings, despite there never being a town there before the conquest and never a port there at all, and the Elizabethan Richard Camden saying the old Hastings was ‘swallowed up by the sea’.  For Freeman the surrender of London in December of 1066 at beorhhamstede must be Berkhamsted, not the ‘royal palace settlement’ at Westminster where King Edward had just built a magnificent new palace to Gallic standards of comfort so convenient to besieged London.  

Obviously, Berkhamsted likes thinking the surrender of London and the northern earls happened in their neck of the woods as a local claim to fame.  In 2016 a bronze bust of William the Conqueror was unveiled to mark the 950th anniversary of the conquest and London's surrender.  Freeman would be pleased.

Recorded as Berchehamstede in Domeday Book, ‘birchtree home settlement’, the town was sited on a hillside above a navigable watercourse northeast of London.  Styled as a Burbio, indicating early establishment of an Anglo-Saxon market, it was likely a convenient place to trade for supplying London, which could be reached from there by navigation.  This is suggested by 52 burgesses and £4 thelony recorded in Domesday. 

However pleasant as a market town, Berkhamsted was far too remote for a sensible siege camp, being 26 miles distant. 26 miles was two days march or a long horse ride in 1066.  Siege warfare requires being close by.  We know King William was not shy of battle, and he had vast experience leading sieges against walled towns from his many campaigns in Gaul. He wouldn't hide out in Hertfordshire while his army surrounded London's walls.  He would have been leading the siege during each day, and retiring to the comforts of his foster-father's royal palace every night.

Moreover, Berkamsted Castle was built after the Conquest, not before.  The only reason it is dated as 11^thcentury now is Freeman’s specious attribution of the surrender there.  No archaeological evidence exists of pre-conquest fortification.   

Reading the Law of Wihtred reinforced my confidence that Beorhhamstede was an Anglo-Saxon term for a royal estate.  A council of Witans gathered at Berghamstyde in 695 according to the preamble to the Laws.  Since King Wihtred was king only of Kent, it makes sense that Berghamstyde is a royal estate in Kent.  The king of Kent wouldn't travel to remote Hertfordshire to enact his laws in 695, and neither would a conquering King of England when besieging London in 1066.

So where was the royal palace in 1066?  Probably pretty much where it is today.  What is now Buckingham Palace was recorded in Domesday Book as the Manor of Ebury, belonging to King Edward and Queen Edith in 1066 and Geoffrey de Magnaville, port reeve of London, in 1086.  It seems King William gave the port reeve the palace as a further boon after the port reeve delivered London to him in surrender in 1066.  By 1086 the king had a magnificent new castle to move to next to the city, now the Tower of London. Geoffrey left the land at Westminster to the Abbey of Saint Peter, and after the dissolution of the abbeys the land fell into private ownership.  What is now Buckingham Palace was reacquired as a royal residence by King George III in 1761.

Google is 20: Talk about making history!

20 years ago yesterday two guys in garage decided to improve the way we find information on the internet and Google was born.  I remember exactly where I was when I first heard the word Google.  It was used as a verb in a geeky seminar I attended in 2000 on XML (extensible mark-up language) data interchange, as in, "So I googled it." 


I made a note in the margin of a page to look up 'google' on Alta Vista.  I never used Alta Vista again.  I never again wrote Google without capitalisation.


Google has globally democratised information (Search), navigation (Maps), communications (gmail and Blogger), translation (Translate), and intellectual and social interaction (YouTube) in ways we could not then imagine.  To a large extent, Google drove global adoption of the internet as an essential requirement for 21st century work, life, love, and education.


I recognise that I am blessed by Google as an historian.  Difficult and rarely used Latin words and phrases come to life in every text in their original context where they were published when I Google them.  Google confirmed to me that Clangendoque at line 100 of the Carmen was both a unique usage and a Googlewhack (unique occurrence with only one result anywhere on the internet) when I first searched for it.  Google confirmed my suspicion that sinu placido meant a calm basin of the sea to Pliny the Elder, rather than a pleasant strand of beach, as I had initially translated it. 


Over the past five years of Carmen translation and research I've made thousands of Google searches, viewed hundreds of topographic maps, written hundreds of emails, published tens of this blog, tested phrases again and again in Translate (though it's no where near as good as William Whitaker's WORDS), and enjoyed the shared music, lessons and ideas on YouTube.


I don't suppose Larry Page and Sergey Brin follow this blog.  But then we never really know if our prayers are heard either.  Sometimes it is enough to be grateful.  Being grateful is an act of generosity not just to the object of your gratitude but to yourself, reminding yourself that others are engaged with you on the journey of discovery and help you each and every day.  So yeah, Google, thanks.

'Haran Apuldran' means 'Anchorage at Appledore'. 'Hoary Apple Tree' is more Victorian nonsense.

Victorians were racist, sexist, religionist, imperialist fantasists, with limited grasp of ancient geography, language, and navigation.  In their rush to re-write English history as leading to the divine right of Germanic peoples, and especially the English, to subjugate and exploit the rest of humanity, Victorians made up a lot of stuff that doesn't hold up to scrutiny.  Most of what they made up is wrong, no matter how many times it got repeated in modern history books and text books.  Think modern American fundamentalists with Adam and Eve riding dinosaurs and Jesus sporting an assault rifle in their school books, and you get some idea of how Victorians distorted history.

Today's example is "hoary apple tree", where King Harold supposedly mustered the English army on the way to Hastings to meet Duke William of Normandy in 1066.


The Anglo-Saxon Chronicle, manuscript D, for 1066 says:  he gaderade þa mycelne here, com him togenes æt þære haran apuldran.
Haran apuldran was translated romantically in 1861 as ‘hoary apple tree’.  Victorians even speculated on the variety of apple from those then growing around Battle Abbey.  Haran Apuldran was better translated long before in 1731 as ‘estuary at Appledore’, which is much more likely for the line of march from London and makes better sense for coastal navigation. 

In medieval navigation harena or arena meant a sandy anchorage or estuarine basin suitable for grounding the round-bottomed coastal raiding and trading vessels, carinas, named as the vessels of the English at line 319 of the Carmen Widonis.  Carinas were very much like Dutch Knorrs, and the names derive from the same ancient Frisian word. 


Per mare • per terram praelia magna parat • In mare quingentas fertur misisse carinas • [Carmen Widonis 318-19]
It is said he has sent five hundred ships to sea so as to hinder our course of return.


Both harena and arena are used hundreds of times to mean safe, sandy, tidal anchorage in the 12th century navigation guide De Viis Maris - the Ways of the Sea.  There should be no further confusion about translating haran as anchorage.

Early medieval Appledore was a small peninsula below the Roman road at the edge of the great tidal basin of the Rye Camber.  A Roman road led across tidal causeways at Tenterden, Northiam and Brede to the cape of Hastingas.  Appledore was an ideal place for muster of local land armies marching from Wessex and Mercia and sea-borne warriors being carried by ships around the coasts from the estuaries of the north, Essex and London.  Travel by coastal transport was faster, safer, and allowed much more efficient carriage of men and materiel.  King Harold, his brothers and his father always travelled by ship whenever possible, as Anglo-Danish sea-lords should do. 

At the time, all of the vast Rye Camber was tidal estuary or salt-marsh.  Appledore was on a secure tidal island or promontory extending below the Roman road leading to London.  It had fresh water, and plenty of space to the north for forage for horses.  It was also likely the place referenced, but not named, in The Anglo-Saxon Chronicle for 893, when Haesten sent a fleet of 250 ships up the Limne – ‘border river’ – from the coastal sea-ford where they destroyed ‘a fort within the fen, whereon sat a few churls’.  A coastal fort requires fresh water and forage for horses, so probably Appledore.

If you're a historian still repeating 'hoary apple tree', stop it.  It's silly and wrong.  Start saying 'anchorage at Appledore'.  That is accurate and sensible and a good translation of haran apuldran.

The map below shows the English camp in 1066 at Appledore, and illustrates a line of march to the cape of Hastingas using Roman roads, fords and causeways.  All the blue would have been tidal seascape in 1066. 

Nathan Bailey's Dictionary and "Harold, the last Danish king"

I have bought another dictionary, perhaps the most wonderful dictionary I will ever own.  It is Nathan Bailey's An Universal Etymological English Dictionary and Interpreter of Hard Words.  Despite the fame of Dr Samuel Johnson and his 1755 adaptation of Bailey's production, Bailey was the first to compile an etymological dictionary of the English language in the form still used today.  Earlier attempts at dictionaries just gave word meanings.  Bailey's dictionary was also the most popular of the 18th century. Bailey innovated many of the forms and practices that Dr Johnson (whose 1755 dictionary was actually the 21st dictionary of the English language) and others emulated.

I already value two things about the Bailey Dictionary that make it infinitely superior to Johnson's effort.  The first is that Bailey recognises the etymology of a word when it entered the English language.  For example, he refers to Teutonic or Ancient German, where Johnson misleadingly says German.  There was no Germany until the first Deutsche Bund in 1814, and certainly none in the 5th to 7th centuries when many words came into use in Britain with Saxon emigration.  It is misleading and inaccurate to say words came into English from German.  The loose German confederation that emerged in the 19th century amalgamated many earlier tribes and diverse linguistic traditions.  Bailey recognised that diversity of antiquity and respected it.  Bailey also recognises a lot more scope for Old French, Greek, Hebrew, Arabic, Belgic, Dutch and Syrian, and even has a separate styling for Sea Term to cover nautical dialects common among mariners from India to Eire.


Bailey also includes many words which were considered too common, uncouth or vulgar for inclusion in later dictionaries aimed at schooling the gentry.  Bailey gives words associated with specialist trades, manufacturing, sea-faring, body parts, and even swearing.


I didn't find the first word to come to mind when I checked for completeness, but I did find Shite: to ease Nature, to discharge the Belly.  No entry for shite in Dr Johnson.


When schooling of children became general during the Victorian era, the priggish governors opted to put Dr Johnson in school libraries and classrooms, thereby narrowing the English mind and denying the richness of English linguistic origins.  If schools had stocked Mr Bailey's dictionary rather than Dr Johnson's, then it is likely a lot more English boys and girls would have spent time consulting the dictionary to their betterment.  Kids like looking up rude and interesting words.  Later Victorian toponymists narrowed English place names even further, as Celtic, German and Latin, ignoring the richness of Nathan Bailey's enquiry.  That narrowing of the English mind is regrettable, and has led to a fair amount of misguided history too.


I can give an example of how the narrowing of the English mind rendered history less accurate, and it is the reason I bought the dictionary.  Nathan Bailey is the only source for the phrase "Harold, the last Danish king" of England, in his entry for Battle Abbey.  The Victorians, perhaps to please a very German royal family, altered Harold to Saxon, ignoring his Danish mother, Anglo-Danish father, and other obvious Danish preferences, like slave raiding along the coast in his ships. 
Because Harold was born in Sussex, Harold became Saxon.  I suppose the children of British parents building the British Empire who were born in Hong Kong became Chinese, or those born in Bombay became Indian, or those born in Johannesburg became African, right?  Historian Edward A. Freeman was delusional in suggesting Earl Godwin had any loyalty to King Aethelred or his sons, or that the Godwin family identified as anything but Danish.  Godwin fought for Danes, was made an earl among Danes, married the Danish daughter of Viking Thorkell the Tall, and gave all his eleven children Danish names, including Harold.
Godwin's daughter Gytha became Queen Edith when she married King Edward.  After his death when she commissioned the Vita Eadwardi, however, she omitted all discussion of her parents' heritage and ethnicity, even though the purpose of the first half of the book (written before Harold died) was to ennoble her family and justify Harold's seizure of the English crown.  If Harold had any noble Saxon connections, it seems likely she would have wanted this advertised.  So if Gytha the family historian was ashamed of her parents' heritage, then it is all the more likely that there was no noble blood to claim.  Her maternal grandfather was Jarl Thorkell the Tall, famed hero of the Battle of Assendun, her maternal uncle Ulf had married Cnut's sister and been regent of Denmark, her cousin was King Sweyn of Denmark, but she chose not to advertise these connections.  She was similarly silent on her father's heritage.  Arguably if Godwin was the son of Wulfnoth then his treason against King Aethelred, raiding and slaving the coast in 1008 with 20 ships, and burning of 80 kings' ships might not be seen as wholly patriotic.  The evidence for Wulfnoth being the son of Ealdorman Aethelmar of Wessex is very thin, but if he was, he was born of a handfast Danish woman, not a Saxon wife, as Wulfnoth is not a noble Saxon name.


Freeman's history inspired Sir Edward Bulwer Lytton to a more transparent flight of fiction.  Harold the Last of the Saxon Kings may have been read by generations of English schoolchildren, but Sir Edward was as inaccurate and misleading as Mr Freeman in making an 11th century Anglo-Danish earl into a Saxon hero.  Harold and his family slaved Saxons to Danish markets to their great profit.  They would never have identified themselves with the subjugated tribe they exploited.  Harold's standard at the Battle of Hastings was the Dacian wolfskin of a Danish sea-warrior.

Dacian wolfskin standard from Trajan's column in Rome, Dacian wolfskin standard with King Harold Fairhair, and detail of Dacian wolfskin standard with King Harold Godwinsson, the last Danish king of England, from the Bayeux Tapestry.

Another Dacian wolfskin standard from Trajan's column in Rome paired with the death scene of King Harold Godwinsson in the Bayeux Tapestry.


I've been saying Harold was a Danish king for four years now, and it's nice to find that an English historian agreed with me.  It's a shame minds and dictionaries narrowed so much in the centuries since.


One notable fan of Bailey's Dictionary was Abraham Lincoln.  He kept his copy of Bailey to hand as he was schooling himself in Indiana from at least 1823.  Bailey's dictionary enabled Lincoln to comprehend works that would otherwise have remained beyond his grasp.  Lincoln's skills as a lawyer and orator are probably due to his fondness for reading Bailey's dictionary for pleasure.


When I was in secondary school I was teased for 'reading the dictionary' because I had a wider vocabulary than many of my peers.  Now I really am reading the dictionary, and am taking great pleasure in doing so.

Finding the 1066 Landfall: Pefnesea or Pevenesae at Lydd and Dungeness

On Saturday we made a sort of family pilgrimage to Dungeness Lighthouse.  It was erected in 1960 using several novel principles of engineering and construction innovated by the late Frank Hay.  Frank Hay's son and two adult grandchildren were among our party.  The grandchildren had not been to Dungeness before and they were captivated by the lighthouse scale, scenery, and seeming permanence.


The principal innovation was the use of pre-cast concrete rings, a very new technology in 1960.  These were fabricated in nearby Rye.  The rings were transported to Dungeness, stacked on a foundation on the shoreline, then interlinked by four internal vertical steel cables, tensioned by anchor points either end to hold the lightweight structure intact through all conceivable forces of wind, sea and shifting shingle that might test a lighthouse on the exposed coastal spur.  Today the lighthouse still stands proud, although the coastal shingle has since shifted gradually to leave the lighthouse some hundred metres or so further from the shore. 

Nothing manmade is permanent in that shifting, liminal landscape and seascape.  The image above was taken from the highwater line on the shore, where it is in 2017.  We know from eyewitnesses that when the lighthouse was erected in 1960 it was sited on the coastal waterline.  The vast shifting arcs of shingle that have accreted to separate the lighthouse and fishermen's cottages from the sea in the very few decades since led me to think again about the mystery of where the 1066 anchorage of Pefnesea/Pevenesae might be found.

Ða com Wyllelm eorl of Normandige into Pefnesea on Sancte Michæles mæsseæfen, 7 sona þæs hi fere wæron, worhton castel æt Hæstingaport - Anglo-Saxon Chronicle Manuscript D for 1066

Scene 38: 'Herein Duke William with a great fleet is crossing the sea and is coming to Pevenesae.'


Pevenesae was an anchorage reached on a single tide with a south wind, as we know from the detailed imagery of the Bayeux Tapestry and the description of the navigation in lines 100 to 116 of the Carmen de Triumpho Normannico - The Song of the Norman Conquest. 

But fearful lest dark night impose losses on your men, and wind roil the sea in contrary gusts, you order the ships to stations so the curved anchor might restrain their sterns.

In the middle of the sea you make a sea-harbour. 

You advise the sails be laid down pending the morning ahead in order that your exceedingly weary people might have rest.

The Norman Fleet clearly came to Pevenesae as a well-known anchorage in the seascape, and not a place in the landscape.  In the Tapestry, the ships are under sail, steersman are guiding by the leeboard, lookouts are scanning for defenders on the land, and the oarports on the ships are all closed.  No one steps out of a ship.  These are visual signals to the savvy mariners of the medieval age that the ships are still making progress toward the landfall which is detailed in the next scene.  At both the launch from Baie de Somme at Scene 36 the landfall at Hastinga in Scene 39, the sails have been taken in, the masts are struck, the steersmen guide ships to the strand with poles, and the oarports are meticulously detailed as opened when the bare boats line up on the strand.  In Scene 39 tousers are rolled only to the knee, and horses can walk on the firm strand.  There are no anchors in either scene, although massive anchors are prominent in Scene 6 when Harold lands at Ponthieu.

I was lucky enough to travel to many Nordic capitals while I was translating the Carmen for the first time.  While in Iceland I visited Viking World, where they have a replica, sea-worthy Viking ship.  Being able to actually walk the deck and inspect the craftsmanship was amazing.  The most striking thing for me was the swivelling oarports.  Touching the oarports and seeing how they close for sea travel and open for estuarine raiding by oarsmen changed my understanding of the Bayeux Tapestry.  Once you know about swivelling oarports, you cannot confuse Scene 6 (a coastal landing with anchors and closed oarports) with Scene 36 (an estuarine launch with open oarports) and Scene 39 (an estuarine landing with open oarports).

The detail from the Carmen about the 'sea-harbour' is one I am particularly proud of deciphering.  No previous editor has understood the line, and I might have missed its meaning too were it not for my Nordic travels and the scholarship of Chuck Stanton on Norman Naval Operations in the Mediterranean.  He illustrated the sea-harbour formation used by Venetian fleets to secure their warships and horse transports as below.  As soon as I saw the illustration I understood line 116 of the Carmen:  In medio pelagi litus adessus facis.  It had seemed nonsense, but once you have seen the sea-harbour formation it makes perfect sense.  Sterns are linked together so that archers on the bows of warships can defend against attack.  Horse and materiel transports are sheltered within the arc of sterns, safe from attack.

Ignorant modern landlubbers have equated Pevenesae with Pevensey, claiming the Normans landed at the mythical Pevensey Bay (named by Victorian real estate developers), but that is nonsense.  Pevensey did not exist as a named place in 1066.  The modern town was founded by charter of King John in 1207 after the burgesses' old port at portus Peuenesel closed to the sea from shingle drift.  There is no archaeological trace of a port at modern Pevensey, despite extensive exploration.  There is no geologic survey evidence for a navigable basin or channel at Pevensey (the west side of the bay would certainly be unsuitable, shallow, silty salt-marsh).  There was no fresh water, no fuel, no forage for horses, no food to be looted anywhere nearby, and no rational reason for settlement on the barren, sea-sprayed spur leading to the Roman-era fort.  The fort itself has yielded no archaeological evidence of sustained habitation during the Anglo-Saxon era.  Any channel through the Pevensey Levels in 1066 must have been toward the east, leading to the sheltered tidal anchorage between Hooe and Wartling. 


The nearest reference to any pre-Conquest site near modern Pevensey is a grant of Caestre near Eastbourne in a 1054 writ of Edward the Confessor.  The writ describes Caestre as having 12 cottages and salt-pan.  Notably there is no mention of a port, mill, market or church, or any other amenity of a town which would have been mentioned as revenue-earners in the writ if they had existed.   The commentary in the linked text speculates Caestre might be Hastings, but all other places in the writ are very close to Eastbourne, implying the abandoned and massive Pevensey Castle, and there was no castle at Hastings before the Norman Conquest.  The only pre-Conquest settlement near modern Hastings was a small Fecamp Abbey cell in Priory Valley to the west, convenient to Hooe.  Beyond Wartling was a tidal salt-marsh that extended more than 22 miles to Pevensey Castle, and more than 20 miles inland into the Weald.  The vast salt-marsh can still be seen on topographic maps, giving an idea of the extent of seascape and salt-marsh when the land was 2 or 3 metres lower.

The other recent breakthrough was realising Lydd, the nearest town to Dungeness,  was probably derived as a local placename from the Norse lith and Anglo-Saxon (mostly northern Anglian) Lyð.  Lith meant arm, limb or estuary, but I think it also meant tidal anchorage.  All the places associated with placenames from lith are liminal.  Lydd, Sussex, is on a shingle bar, a natural tidal anchorage in 1066 or thereafter for anchoring or beaching while awaiting the next flowing tide to enter the Rye Camber. 

Pevenesae, Pevenesel, Pevenessellum in accounts of the Norman Conquest were all references to a shingle spur or island at the gap leading to the Rye Camber, Hastingaport, Old Romney and other medieval harbours.  In the Anglo-Saxon Chronicle too, all references to Pefenesea and Pevenesea make it clear that it is an anchorage for ships.  Ships shelter there, ships lay to there, ships muster there, ships are attacked and seized by Godwin there.  The only reference which indicates some land feature is in a 1087 entry, long after the Norman Conquest, when there was a castele there.  Castele can mean subsidiary fortlet or cell of an abbey, but it can also mean lighthouse or signal beacon in Old French.  A lighthouse at the anchorage to guide ships from the Gallic coast safely to shelter would make good sense.


In 1066 there would have been another shingle bar where Lydd is now, 'almost at the ness', so Peue (almost) nes (ness) sel (isle) or Peue (almost) nes (ness) sae (waters) would be apt, toponymic placenames well suited to guiding navigation.  The Norman accounts and Bayeux Tapestry all agree and make better sense once you know about liminal anchorages.
The Norman fleet crossed on one tide and anchored at Pevenesae/Peuenesel/Peuenessellum while they rested and awaited the next flowing tide which took them into Hastinge portus, the port of the cape of Hastingas in the Brede Basin.


Now that makes sense!  And I'm grateful to Frank Hay and Dungeness Lighthouse for helping me figure it out.