Thursday, July 16, 2009

DiGlossia in Anglo-Saxon England

As regular passers-by know, I'm interested somewhat avidly in the Adventus period. One of the issues of that period has been what happened to the Celtic speakers so that few Celtic words became part of Old English. The old model of course was that the English came in and pushed the Celts west to Wales and Cornwall, north into Scotland, and off the island to Brittany. That view has been undergoing something of a revision in recent years, sped by the genetic studies of earlier this millenium, but still there is a lack of clarity on the issue of language borrowing and contact. I've written about some of this previously here and here.

So, I finally got around to doing some reading. Believe it or not, I have made inroads into the pile of books and articles that have stacked up since coming to Chicago to finish the ol' PhD and dissertation and all that. An article that was on that pile and is now in the "file" pile is "Diglossia in Anglo-Saxon England, or What Was Spoken Old English like?" by Hildegard Tristram in Studia Anglica Posnaniensia 40 (2004): 87-110. The article essentially argues that the differences between Middle English and Old English are due to the main population of Anglo-Saxon England being "between two languages" to steal a book title (just not the two languages of the that book).

1) Written Old English was likely kept relatively unchanged over the long period of 550 years of the period by deliberate efforts. Dialectal variations are remarkably few; there is very little grammatical change in the structure of the language. It isn't until the reign of Henry I that evidence of such change begins to appear.

2) The earliest Middle English texts show great typological change, grammatically and phonologically. "With apparent suddenness appeared the drift away from syntheticity to analycity."

The paper addresses the over all explanation of this sudden and accelerated movement in the 12th century. Tristram proposes four possible scenarios to explain this.

A) Punctuated Equilibrium Model proposed by Robert M. W. Dixon The Rise and Fall of Languages. Cambridge: Cambridge Univ. Press, 1997. Just like in evolution, this model says that the rate of change in the evolution of languages. A language may change slowly over time and then be subject to radical typological changes caused by a number of factors. There are a number of unanswered problems with this view.

B) A traditional explanation that says that the reduction of unstressed vowels was caused by strong stress accents on the first syllable in Old English words. This led over time to the loss of inflectional endings. (In contrast Tristram notes that German with the same type of stress pattern did not undergo the same type of erosion of the inflectional endings, and notes the same pattern in Old Norse, and even in modern Icelandic.)

C) This explanation says that many endings in OE were simply redundant, and so lost any meaning, and were discarded. Tristram again points to other languages in the Germanic group where the same conditions apply, but the same results did not occur.

D) Language contact with Old Norse caused the change, creating a "creole" situation as the languages were fused. However, Tristram rightly notes, ON and OE were close enough that with minimal effort the two linguistic groups could understand each other. Far more interesting to me and in my mind far more important is that the earliest Northumbrian poems we have from the eighth century already shows signs of "inflectional attrition." (I love that term!)

Tristram naturally has a different explanation. First to be noted is that the areas that deviated the most from what had become in 10th, 11th, 12th centuries "Standard Written Old English" in terms of the inflectional endings were Northumbria and the South and South West(though the degree of deviation is much different in each area). These are areas btw in which the Celtic peoples seemed to have maintained a longer period of influence culturally and linguistically. It is in the SW for example that the "periphrastic-Do" construction arose and periphrastic verbal aspect.

Tristram bases her conclusions on a number of factors, most importantly the recent genetic results showing that in large areas of England the British population remained unchanged except for their political affiliation to Anglo-Saxon overlords rather than the Romans. These Brits depending on where they lived in the island spoke British Latin and Brittonic or solely Brittonic at the time of the Adventus Saxonum. The areas where Brittonic was strongest and British Latin least strong are those areas in which we see the earliest and most sweeping change in Old English, including loss of inflectional morphology. Under the Anglo-Saxons, they likely maintained their native language, in some areas 7-8 generations before making a complete shift to only Old English.

This resulted in Brittonicisms entering Old English, but becoming exhibited in the written language in the early Middle English period. In the North: invariable case and gender inflection of nouns, pronouns, and adjectives just like in Brittonic; invariable article, and fixed word order both features of Brittonic. In the SW early Middle English, there is the previously mentioned rise of periphrastic aspect like Middle Welsh and the periphrastic DO in a variety of uses all matched in Middle Welsh. There are other features that appear later in Middle English that are shared with Brittonic or Middle Welsh. Tristram points to one example of this: clefting. She notes that clefting appears in OE in the late West Saxon Gospels, but only more commonly in ME in the 13th century. It already appears in Old Irish and Old Welsh.

Finally, Tristram outlines "levels" of Old English: the written language which was somewhat artificial, the language of the elites, the ruling class, and the "low" level, the language of the common person in Anglo-Saxon England. Middle English's differences from Old English are to be explained largely by the third level coming into the foreground after the removal of the Anglo-Saxon elite, post-11th century invasions. Thus, the elite control of the written language was lost, and the elites were gone to be replaced by a new elite, but this elite spoke and then wrote the language that had been that of the "common man" in Anglo-Saxon England.


Anonymous said...

Interesting... I wonder if glosses and marginal notes deviate more from 'proper' Old English than full works? This would include things like the translations of Caedmons hymn that I believe come in a couple variations (Northumbrian, West Saxon etc). Given how little full Old English works there are it would be fairly easy for schools to enforce 'proper' Old English. Interesting to think a version of the 'Queen's English' may have been around even then vs. regular spoken English.

Anonymous said...

If you want to get that late on these questions, the Chris Lewis paper I blogged about a while back comes to mind, where he was suggesting that the reform movement of the tenth and eleventh centuries did not just generate most of our OE source material, but also invented a learned vocabulary to conduct political argument in (he referenced someone else on this, I can look it up in my notes if it would help). Makes me think that Michelle is right that we should be looking outside the mainstream for where the languages were really mixing, in the same way as I wouldn't look at Carolingian court poetry for Romance symptoms in Latin...

Carin said...

Michelle's suggestion that glosses are where we'd fine a more personal/less standardized language that might reflect the way the language was actually moving seems intuitively right from a modern p.o.v., but for the AS period, a lot of the evidence runs the other way, and the use of that evidence for tracking the kind of language change we're talking about here is problematic.

Consider, for instance, that glosses - at least continuous interlinear glosses - are a major vehicle for transmitting the "authorized" language of the Benedictine Reform (cf the literature on "Winchester Words"). More generally, a glossed text is likely to be the product of a school, and thus to have come from an environment where scribes were aware of standards.

And if we are interested in inflectional and syntactical change, glosses are problematic evidence for two reasons. One, OE words and phrases used to gloss Latin constructions are more likely rather than less to retain (perhaps old-fashioned) OE inflectional distinctions, because their purpose is to clarify the grammar of the Latin. (As a modern comparandum, consider that even if you rarely use "whom" in your own, informal spoken English, you will use it rigorously and correctly if you are trying to explain to students how to translate Latin relative pronouns.) Two, it is very much an open question whether the syntax of continuous OE glosses reflects "real" OE syntax or the Latin of the target text or a combination; even where the OE does not track the Latin word-for-word, an apparently simplified, more direct, more modern word order in the OE may reflect pedagogical practice for clarifying Latin syntax that is not specific to any vernacular.

All of the above reflects two useful generalizations about glossing language: first, that it is in many cases likely to be just that, a specialized language for glossing, a hyper-learned idiom; and second, that the glosses we have very rarely reflect a spontaneous, ad-hoc response of a particular reader and are much more likely to have been transmitted along with the text.

Anonymous said...

Perhaps I shouldn't have said glossing. I mean short marginal notes or translations of very short passages like Caedmon's hymn. I certainly do not mean complete interlinear glosses. Those would be the products of schools or at least teachers. I mean marginal notes or notes on fly leaves in personal copies of books like the Stowe missal.

Carin said...

Fair enough! I'm glad you said "glossing" the first time, though, because it got me thinking about how rare language unmediated by a written standard is in our period.

Larry, I'm not at all convinced that Tristram's carts aren't pushing her horses, but it seems to me (not having read her article) that the part about changes manifesting themselves in eME because of the removal of the structures that perpetuated the written standard is old news, and the new part is the claim of Celtic influence in certain regions. Yes?

theswain said...

Hi Carin,

Yes, the old news is the appearance of those features in eMiddle English. Tristram spends time debunking the current theories as to why those seem to appear all of a sudden in the 12th century. She then proposes that these features appear in those areas where Celtic speakers maintained their language longest and were bilingual, Old English being for a long time their second, learned language. If she's right the implication is that early Middle English is actually the spoken Old English of the lower classes in the Anglo-Saxon period.