The P-Q split you allude to as the Brythonic-Goidelic split cuts right across Italic too. Compare the words for four and five:
PIE *kwetwer-, *penkwe-
Oscan petiro, pumperias
Latin quattuor, quinque
Welsh pedwar, pump
Irish ceathair, cúig
In P-Celtic and P-Italic, the original /*kw/ is assimilated to /p/. In Q-Celtic and Q-Italic, /*p/ is assimilated to /*k(w)/, conversely.
For clarity, there is a theory that the P-Celtic languages constitute a sub-branch, but the more widely accepted idea is that the change happened separately in non-Celtic Oscan, in Gaulish, and in Brythonic.
Also, there are serious challenges to the idea that the Celts arose in Central Europe. More evidence is coming to light from Iberia and from Eastern Europe complicating the picture nicely.
I’ll ask this while all the experts are here:
Apparently the British Isles were settled by people who could walk across the glaciers. Was there a time when they were cut off from the mainland until they developed the ability to cross the water? Or did they have boats the whole time? Or is that unknown?
The earliest people in Great Britain could walk across land to get there, during a glacial maximum. Even in nonglaciated areas, sea levels were lower. The now-submerged land bridge that once connected Great Britain to the continent is called Doggerland. What is now the Dogger Bank on the sea floor was once the Dogger Hills. They’ve found spear points & stuff on the North Sea floor.
The need for assimilation with penkwe seems clear, especially since counting numbers need to be said quickly. (Imagine a shepherd counting his flock of frisky sheep:
“One, two, three, four, penkwe, six …” 
)
Are you suggesting that the assimilation adopted in the words for four and five motivated more general P->Q (or Q->P) sound changes?
Yes, I am, to a limited extent. The *kw- > *p- holds generally true in Brythonic, although assimilation of this /*p-/ to to this /k-/ in Irish is a special case of assimilation. But the examples I gave are probably the best-known examples usually cited when discussing this topic. It’s a general rule that PIE initial *kw- becomes *p- in Brythonic, *k- in Goidelic, and *kw- in Latin. Oscan I’m not so sure about, but I would like to learn more of it. The Sicel language spoken by my ancestors was either a dialect of Oscan or a closely related language.
I’m glad others also find the Atkinson-Gray clading unlikely. I linked to it because it has date estimates, and shows the Germanic breakup much later than that of Celtic, the question I was answering. Atkinson and Gray aren’t even linguists but their model has become ubiquitous since their publication in Nature.
Ringe’s model is much better, I think. Slightly after the middle of the linked pdf is a clading chart with roughly-estimated time depths. It shows West (OE/OG) and North Germanic (ON) diverging about 200 AD, with Gothic (GO) off about 1 AD. Brythonic (WE) and Goidelic (OI) split 500 BC. It shows Balto-Slavic (LI/OC) as closest to Indo-Iranian (VE/PE) separating as late as 2500 BC, and Greco-Armenian as second closest – as expected, but opposite to Atkinson-Gray.
An interesting comment Ringe et al made in their early work is that it was easy to produce “perfect” trees when Germanic was omitted, but placing Germanic was hard.
None of this answers Dr. Drake’s specific question, which would require an expert – I’m just an interested layman. However, writers like Thomason write that Norse influence on some OE dialects ca 900, despite “less intensive contact” was encouraged by “the pre-existing close typological fit … being closely related genetically, with a maximum separation of perhaps 1000 years.” Obviously the Welsh-Irish split was much older than that of Norse-Saxon. Is there any evidence of “close fit” influence between any Brythonic and Goidelic languages?
I think so, too. I know who Ringe is, and what is on this chart in my area accords very well with evidence I’m familiar with. An interesting chart! That puts common Germanic only a few hundred years before the Germanic invasion of Britain.
ETA: There are borrowings from Brythonic to Irish & vice versa, but the two languages are not a close fit. Both underwent parallel but distinct series of quite dramatic sound changes and patterns of syllable loss. The inflectional systems in Brythonic, in particular, were ravaged, and in Old Irish the verbs in particular are just really, really, really weird.
On the “history” of Anglo-Saxons-- using the term loosely for a reconstruction largely from legendary materials (seldom do the Roman chroniclers have all that much useful to tell us).
“Saxon” appears to have been a generic term for a wide variety of northern Germanic tribes of which Angles and Jutes were two of the larger. Their ruling families claimed descent from Woden, probably an actual historical leader around 200 BC since he was not one of the primordial Indo-European deities (we cannot of course rely on much “biographical” data from the legends, but can consider it likely that he had a very fast horse, and lost one eye); the cult emerges first with the Cimbri and Teutones, which exploded out of peninsular Denmark reaching southern Gaul (the occasion on which Massilia, now Marseilles, sought Roman alliance giving the Roman Republic a strong foothold in Provence).
The three most prominent Wodenite families were the Walsings ruling Angeln, Helsings ruling Jutland, and Turings ruling Thuringia. “Angeln”, now Shleswig in northern Germany, was at that time a three-cornered island, as the Elbe used to fork and empty into both North and Baltic Seas, while the Kiel Canal, now artificially maintained, was a natural channel all the way across (before the west end silted up leaving the east-flowing Eider river). The Walsings (called Volsungs in the Norse sagas) were reputed to have gained their fabulous wealth by slaying a great dragon; actually, they got their money the old-fashioned way, by controlling the sea routes to the amber and fur sources in the Baltic country. Jutland is now “peninsular Denmark” but at that time was an island, which had no Danes on it (the Danes then being in western Sweden and some of the islands in between). The Helsings at one time also controlled a lot of bases in the Baltic trade, giving their name not only to Elsinore in Denmark but also Helsinki in Finland, but by the 4th century were subordinates to the Walsings: Jutland was referred to as a presidium rather than a regnum as its “king” (or “president” maybe we should say?) ruled at the pleasure of the king of Angeln. Here are the stories (as, of course, preserved fragmentarily in several late sources, so take the accounts with a grain of salt) of what happened to those kingdoms:
(per Saxo Grammaticus) Horvendil of Jutland was murdered by his brother Clodo so that Clodo could take his wife Gertrude. Hamlet, Horvendil’s son, played stupid to avoid getting killed, and Clodo did not dare kill the boy himself for fear of angering the people. Eventually Clodo sent him on an embassy to the queen of Pictavia, where inheritance was matrilineal: the king of the Picts was whoever the queen married, and their daughter would become the next queen. Clodo sent Hamlet with a letter asking the queen to kill the bearer; when the queen asked Hamlet what his message was, he answered boldly “Doubtless it is my uncle telling you to kill me”, so she was impressed and married him. He made sure they had both a daughter (to inherit Pictavia) and a son (to inherit his claim to Jutland) before gathering an army to kill Clodo and crown himself in Jutland-- he did not ask the permission of Angeln, resenting the Walsings for allowing Clodo’s usurpation, and so the Angles invaded, killed him, and put a cousin from the Walsing line as the new ruler of Jutland.
(per Tolkien’s reconstruction of the “Battle of Finnsburgh” fragment) In about the 440’s, king Hnaef of Denmark went for a Yuletide visit to his sister and her husband, king Finn of West Frisia (at the mouth of the Rhine; most Frisians lived further east, and are now confined to the offshore isles from the Netherlands). Hnaef brought along Hengest, the Walsing ruler of Jutland (more or less as a hostage, to ensure there was no trouble back home over the winter). Unfortunately, Finn’s entourage included Guthwef, great-grandson of Hamlet and the Helsing claimant to Jutland. Hengest’s Jutes and Guthwef’s Jutes exchanged some angry words, spoiling the welcome banquet. Finn evacuated his palace for the use of his guests, moving into a smaller fort. During the middle of the night, Guthwef’s party broke into the palace and accidentally stabbed king Hnaef: surely they were trying to kill Hengest, but perhaps there were two royal bedrooms in the palace (often a queen liked to have her own space and not sleep with the king all the time) and they guessed the wrong room. The poet has Hengest discovering Hnaef with his life-blood draining: “What is this red I see? Is this the day dawning, or is this a dragon flying, or is this the roof-beam catching fire?” Of course, Hnaef manages the full repetition, “Nay, this red that you see is not the day dawning, nor is it a dragon flying, nor is it the roof-beam catching fire” before expiring.
Hnaef and Hengest’s men chased Guthwef’s men back to the fort where Finn’s men came charging out in response to the alarms; in the confused fighting the Danes managed to kill Guthwef and also Finn’s two eldest sons. Finn ended up besieging his own palace; he was urged to burn the Danes out but did not want to set fire to his own house, and then the weather turned foul, and everyone realized the fight was useless and based on misunderstanding. A truce was called for everyone to burn their dead. In a touching scene, the queen (Hnaef’s sister) insisted on moving her sons from the Frisian pyre to the Danish pyre, laying them by Hnaef “for as his sister’s sons, they ought have died by his side” (the maternal uncle was a culturally important relationship). The problem was that the Danes could not go home for months: they had been planning on staying all winter, and weather too poor for sailing was already setting in. They agreed to switch into the fort, giving Finn his house back, and both sides swore an oath to speak no word of what had happened. In the spring, Hnaef’s young sons came up to Hengest and laid in his lap Hnaef’s sword, Heomleodring, “the sharpness of whose edge was well known among the Jutes”, in a silent appeal to him to decide what to do. Hengest took everybody home as if all was well, but roused the Danes and Jutes together to go back and kill Finn and all his men. Finn’s widow was taken back to Denmark, but Hengest kept Finn’s young son “Half-Dane”.
(various sources from here) Though the poet does not say so, evidently the whole Jutish nation evacuated and moved to West Frisia at this time, letting the Danes take over Jutland: for archaeological evidence does indicate that the Jutes were bringing artifacts from the Rhineland and not Scandinavia when they hit Britain, and linguistic evidence indicates that many Frisian speakers came along with the Jutes to Britain. Hengest was not happy in the new location long: about 451, Attila the Hun intervened in the politics of the Franks, then in Gallia Belgica, forcing them to accept as ruler Merovech, a son of Chlodion of Tournai’s wife, but not of Chlodion; Chlodion had accepted this cuckoo’s egg as his heir, but many Franks were not inclined to and Aetius, generalissimo of Rome, was pushing another candidate; Attila was seeking a pretext for a fight with Rome, and accepted a wedding proposal from Galla Placidia, a sister of the puppet Emperor who was bored at court, and Attila demanded a large territorial dowry. So when Hengest visited Vortigern of Britain (this was a title, not a name: vor-tige is “over-chief”) to discuss a marital alliance, he also expressed a wish to move. Vortigern fatefully invited him to settle on his northern border, to help protect against the Picts and the Saxon pirates. However, once Hengest’s men crossed the Channel they detected they liked Kent much better, and when Vortigern fought back, they also took the isle of Wight (of the two Jutish groups, the Cantuarii were ruled by descendants of Hengest but the Victuarii perhaps for a while by Half-Dane and sons). Vortigern is said to have wandered the country in disgrace, shunned by all, until he died of a broken heart; a son of his, Faustus, fled Britain for Gaul, where he became a bishop and appears in the correspondence of Sidonius Appolinaris, lamenting the sack of his cathedral by (Arian heretic) Visigoths.
Gaul was falling apart in this period: the Visigoths and Burgundians stopped pretending to be “federate” allies of Rome, carving out their own states in the south; in the northwest, the peninsula of Armorica began to be called Brittany, as it became full of British refugees who allied with the Bagaudae (a generic derogatory term for low-class rebels) rather than the remaining Roman authorities; general Aegidius and his son Syagrius maintained a sphere of Romanitas centered around Soissons as a capital, but did not respect the transitory “Emperors” in Italy anymore, Syagrius taking the title “King of the Romans”; the Franks were in civil war. Merovech’s son Chilperic was asked to leave, and went to Thuringia since he had married a Turing princess; there his son Clovis would marry his maternal first cousin Basena (also a Turing princess), and later he would marry Clotilde of the Burgundians, who would convert him to Christianity giving him respectability among the Gallo-Romans; with the strength of all these connections he would unite the Franks and conquer much of Gaul, but in the 450’s and 460’s Gaul was in a mess, and Saxon bands not only took some formerly Frankish territory in Belgium but penetrated into Gaul proper taking a territory around Angers. Britain on the other hand temporarily recovered under “Rigothamus”, again a title rather than a name (rig-utham is “king-supreme”) but the personal name surely was Arthur (I tend to assume that prominent figures in stories really are based on real persons, however little the biographical info can be trusted).
Rigothamus beat back several attempts by Saxon pirates to establish bases on the east coast of Britain (raids by the Saxons had been a recurrent feature of British history for some centuries already), and launched an expedition to the continent, to deter Euric of the Visigoths from expanding north of Aquitaine and to thrash the Saxons of Anger. He was allied with the Burgundians, whose capital at that time was Avallon: which is probably where he died (that would explain much). When he was gone, a Saxon chief named Aelli crossed the Channel and conquered Sussex (the “south Saxon” country). One of his generals went further inland and took Colchester (Roman Camulodunum, from Brythonic cam-lug-dun where Lug was the god of the sun and of astronomy and lore in general, a Lugdun being an observatory/temple, so that a Camlugdun is “temple of the allies”; thus Camulodunum is not necessarily our “Camelot” since the name might have been used for several places that served as capitals of various confederations, much as we find several places in Gaul all called Lugdunum, so the various sites which claim to be “Camelot” might all be right); his line refused to acknowledge the overlordship of Sussex, and called themselves “kings” in Essex (the “east Saxon” country) although it should be pointed out that none of the other “kings” ever recognized Essex as being of equal rank. The more important Wessex (“west Saxons”) was founded by Ceawlin and Cynric, whose relationship is unclear: Cynric was an adopted son, or a son-in-law, or a maternal grandson; the name indicated royal kinship, so perhaps Ceawlin had no prominent family but married his daughter to someone with Wodenite pretensions? It is not clear. In any case Wessex took Hampshire and Wight early and beat the Britons badly in Devon at the battle of “Badon” (precise location unclear) where “Arthur’s men” fought on the British side. The chronicler doesn’t say that “Arthur” fought there, so Arthur would be surely be dead by this time, although whatever “state” he ruled was still somewhat functional: Logres was perhaps the name of the state, subsequently becoming a generic word for “English territory” in Welsh but obviously not derived from Anglo-Saxon so it must have meant something else before.
After these Saxon penetrations of the southern coast and the collapse of any residual overall leadership among the Britons, Angeln was completely abandoned, as the entire nation of the Angles migrated en masse to the eastern coast of Britain. Bede writing over 200 years later said that Angeln was still uninhabited: like the modern Netherlands, it was largely below sea level (sea level was higher then) requiring intricate dike-work to keep its fields usable, so when its people left it took a long time before it was reclaimed. The kings of East Anglia, with their original capital at Walsingham, were from the senior Wodenite lineage. Mercia, which meant the “border” country, was from a cadet branch: it was common to carve off a piece of the kingdom for a prince, to try his mettle before he succeeded to the throne (kingship was not strictly hereditary but somewhat elective, candidates being restricted to members of the royal family but the senior heir not being guaranteed acceptance) and sometimes these “half-kingdoms” would go their own way. The little kingdoms of Deira and Bernicia which eventually merged into Northumbria were from lines which claimed Wodenite ancestry and were recognized as such by the others, but that may have been a polite fiction.
Bob X, may I recommend a more critical evaluation of your sources? I wrote out a critique of some of this, but then I just turned into that mean, pompous guy on the internet.
I am perfectly aware that much of this is doubtful: I said so at the top. By the nature of things, the “Dark Ages” are difficult to reconstruct. I am giving the best that can be done given the nature of the source material that we do have.
There really isn’t such a thing as “Q-Italic”. What happened to the number four is unique to the numerals; other instances of *p from Proto-Indo-European remain “p” such as pater etc.
Yes, but some of your information is flat-out wrong (e.g. the meaning of Camulodunum).
If you have a better etymology for the name, then offer it.
Camulodunum is widely accepted as the dunum “fortress” associated with the god Camulos. http://www.camulos.com/
It cannot be from Brythonic cam-lug-dun because (1) at the date when the name is attested, intervocalic -g- was still preserved in Brythonic, and so any form based on lug- would still have the G: compare luguvalium; (2) most Celtic names are compounds of two words, not three; (3) the prefix cam- on its own means “cooked” (as in Cambridge), which doesn’t make any sense in context.
By the way, Camelot isn’t attested until the twelfth century, so whatever it is, its association with Arthur is purely literary without an ounce of history to back it up.
The first element wouldn’t be that cam- but a different cam- meaning “host; army” similar to the Italic root campus “battlefield” (only later generalizing to “field”, initially specifically “field of combat”). I didn’t know how the L got in there so I guessed it was an eroded Lug but as you say the disappearance of the G would be hard to explain; so thank you for pointing me to this deity Camulo god of “armies” (I still don’t know what the L suffix does here, but maybe you do?). Anyway, my point was that the name could easily have been used for several different places so that there is no necessity to think Colchester was the original Camelot.
So? Lots of things hang around in oral tradition for a very long time before they get written down. I think it is highly improbable that someone in the 12th century made up the name by the equivalent of picking Scrabble tiles out of a box, given that it does strongly resemble forms like Camulodunum which definitely were in usage in the right period, as any purported concocter-out-of-nothing in the 12th century would not have had a way of knowing.
This is the attitudinal point that I was making earlier when I said “I tend to assume that prominent figures in stories really are based on real persons, however little the biographical info can be trusted”; similarly with place-names, where to my mind the Occam’s Razor explanation for why a legendary king Arthur would have a legendary royal residence named Camelot is that a prominent king back then had a personal name sounding much like “Arthur” and his capital was a place whose name sounded much like “Camelot”. 19th century skeptics thought the Trojan War was a “minstrel’s fancy” and mocked Schliemann for thinking there was actual geographic data to be extracted from Homer and a real city of “Troy” out there to be found. The gap between the Trojan War and Homer’s epics is centuries, but of course major events like that were not going to be forgotten. I think the tedious list of names that interrupts Iliad Book 2 is, actually, solid data; I think Agamemnon and Menelaus really were the kings of Mycenae and Sparta, but I doubt that Achilles actually had a great long argument with the high king over the disposition of a female prisoner etc. etc. (though Akhilleus really was a preserved name of a commander).
Camlann is, though.
Bobx, no mention of Cerdic?
What islands? I can’t find any on my map.
“Helsing” as in Helsingborg/Helsingør (Elsinore), Hälsingland and Helsingfors (Helsinki) is a descriptive term and means a place where the water gets narrow. If the people in Helsingborg were called helsingar it’s because they lived there, not the other way round.
Bob X, Thank you for your great post about Hamlet, etc.! Interesting and beautiful!
I’ve read a suggestion that “Cerdic” sounds more like a Brythonic name than a Germanic one. Valid?
Yes; it is a form of what becomes Modern Welsh Caradog, or the place-name Ceredigion. I don’t know enough about Old English to assess whether this must be the etymology, but it’s certainly a possible one from the Celtic side.