Welsh and English

Treating language families as trees that branch is probably the most common metaphor in historical linguistics; though it is a rather poor one. Languages actually work something more like bacterial lines: large groupings of similar organisms with tiny variations between them but broadly the same DNA. The bacterial analogy also works better because in a tree, once a branch splits it cannot be rejoined with other influence; but bacteria can and do exchange material between one another. Of course, the metaphor suffers for being relatively less grand and comforting than a tree, so Tolkien would likely not have tolerated it. And besides, the diagrams can be so beautiful 

(taken from a webcomic called Stand Still Stay Silent)

How do languages "exchange material"? This question is the basis of contact linguistics, a relatively new sub-field. The lowest level of contact is exchange of vocabulary. Powhatan does not need to have profound influence on American English to lend the word 'raccoon' for an animal the settlers had never seen. Words can even spread so far that no one is sure of their origin; take 'orange', which spread to English through at least five languages and probably originated in South India. There is obviously another type of vocabulary exchange, of the type experienced by English with a mass influx of words from French, Latin, and Greek. And there are much higher levels of contact: languages can borrow whole grammatical structures; or they can be so influenced by another language as to lose all of their vocabulary but keep their grammar, becoming creoles; or the community gradually switches to another language entirely. The position that English is a creole between a Germanic base and Romance superstrate has been thoroughly refuted. 

Contact take place within a whole area, usually termed a 'sprachbund'. For example, language of the Balkans such as Romanian, Serbo-Croatian, Greek, Albanian, and Turkish have developed a large number of similar grammatical features, despite being distantly or wholly not related. Other famous sprachbunds are in the Caucasus, in Southeast Asia, in Mesoamerica, and even the whole of Western European. The last is termed 'Standard Average European' and includes English but not the Celtic languages. So the 'climate' a language finds itself in influences it, in the sense that neighboring languages usually influence one another. The (literal) climate of an area also impacts languages; see my last post on sonority.

Contact is not necessarily symmetric. When a language is imposed from a higher level and imparts characteristics onto local languages (as e.g. French on English), it is termed a 'superstrate'; when a local language imparts effects onto a new language in the area which eventually displaces it (as e.g. Irish on Hiberno-English), it is termed a 'substrate'. Welsh undoubtedly has an English superstrate, and Welsh English has a Welsh substrate. However, the number of Welsh loanwords in English can be counted on one hand. The issue of Brittonic influence on English is controversial, mainly because there appears to be incredibly little. Advocates of the Brittonic theory usually assume that 1) huge numbers of Britons switched to English and 2) their influence was mostly on common English, not the conservative, high-prestige register used in written Old English. Both these assumptions fall apart rather easily. The consensus in the archaeological community until fairly recently was that the Anglo-Saxon invasions were overblown and that the English had a large amount of Briton in them. But a very recent genetic study of Anglo-Saxon graves (Gretzinger, Sayer, Justeau et al, 2022) proved that they indeed had overwhelmingly Germanic heritage. The very cause of any debate is overblown itself. Furthermore, many of the supposed Brittonicisms in English only developed in the later Middle Ages or even Early Modern period. 

The first matter is phonology. English is unusual among the Germanic languages for still having "th"-sounds (dental fricatives) and keeping "w" and "v" separate; and it shares these sounds with Welsh. But the influence could run either way. The continental Germanic languages lost their "th" sounds in a wave spreading from south Germany that reached Dutch at about 1200 A.D., at which point there was no feasible way for former Brittonic speakers to still be influencing English. The continued presence of the sounds in Welsh could well be argued to be English influence; Breton, a close relative of Welsh in France, has lost them, and so did Irish, Scots Gaelic, and Manx. Although English vowels, on the other hand, have undergone massive shifts, similar developments occurred in German; English is only unusual for lacking front rounded vowels like German ü, ö. These vowels are also found in French, yet disappeared from (written, highly conservative) English at about the time of the Norman conquest, three centuries after any possible Brittonic influence. But Middle Welsh developed the ü vowel after contact with Old English and lost it after Middle English did. So, to summarize, there are no sounds English has definitively borrowed from Welsh at any stage.

Supporters of the Brittonic substrate theory typically fly straight to grammar. The most shocking development in English is the near-total death of inflection. Or at least, it is shocking when comparing only with Old English and ignoring all in-between stages or what happened to Old Norse or Dutch. Dutch and the continental descendants of Norse have both lost noun cases, reduced their gender system, and severely reduced their . The loss of cases and most verb inflection occurred gradually over centuries, not all at once in 600 A.D. The evidence more suggests Norse influence as the catalyst. The inflectionless trend spread from the North, where there were more Norsemen, and the fact that there are hundreds of Norse words among basic English vocabulary is a smoking gun for strong contact.

The two strongest candidates for Brittonic influence are do-periphrasis and continuous verb forms. Respectively, these are the tendency to use "do" as a meaningless auxiliary verb for negation and questions and the verb construction BE + participle. But both features developed slowly and did not reach their present forms until the Modern period. Do-support first appears in the 15th century, and continuous verb forms likewise originate in two Old English constructions, which 

When Tolkien says Latin is an "old" language and that Old English is a "middle" language, he uses terms unusual in modern linguistics. But the trend he refers to, which has been happening to languages across Europe for the past several thousand years, does have a name: becoming more 'analytic'. "Old" European languages have rich, mostly regular inflection and are called 'synthetic'; "middle" languages lose a great deal due to sound changes and become more irregular; "new" languages like Modern English very have little inflection left. A classic example of a totally analytic language is Chinese, which (more or less) lacks any and all inflection. But there are more factors to language than being analytic or synthetic. The latter category splits into 'agglutinative' and 'fusional', where the morphemes indicate one category in the former and many in the latter. For example, the Turkish ending -lar- indicates just plurality; but the Latin ending -ae (can) indicate not just plural but also feminine and nominative. There is a trend from agglutination to fusion. In fact, there is a trend from analytic to agglutinative again! As syntax becomes more fixed and small words become purely grammatical, words smush together and become inflected. The process of an independent word becoming a grammatical affix is called "grammaticalization". The English affix -n't is actually inflectional by most arguments, and it does not seem a stretch to say that contractions will become true inflection at one point in the future. So there is really a cycle, from agglutinative to synthetic to analytic and back to agglutinative. Over the course of its thousands of years of recorded history, Ancient Egyptian seems to have gone through the whole cycle, starting as fusional, becoming analytic, then agglutinating, and trending back towards fusional again by the time it went extinct as Coptic in the 16th century. So fear not, fans of fusion! Rich inflection will return one day in glory. 

-LN