Our last little installment of dialects! I know that this is a historical linguistics’ blog, but, today, let’s talk about Modern English, shall we?
Before we can do that though, we need to talk about something else: the distinction between a dialect and an accent.
Up until now, I haven’t made this distinction because it hasn’t been truly necessary; you see, when talking about Middle English and Old English, the term dialect holds quite true. When it comes to modern English, however…
Not as much.
Although often used interchangeably, the terms dialect and accent actually refer to two different things in linguistics. So what is an accents and what is a dialect?
Well, an accent is one part of a dialect.
That… didn’t clear things up, did it?
Alright, an accent refers to how people pronounce words, while a dialect is much more all-encompassing and includes pronunciation, grammar and vocabulary.
As I have been focusing on the England in this post, I’ll be focusing on the accents (commonly called dialects) in British English, but don’t fret! I’ll come back to other varieties of English (like American English) in a future post.
Yeah, that was a bit creepy, but hey, what can I say – I am a horror-flick fan.
Anyway, the accents of (British) English!
Trudgill divided the accents of English into ten (!) different accent regions. In no particular order, with their accent name in parenthesis following, these are:
East, North, and South Midlands
East Anglian (traditional)
Central and lower North
Trudgill divided the accents into these groups based on a simple sentence: very few cars made it up the path of the long hill.
Ignoring the function-words here (that is it, of, and the), Trudgill recorded the pronunciation of these eight words and noted the following:
"y" in "very"
"ew" in "few"
"ar" in "cars"
"a" in "made"
"u" in "up"
"a" in "path"
"n" in "long"
"hill" in "hill"
In addition to these features, the absence or presence of the so-called trap-bath split was also recorded (under the feature path). The trap-bath split is a vowel split by which some words come to be pronounced with a long /ɑ:/, mostly in the southern English accents, and short /a/ in the northern ones. If you are unsure of how that would sound, check out the sound examples at the Pronunciation Studio.
Using this fairly simple sentence, it was possible to discern some general patterns of accent “boundaries”, thus creating the accent-boundaries of modern (British) English! Using the results, it was then possible to divide the country into six major dialect areas:
Isn’t that quite amazing? (and, as usual, a bit ridiculously oversimplified)
However, there is one accent that I haven’t mentioned yet: Received Pronunciation, or RP.
Also known as Received Pronunciation, the Queen’s English, BBC English, Standard British pronunciation or Southern British pronunciation, RP is a highly prestigious “standard” accent in Britain. However, very few British English speakers actually speak RP: Trudgill estimated only about 3% in 1974. This has since been questioned but the highest “guestimates” appear to be 10% – which is really not a very high number any way.
And there you have it – the British English dialects!
I hope you enjoyed that little tidbit, but check out the references if you want to learn more – because, naturally, I can’t go through all of the details here (nor, if I am frank, do I know them) and there is a lot more to learn!
Join me next week when we go back in history again, and take a look at Old English morphology! Until then!
If you want to learn more about the difference between dialects and accents (and dialects generally), check out this OED blog post.
On a more formal level, Trudgill’s study was reported by Ossi Ihalainen in The Cambridge History of the English Language, Vol. 5, where you can read more about the study. Or go straight to the source, which in this case is The dialects of England by Peter Trudgill (1990).
I’ve also had a brief look at Benedikt Szmrecsanyi (2013) Grammatical Variation in British English Dialects.
While I would normally move on to modern English dialects at this point, I did promise you a treat. So allow me to introduce you briefly to two marvelous sources in the study of Middle English dialectology: LAEME and eLALME.
The abbreviations stand for A linguistic Atlas of Early Middle English 1150 to 1325 (LAEME) and An electronic version of a linguistic atlas of Late Mediaeval English (eLALME). There is also a LALME, which stands for A linguistic atlas of Late Mediaeval English, and is the print version of eLALME (the latter was created years after the study itself was conducted and LALME printed).
Before I dig into these marvelous things, I can hear you asking “what is a linguistic atlas?”, so let’s start there!
A linguistic atlas is basically what you’d expect: a collection of maps. In this case, it is a collection of linguistic maps, which show the geographic distribution of speakers of a language. It can also show isoglosses, which separate areas that have a particular linguistic feature, of a dialect continuum.
Okay, so, what is a dialect continuum, you ask?
A dialect continuum is a situation in which a dialect spoken in one area differs only slightly from the dialect in the neighboring area. The further away we go though, the more differences we will find. Eventually, the varieties might be so widely separated that they are no longer mutually intelligible.
Both LAEME and eLALME are linguistic atlases, as their titles tell us. Specifically, though, they are linguistic atlases of Middle English, not modern. Their mapping of dialects concerns the dialects of Middle English, which we took a look at in last week’s post.
Now, obviously, when working with historical dialects, you’re in much more troublesome circumstances than when working with modern dialects. Why? Well, the introduction of LALME (reprinted in eLALME, of course) describes it pretty succinctly:
“It is rather as if the compilers of a modern dialect atlas had access to any number of speakers, all willing to be interviewed but very few of whom divulged where they came from”
§2.3.1. from A linguistic Atlas of Late Mediaeval English, Volume 1, General introduction
Can you imagine what that’s like? Sounds hard, right?
Well, I’ll wager that creating these tools was very hard. But they managed! How, you ask?
With the use of so-called anchor texts.
An anchor text is a text that can be localised to a specific place, or near it, on non-linguistic grounds. This may be letters that, for example, were written by a nobleman whom we know lived in East Anglia for all of his life. It could perhaps be records of courts, or manors, or legal instruments, and so on.
Once you’ve done that, you can start looking for linguistic features in those texts. Let’s say that an anchor text in Middle English uses a specific spelling for a word. Okay, so that spelling might then be characteristic of the dialect in that area.
And then, you look at another text from another place. That text shows a different spelling for that same word. And then you look at a third text, from somewhere between those first two areas. That text uses a spelling with some characteristics of both the first spellings.
Brilliant, you have a dialect continuum!
Alright, so that was ridiculously over-simplified, but you get the idea.
Anyway, my over-simplified explanation (kind of) describes what they did for LALME. They used a questionnaire, consisting of many words, looked in the manuscripts for those words, and voilà! There you are, a dialect continuum of the late Middle English dialects.
Alright, again, that was ridiculously over-simplified.
LAEME took a slightly different approach. It kind of had to as there are so few documents left from early Middle English. So LAEME used a corpus-method instead. They transcribed all early Middle English texts (or long passages from the really long ones) and put them into a database.
This allowed them to elicit all variations from the surviving manuscripts, meaning that they weren’t limited to particular words. Having done so, it is then possible to look for linguistic features that are used in one text and compare them to another, thus seeing dialect variation (see LAEME’s Introduction, Chapter 1, §1.5.5. for more info).
Now, one last amazing thing about these tools: the fit-technique.
Basically, what this means is that you take features of a dialect and, as more linguistic features are added, the area from which the person comes becomes more and more constricted.
Let’s say, for the sake of an example, that you have two Swedish speakers. You ask them to pronounce the word “räka”.
One of them says “räka” with a fairly open mouth, producing [æː]. The other, on the other hand, pronounces it more like “reka”, with a reasonably closed mouth, producing something more like [e:]. Okay, there is something different here, so you ask them some more questions, and hear the same thing. So you go to your “anchor person”, who you know for a fact is from Stockholm. You ask him/her to pronounce “räka” and your other words, and you get “reka” in reply.
Great, one of your participants is probably from Stockholm (obviously, though, one shared feature is not enough). Then you look around, and you find someone who you know is from Gothenburg. You ask the same of him/her and get “räka” in response. And there you are!
That, though again grossly over-simplified, is the fit-technique. You take an un-localised text and compare it to the variations you find in texts that you can place. The more features you add, the more you can constrict the area that text is likely to come from! Isn’t that quite amazing?!
Before I get to the “warnings” of this post, I’d like to encourage all my readers to check these resources out. Truly, they are quite amazing!
You can find LAEME here and eLALME here. Do take a minute, really.
Right, so on to the warnings.
The most important thing first:
While we might be able to say that this text was written in the northern dialect, it does not mean that it was actually written in the north.
For example, the text you are looking at was written in a dialect from southwest Yorkshire according to LAEME or eLALME. That does not mean that it was actually written in southwest Yorkshire.
Neither LAEME or LALME are geographical atlases, they are linguistic ones. Your results simply mean that the person who wrote this text was likely from southwest Yorkshire (or at least wrote in that dialect). The text itself, however, might have been composed in London. Or in Sussex. Or in Worcestershire.
You get the drift. Point is, you cannot conclude with certainty that a text was written in a specific area, only that the person writing the text wrote in the dialect from that area. Obviously, it could have been written there, but you can’t say for sure.
That’s probably the most important thing to remember here but one more little thing: the fit-technique works somewhat better for eLALME than for LAEME. This is simply because we do not have as much data from the early Middle English period as we do for the late period. In the introduction to LAEME, Margaret Laing (LAEME’s compiler) and Roger Lass write:
For much of LAEME, the display of linguistic data in map form at all is a convenient but highly generalised abstraction.
LAEME, Introduction, Chapter 1, §1.5.3.
and this must be kept in mind when using the resource.
Did I get carried away? I did, right?
Well, I’m done now, but again: Check these resources out!
For some more information on these fantastic resources, check them out by following the links in the post (or here: LAEME and eLALME). You can also have a look at the Angus McIntosh Centre for Historical Linguistics introductory remarks (with, in the case of LAEME, a link to Rhona Alcorn’s beginners guide to LAEME) to the resources.
Today, I’ve promised you the Middle English dialects and that’s what you’ll get!
First, a bit of a recap, though.
When I say Middle English (or ME, which is the usual abbreviation), I am talking about English as it was between (roughly) 1066 – the Norman invasion – and 1500. Now, obviously, there is no exact date: people didn’t talk Old English one day and woke up the next speaking Middle English. But it is a convenient way of dividing the history of the English language into manageable chunks.
We must also remember that we can divide Middle English into two parts: Early Middle English (c. 1066-1300) and Late Middle English (1300-1500). The closer we get to our own age, the more recognisable the language will be.
So, now we know that. Let’s look at the dialects!
The Middle English dialects are commonly divided into five distinct dialects: Kentish, Southern, Northern, West Midlands, and East Midlands.
The Kentish dialect is also found in Old English but during Middle English, the area in which the dialect was spoken diminished. By late Middle English, it was spoken only in Kent and Sussex.
The Southern dialect was also spoken in (west) Sussex as well as south and southwest of the Thames. This dialect is a descendant of the West Saxon dialect in Old English and was quite conservative. It didn’t show a lot of influence from other languages which makes it an interesting topic of study!
The Northern dialect is an interesting one too. It is in this dialect that we find very rapid developments in morphology and syntax. This may be due to intense contact with Old Norse, but that’s simply a hypothesis – it is quite possible that these changes would have happened regardless.
Last, the East and West Midlands dialects. These dialects are a bit of an intermediate between the conservative dialects of the south and the fast-moving ones in the north.
What is particularly interesting about Middle English is, of course, its spelling. You see, during the Middle English period, there was no standardised spelling, meaning that people spelled according to their own dialect – which gives rise to some interesting variations.
Which is actually what we’ll look at next week! You see, this was primarily just a little primer so that you’ll know a bit about the Middle English dialects; next week, we’ll get to the really interesting stuff: a brief introduction to two amazing resources when it comes to studying Middle English dialectology!
I won’t tell you which though (though I’m sure some of you are already quite familiar with them) – consider this a teaser.
Until next week!
For this post I have used the Harvard’s Geoffrey Chaucer Website’s entry on the Middle English dialects as well as this entry (which appears to be by Randall (2000), although the link itself gives no author).
I know that we’ve talked quite a bit about English and its history. (You’ll have to excuse me, but the topic is what I’ve studied for years after all.)
I also know that it’s been a bit disjointed. One week, we’ve been talking about English and the next week about something else entirely. That’s what happens when you’re several people working on the same thing (and it’s a good thing too!).
However, now, it’s just little old me. So, I’d thought we’d run through a standard little “course” on the topic and go through it a bit more systematically (don’t worry, we’ll do something similar with other languages following this one).
Originally, we were supposed to start with Old English phonology today, but, I went back and had a look at our previous posts of Old English. Doing so, I suddenly realised that we never really talked specifically about the Old English dialects.
So let’s do that! But first…
I think I need to give you a very brief reminder about what Old English actually is. As you know, English is usually divided into time periods (and if you want all of them at once, take a look at Rebekah’s earlier post here. Otherwise, get back to me next week when I’ll talk about Middle English).
Old English is the English language as it looked until roughly 1066. This is not from the very beginning of the world, so to speak, but from roughly the time that we start getting written records of English (ca. 450 AD – before that, we usually talk about “Proto-English”).
That’s it (for now).
Now, next step: when I say Old English, what I am actually saying is the West Saxon dialect of Old English.
But it was not the only Old English dialect.
I’ve shown you this map before in my Early Germanic Dialects series:
But, while I warned you about how Old English tends to equal the dialects of West Saxon, I didn’t actually say anything about the other dialects.
Let me fix that!
So. Old English had four commonly recognised dialects: West Saxon, Kentish, Mercian, and Northumbrian. Each of these dialects* was associated with an independent kingdom in the British Isles.
Of these dialects, we know most about West Saxon. However, the earliest surviving Old English materials are actually written in Northumbrian.
Spoken from the Humber (now in England) to the Firth of Forth (now in Scotland), the Northumbrian dialect is recorded in texts like Cædmon’s Hymn, a short poem composed between 658 and 680. It is the oldest surviving Old English poem and one of the oldest surviving samples of Germanic alliterative verse. This is made all the more impressive by the fact that it was, supposedly, composed by an illiterate cow-herder.
We also find surviving examples of Northumbrian in Bede’s Deathsong (a five-line poem that supposedly is the final words of the Venerable Bede), the runes on the Ruthwell Cross from the Dream of the Rood, the Leiden Riddle, and the famous mid-10th-century gloss of the Lindisfarne Gospels.
Northumbria was, however, overrun by the Vikings during the 9th century. As a result, most of the written records of the dialect have been lost.
The same is the case for Mercian.
The Mercian dialect was spoken as far east as the border of East Anglia, as far west as Offa’s Dyke (bordering Wales), as far north as Staffordshire and as far south as South Oxfordshire or Gloucestershire – basically, it was a pretty huge dialect.
But then came those pesky Vikings… And Mercian goes the same way as its sister dialect, Northumbrian. (The two dialects together are often talked about as Anglian.)
As with Northumbrian, we do have some surviving textual records of Mercian, but very few. These include the Old English martyrology, which contains 230 stories about the lives of saints and was probably compiled in Mercia – or by someone who wrote in the Mercian dialect anyway. We also have six hymns in the Vespasian Psalter that are written in Mercian, but that’s really pretty much it.
And then, we have Kentish.
Now, Kentish didn’t quite suffer the same fate as Mercian and Northumbrian. Despite that, according to Baugh and Cable, even less material from Kentish survives than from the other two dialects. We could speculate as to why, but that is an exercise in futility – it happens sometimes, unfortunately.
Kentish, as the name tells us, was spoken in the county of Kent. It was eventually submerged in the West Saxon dialect. Most of our surviving textual records are early law texts, for example from the Kentish kings Hlothere and Eadric. However, the surviving materials were late 12th century copies and studies have shown that they have been altered and “modernised”. That means, unfortunately, that little of what survives of the dialect is truly representative of the dialect itself.
And thus, we are left with West Saxon.
Originally spoken in the kingdom of Wessex, West Saxon is typically divided into two: Early West Saxon and Late West Saxon.
Now, Early West Saxon is the language used by Alfred the Great. Aside from keeping the Vikings at bay, Alfred avidly encouraged education. He even translated some things himself. However, this is not the dialect we mean when we say Old English.
What we mean is the Late West Saxon dialect – yes, I know this is getting confusing. But, following the Athewoldian language reform, started by Bishop Æthelwold of Winchester, Late West Saxon emerged. Some even argue that Late West Saxon is not a direct descendant of early West Saxon! This is the dialect we talk about when we say Old English.
We have quite a bit of surviving evidence from Late West Saxon – if I were to try to count them up, we’d probably be here ’til New Years. So I won’t. But I will say that this was the first standardised written language in England, sometimes referred to as the “Winchester standard” (as it was primarily used in and around the monastery at Winchester). This is the language that you find in evidence in the Old English poem Beowulf (though it is worth mentioning that you also find some Anglian features in the poem).
And those are our four Old English dialects!
Next week, we’ll continue with something else tricky: the Middle English dialects. Join me then (if you dare)!
*This post actually triggered a very interesting discussion – are the Old English dialects really dialects or languages? As you know by now, the separation between language and dialect is a tricky one (linguistically) (and if you can’t remember why, check out Lisa’s post on this topic here), but play with the thought for a bit: should the language/dialect of an independent kingdom be considered a dialect in this instance – or is it a language, regardless of the close similarity to another nearby kingdom’s language?
Insights from the ISLE Summer School, 24-28 June 2019
This week, I had the pleasure of attending the International Society for the Linguistics of English (ISLE) Summer School. The summer school is bi-annual, and explores different themes each time – this year, the theme was using the past to explain the present, with the description: “A special focus will be on evidence for past states of English and Scots, with reference to the functioning of writing systems in manuscript and printed contexts.”
With a theme like that, there’s no wonder that this summer school caught the interest of two HLC:ers: Sabina and myself (Lisa)!
The summer school was organised at the University of Glasgow by the ISLE president, Professor Jeremy Smith. On the first day, he held a workshop which led us to think more about how the past can help us explain the present, and he emphasised the importance of considering that the old languages and writing systems we study were produced by people who were as much conditioned by social factors as we are today. In fact, the name of this year’s theme is a scrambled version of a pioneering publication by famous sociolinguist William Labov, On the use of the present to explain the past, which explored the idea that humans are not so different in history and today, and thus we can use our knowledge of today’s languages, and the people who speak them, to make inferences about history. Likewise, through looking at material culture (for example scribal practices, and the look and material of manuscripts), and through exploring the social context in which they operate, we can learn more about what drives language change.
The exploration of manuscripts continued into the workshops in the morning of the second day. Professor Wendy Scase from the University of Birmingham held a workshop about writing systems, and made us aware of the social factors which may condition how we write. The traditional view of spelling is that it follows pronunciation, but it’s not usually that straightforward, and there are often social cues in what spelling systems we adhere to.
One simple example is, of course, the differences between British and American English; the use of colour or color says nothing about pronunciation, but reading one or the other immediately tells you something about the writer. Consider also things like “heavy metal umlaut”, as found in the band names Mötley Crüe and Motörhead; these umlauted letters are pronounced a certain way in the languages who use them in their writing systems, such as Swedish and German, but these bands use them as a form of identity marker. If these social identity markers are used in the present day, we should be aware that this may also be the case in the past. As an example of this, a mediaeval writer may have chosen to use the italic script to advertise to the reader that they are a humanist.
As the second day progressed, we received introductions on how to use historical corpora by Dr Joanna Kopaczyk (University of Glasgow) and Dr Kristin Bech (University of Oslo). While these workshops were more focused on presenting resources for doing research in historical linguistics, the theme of the week still ran like a red thread through them: for example, we were reminded that when looking at historical written text, the scribal practice should not only be taken to be dialectal, but can also be socially conditioned.
On the third day of the summer school, we went on a field trip to Ruthwell Cross, in Dumfriesshire. The runes inscripted on the cross make the earliest evidence we have of Anglo-Saxon in Britain, and it was interesting to learn about some of the unique features of the runic system which are only found on this monument, which again led us to think about what the purpose was behind using these particular symbols.
In the final two days of the summer school, all participants presented their PhD research, and we reflected on the mechanisms behind language change in a discussion led by Jeremy Smith. In this discussion, we looked at different examples of words or expressions which use and meaning had changed in the history of English, and whether social factors may have driven these changes. In the HLC’s weekly etymologies on facebook, we have sometimes demonstrated how social associations may trigger the meaning of a word to become more negative or positive – an example being the word ‘villain’, now a pejorative term, which developed from simply referring to someone living on a farm. This is only one type of language change that can be socially conditioned, and this week we’ve come to learn even more about how identity markers and other socially conditioned factors play a role in how we express ourselves, both in writing and speaking. This is why it’s so important for historical linguists to approach our textual sources with the same sociolinguistic awareness with which we would approach today’s spoken data.
Personally, I found this week to be incredibly inspiring, and in our final discussions you could tell that we had all received plenty of input and inspiration for continuing our research with some more attention to material culture and social practice.
And EGD is back! Today, we’re going to be talking about something close to my own heart: English! This is Early Germanic Dialects thought, so, naturally, we won’t be talking about modern English, but, Old English.
Now, before we start, let’s make one thing very clear: Shakespeare is not Old English. Nope, nope, not even close. In fact, some native speakers of English (and I’ve experimented on this with friends), don’t even recognise Old English as English. Let’s compare, just so you can see the differences. These are the first two lines of the epic poem Beowulf:
Hwæt! Wé Gárdena in géardagum
þéodcyninga þrym gefrúnon
Listen! We of the Spear-Danes in the days of yore
of those clan-kings heard
of their glory
A bit different, wouldn’t you say? And now, of course, you’re wondering how it went from that to this? Well, that’s a different story (but we’ve told it in bits and pieces before).
Let’s today simply focus on Old English, shall we?
Right, so as per usual, let’s start with a bit of a history lesson!
As you might know, while English is today the dominant language of the British Isles, this was certainly not always the case. In fact, the tribes that we eventually consider “English” were all invaders or immigrants: Saxons, Angles and (maybe) Jutes! The native population of the British Isles were, the stories tell us, treated rather horridly – primarily thanks to the Celtic king, Vortigern, who ruled there during the mid-fifth century, who made a really bad call.
You see, Vortigern had a problem: the Picts and Scots kept attacking him and he simply couldn’t deal with these vicious barbarians on his own! So, he called in reinforcements! That means, he invited Saxons to come over to deal with the problem.
And they did. Then, I suppose, they were chatting amongst themselves, and with their buddies who were already living there, and thought “wait… If he can’t deal with these people… How would he possibly be able to deal with all of us?”. After, I imagine, a bit of snickering and laughing, they went off and told Vortigern – pleased with himself after the Picts and Scots had been pushed back – that they weren’t intending to leave. I imagine that left him less pleased.
It is actually from this period in time (or somewhat later), around the year 500, that we get the legendary myth of King Arthur. During this time, a great battle was fought at someplace called Mount Badon (which we can’t really place), and the British people succeeded in stopping the Anglo-Saxon expansion for a little while, and they may (possibly, maybe, we don’t really know) have been led by a king called Arthur (kinda little historical evidence for one of the most widespread myths out there, right?). Despite this success, a great deal of southern Britain was in the hands of the Anglo-Saxons by the year 600, and the areas under British rule had been reduced to distant corners of the west, such as Wales and Cornwall. What we end up with, is a geographical division that looks something like this:
Now, naturally, when people come together in close quarters and multiple leader-types, what follows is about 300 years of squabble about the ‘overlordship’ of this green area. Then… Then, they had other things to worry about – the Vikings had arrived.
But we’re not gonna talk about that today, so check it out here if you want!
So, the Vikings arrived, and this led to a long war. Eventually, King Alfred the Great of Wessex forced the Vikings to peace-talks (mostly because he kept beating them, though he might have been pretty much the only Anglo-Saxon king who could boast about that), and the Danelaw was formed.
The descendents of Alfred managed to keep things pretty smooth for a while. Specifically, until 978, when King Edward was murdered. Enter: Æthelred the Unready (and no, that is not a nickname that history added: his own contemporaries called him unræd, loosely translated as ‘ill counsel’). Basically, he did most things wrong (even attempting to order the death of all Danes in the country). The, probably, largest mistake that Ætheldred did though, was the decision to kill the sister of King Swein of Denmark.
Riled Vikings? Really, that’s a bad idea.
And in 1013, Æthelred was shown just how much of a bad idea that was, when a pissed-off Viking army landed on his beaches. The army of Danes met little resistance and Æthelred was forced to flee to Normandy. However, Swein died just a couple of months after that, and Æthelred returned to England – only to be re-invaded by Canute the Great, son of Swein, in 1015. Æthelred eventually died in 1016, and his oldest surviving son Edmund died soon after, leaving Canute the ruler of England.
Canute’s sons, Harald Harefoot and Hardecanute, ruled after his death, until 1042, when the son of Æthelred and Emma of Normandy (Hardecanute’s adoptive heir) Edward took the throne, which he held onto until his death in 1066. And we all know what happened after that… Enter the Norman invasion. Though Harold Godwinson, Earl of Wessex, was acclaimed king after Edward, he held the throne for only nine months before he fell at the Battle of Hastings, thus putting a bloody end to the (fairly bloody) Anglo-Saxon state.
Alright, let’s talk language!
Though we have a number of surviving texts from Old English (a lot more than many other of the EGDs that we’ve been talking about), a lot is, of course, lost to us. What does survive, and what we really mean when we say “Old English”, is the late West Saxon dialect. The reason for that is simple: most surviving texts are written in that dialect. But, when studying Old English, it’s worth keeping this in mind: we’re not (necessarily) talking about a unified language; we’re talking about a dialect that happens to be primary in the surviving materials.
Anyway, first, as per usual, let’s look at some phonology!
Most letters of the Old English alphabet are fairly uncomplicated for a speaker of modern English. Some, however, have surprises in store.
One of those letters is the letter <g>. This letter is pronounced as in modern English ‘good’ only when it follows [ŋ] or when it’s doubled:
cyning ‘king’ frogga ‘frog’
Before the front vowels i and e, after them at the end of a syllable, and also in a few instances where <j> or <i> originally followed but has since disappeared, <g> is pronounced like the first consonant in modern ‘yes’. Before back vowels, though, <g> was pronounced [g].
Elsewhere, <g> is pronounced as a back fricative (remember Rebekah’s phonology lesson on consonants?), unless it is a sequence of <cg>, in which case it is pronounced as the first sound in modern English ‘giant’.
Another sequence that has a surprise in store is the letter sequence <sc>. Although a modern English speaker might expect that <c> here actually corresponds to [sk], it doesn’t. Instead, it would have been pronounced something like [ʃ], that is, the first sound in modern English ‘ship’ (as, indeed, also Old English scip).
Last, in this part, we have the letter <h>. While seemingly simple enough, <h> is pronounced [h] only in initial position and before vowels:
But before consonants, and when occurring in word-final position, <h> is pronounced as [x], a sound today found in German nacht or Scottish loch:
feohtan ‘fight’, here pronounced with [x].
In the vowels, Old English shows a number of changes that are not found in the languages discussed so far in our little EGD series. For example:
Like most other Germanic languages (except Gothic), Old English originally changed the vowel [æː] into [aː], yet under most circumstances (though especially before w), it changes back to æ:
Similarly, in most cases, the change of short [a] (which usually also changes into [æ]) systematically fails to take place when <a> is followed by a single consonant, plus <a>, <o>, or <u>:
daga (dat. sg.)
Except before nasal consonants, where long and short <a> instead becomes long and short <o>:
Now, something rather interesting before we move on: in Old English, we find evidence of a process known as assibilation. This process, which is shared only with Old Frisian of the Germanic dialects, means that the stops k and g becomes [tʃ] (as in church) and [dʒ] (as in drudge) respectively. This process is also the one responsible for correspondences like skirt/shirt, where shirt is the assibilated Old English form, while skirt is borrowed from Old Norse, which did not undergo this process, and thus retains a hard [k] sound. Interesting, isn’t it?
Now, I’m going to break tradition a bit and not really talk about morphology. Instead, I want to say a few words on syntax, that is, word order. Why? Because the syntax of Old English is not quite the same as the syntax of modern English. In fact, it’s rather markedly different.
Most notably, Old English is significantly more inflected than modern English: it inflected for five grammatical classes, two grammatical numbers and three grammatical genders, much like modern German. While this may be frustrating to students of the language, it did mean that reliance on word order was significantly less than it is today because the morphological form would tell you who was the subject, object, etc. This means that Old English word order was a bit less rigid than in modern English (in which, it is the only thing that shows you that there is a difference between the dog bit the man and the man bit the dog).
Generally speaking, the standard rule for Old English is that it has a verb-second word order, that is, the finite verb takes the second position in the sentence regardless of what comes before it. So it really doesn’t matter if the first element is the subject or the object, the verb holds its second position (in which case, the declension of the words become important for understanding the sentence correctly).
However, this holds true only for main clauses. In subclauses, Old English is (generally speaking) verb-final, that is, the verb winds up at the end of the sentence. Students of modern German (such as myself in fact), may recognise this kind of word order.
On the topic of syntax, I would like to wrap this post up with a cautionary note.
If you’re reading Old English poetry (and sometimes even when you’re reading prose): chuck these ‘rules’ of Old English syntax out the window. They won’t do you any good: in Beowulf, for example, main clauses frequently have verb-initial or verb-final order while verb-second is often found in subordinate clauses. So heads-up!
Right, that’s all I had for today, though, obviously, this is a very small appetizer in a huuuge buffet. If you’d like to learn more, we, as always, refer you to Robinson’s great book but, to be quite honest, the chapter on Old English is quite dense and even I had to refer a couple of times to Wikipedia and other sources just to make things clear. However, it is a good starting point so do enjoy!
As always in our EGD-series, our main source is Robinson’s Old English and its closest relatives (1992).
For this post, we’ve also taken a look at:
The passage of Beowulf, with its translation, is by Benjamin Slade: you’ll find it – and the rest of the translation of Beowulf – here
In many varieties of English, a W is a W. In these varieties, W sounds like [w] like in ‘wise’ and ‘wonderful’ and ‘wowza’ (unless it’s at the end like in ‘draw’ or ‘stow’, in which case it’s quiet as a mouse.)
However, in Scottish, Irish, New Zealand, and certain American dialects, wh-words are pronounced a little different. In words like ‘which’ and ‘whale’, the H makes the W kind of…H-y.
Why is that? Why are those words even spelled with an H to begin with? As with many questions about the bitter rivalry between English pronunciation and English spelling, we have to look to the distant past…
Or, you know, the fairly old past—the Old English-y one. Old English inherited from Indo-European (with a few twists and turns through Grimm’s Law) a sound we linguists like to call a voiceless labiovelar approximant1. In IPA, [ʍ]. That’s fancy language-people talk for a kind of voiceless W. In OE, this sound was spelled ‘hw’. ‘Which’ was ‘hwilc’ and ‘whale’ was ‘hwæl’. Perhaps the real poster child for this phenomenon is the first word of Beowulf: Hwæt! (ModE ‘what’)
During the Middle English period, the spelling of this sound was flipped to our modern ‘wh’, most likely due to the influence of French scribes who came to England with the Normans. It was also sometime during this period that some dialects began to see a merger between the pronunciation of ‘wh’ and plain old ‘w’. For a while, the merger was seen as uncouth, and educated speakers deliberately maintained the [ʍ] pronunciation of ‘wh’. Now, we find more dialects than not where the merger is complete and both spellings are pronounced [w]. But as mentioned before, there are several varieties of English where the original [ʍ] is hanging in there.
English has its share of strange, purely historic spellings, but this isn’t one of them. Your [ʍ] dropping friend isn’t mispronouncing ‘white’ or being pedantic; they’re just kicking it old school.
1This sound is sometimes traditionally/erroneously called a labiovelar fricative.
Historical linguistics is often synonymous with the study of language change over time, and investigating what the reasons for that change are; are the changes being triggered by processes internal to the language, or did they come about through influence from another language? We know that English has changed significantly in its history – I recommend going back to Rebekah’s post about the English periods for a recap. Exactly what mechanisms are behind some of these changes are still under debate, but we do know that English has been greatly influenced by the languages it has been in contact with throughout history and in particular the contact with Old Norse and French. (Read also Sabina’s post about the creolization hypothesis for more about these contact situations)
Looking at these two contact situations, they had quite different effects. This can partly be explained by the different relationships these language had with English during the time of contact. Let’s investigate:
Remember the Vikings and their language, which Sabina taught us about two weeks ago? This language first became introduced to Britain through Viking raids in the 8th-9th centuries. The very earliest evidence of influence from Old Norse is from this period, and it shows up through loan words which have to do with seafaring and similar themes. The Norsemen eventually started settling in the British Isles, however, and in England, this meant some drastic political changes: wars between the Danes and Anglo-Saxons led to the establishment of the Danelaw, an area covering most of the East and North-East of England, which was under Danish rule for some time (although, the power shifted between Anglo-Saxon and Danish for the duration of the Danelaw). The relationship between Old Norse and Anglo-Saxon English (Old English) thus changed in the Danelaw, in that Old Norse became the language of the rulers. In this period we see many Old Norse words to do with law and government entering English.
Another effect of this time is that speakers of Old Norse and Old English lived side-by-side and dealt with everyday communication. If you’ve paid attention during Sabina’s Early Germanic Dialects series, you will know that Old Norse and Old English are quite closely related languages, both being descendants of Proto-Germanic. In fact, these two languages had not been developing separately for very long before they came into contact in England again (only about 4-500 years or so, quite a short time in the grand scheme of things). This fact, along with some evidence in records from the time, suggests to us that speakers of these languages could understand each other, with help from some accommodation from each side. I imagine this to be something like when I, as a Swedish speaker, am talking to a Norwegian speaker and end up speaking something we jokingly call ‘Svorska’, a combination of Svenska (=Swedish) and Norska (=Norwegian).
It is through this close relationship between the languages in the Danelaw that we see some of the deeper effects of Old Norse influence in English. First of all, the type of loan words that enter English in this period is the type of words that are most reluctant to change or influence, such as the verb take, or the noun sister; words that are used so frequently and are so fundamental to the language that they rarely get replaced by innovations. There are also some grammatical changes that have been suggested to be triggered by contact with Old Norse, such as changes in word order and simplification of the case and gender system on nouns. While some suggest that these grammatical changes were directly influenced by Old Norse, others argue that the changes were already underway before the contact, but that the contact triggered them to happen quicker.
How does contact with another language affect the grammar, if it’s not a direct transfer? Well, this can partly be due to grammar becoming compromised when people accommodate to a closely related language – people taking “shortcuts” to make themselves understood, kind of. It can also be due to that children learning their first language become presented with a mixed-but-similar-enough input that the grammar they learn as a first language is slightly different from what came before.
This Danelaw way of speaking came to influence the other English dialects, and Scots as well, through large enough numbers of people bringing their dialect with them from the Danelaw to other parts of Britain, even after the Norman conquest (immigration from the North to London is of particular importance for the Northern dialect forms entering standard English).
So what about that Norman conquest?
Well, in short: the Norman army conquered England in 1066. English nobles were replaced by French ones, and Norman French as well as Parisian French became the language of rulers across England. This contact situation was in some respects similar to the situation in the Danelaw, in that French, like Old Norse, became a language of invading rulers, but the situations differ in one very important way: French never became the language of the people. The French-speaking nobility were always in a position of power, and often had to speak English in order for their subordinates to understand them. Furthermore, the French-speakers were much fewer in numbers than the Old Norse-speakers were. After French loses its influence over England, the nobility starts to shift to English altogether, and English successively regains its position as the national language of England and the process of standardisation begins.
What English gets from this period is a whole lot of vocabulary. We especially see a huge influx of French vocabulary items entering English when the nobles shifted from French to English.. These words are often relating to “higher” contexts, such as art, music, religion and government, but there are also everyday words, such as joy, entering the language as a result of this contact. The legal and “governmentary” words largely came to replace the similar words that were previously borrowed from Old Norse, but mostly the French vocabulary expanded the language so that there were more word choices (compare the French borrowing joyous to the Old Norse borrowing happy).
The influence was not only in form of individual word transfers, however. The sheer number of words that entered English caused some alteration to the grammar: for example, the French suffixes -ment and -able became productive suffixes, meaning they can now combine with any word stem, whether Germanic or French (or other), and the sounds /f/ and /v/ became “upgraded” from allophones to phonemes.
Most languages do not develop in a straight line from their origin to the present day, and English is certainly no exception – it is usually estimated that 70% of the English vocabulary is loan words! Not all of these are from Old Norse and French of course, but they certainly make out the largest chunks of the borrowed vocabulary.
Is English uniquely mixed, though? A lot of people would like to think so, but there are plenty of other languages in this melting-pot continent of Europe which have experienced intense contact during long periods of time – for example, 40% of the Swedish vocabulary is estimated to be from German. However, there is no doubt that English has been greatly affected by these conquests, which sets it apart at least from its Germanic sisters in terms of its vocabulary and grammar.
You may have heard the word eh being used before. Often, it’s found
at the end of sentences; for example, you might hear someone say ‘nice day,
eh?’. Usually, eh serves to mark a
question or initiate some kind of response from the listener, though it can
also be used to signal agreement or inclusiveness. We call these kinds of words
‘tag particles’ – they have no set meaning on their own but are often used for
a particular communicative function.
The tag particle eh has a long history, dating back in literature to the 1600s. It has been noted across a far-ranging spread of dialects and varieties, including Scottish English, Canadian English, Guernsey English and New Zealand English, suggesting a common British origin. In each variety it shares several semantic and social functions and it is frequently associated with national identity and vernacular use. However, over time these different varieties have also developed dialect-specific uses of eh. Today we’re going to focus particularly on the use of eh in New Zealand English, where it has the shortest but nonetheless very interesting history. But first, we cannot talk about eh without briefly mentioning its prominent role in Canadian English.
Eh has long been recognised as a typical feature of Canadian English,
and it is so prevalent and so well-known that it is often the subject of jokes
or caricatures of the Canadian accent. Already in the 1970s and 80s it was
being used in advertisements, indicating that this particle was becoming
widespread and nationally recognised.
Canadian eh has with time become associated with national identity, and this
has endowed it with the status of a purely Canadian feature, or ‘Canadianism’,
despite the fact that eh also plays
this role in a number of other accents. The Canadian variant is typically pronounced
as the short, front, mid-high vowel [e], and has a rising intonation. The main
function of eh is to mark informality
and inclusiveness, as well as seek agreement from the listener. Eh has been found to be widespread
across Canada geographically and socially, although it is more frequently used by
the lower classes, who tend to make more use of addressee-oriented devices in
general. Though it has several functions, Canadian eh is most commonly found in:
Opinions: ‘nice day, eh?’ Statements of fact: ‘it goes over there, eh’ Exclamations: ‘what a game, eh?’ and fixed expressions, such as: ‘I know, eh’ and ‘thanks, eh’.
It is also found in questions, requests for
repetition, insults, accusations, and narrative functions, although the
questioning and narrative function of eh
is often seen by speakers as uneducated, lower class, and rural.
To jump forward a few centuries to a more
recently developed English accent, eh
is commonly found in New Zealand English as well. New Zealand English (NZE) speakers
tend to prefer eh to other possible
tags, leading to its highly salient nature. As in Canadian English, eh is a well-recognised feature, and is
also showing signs of growing national awareness, exemplified in its use in a
nationwide advert promoting New Zealand’s national soft drink; L&P. This
soft drink is an iconic feature of New Zealand, originating and being produced there,
and it is partially named after the small town it was created in.
Notice that the spelling here is aye rather than eh. This is most likely because in NZE eh is realised as the diphthong [æe], as in ‘face’, with a slight palatal approximant gesture (meaning that
the vowel is followed by a slight ‘y’ sound), unlike Canadian eh which is realized as [e] in IPA. New
Zealand speakers generally pronounce eh
with a falling intonation, which distinguishes eh from most other varieties of English who typically have a rising
intonation, Canadian English included. Eh
most commonly occurs at the end of sentences, but is also likely to occur
mid-utterance, unlike in most other varieties. For example:
‘the phone will be non-stop eh with all the
girls ringing him up and stuff’
Eh performs a number of functions in New Zealand English and tends to
be used to a greater extent by working-class speakers and in informal contexts,
which overlaps with the patterning we find for Canadian English. The array of semantic
roles eh has acquired are both New
Zealand-specific and share significant overlap with the Canadian variant. In
New Zealand English its most common purpose is to signal, recheck or establish
common ground with the interlocutor, but eh
can also be used to checking the comprehension of information, confirm shared
background knowledge or seek reassurance of the listener’s continued attention.
However, question and answer sentences discourage eh, quite unlike the Canadian variant. This wide range of usage may
be partially due to the historical developments it has undergone since it
arrived on New Zealand’s shores.
where did this eh in New Zealand
English come from exactly?
Whilst we cannot know for sure with the
current information we have, it seems very likely that eh came from Scots, where it is still found today. Previously, the
general assumption was that New Zealand English was generally derived from the
English of South East England, but now we know that a surprising number of words
came from the north of Britain, particularly from Scots. The use of Scottish eh, or rather e (as it is commonly transcribed), is prevalent in some Scots varieties
such as Hawick Scots and also in Edinburgh. Just like New Zealand English, it
too has a falling intonation, although it is pronounced [e] rather than [æe]. E typically occurs with be and have, for example:
‘he had a stroke, e?’
There are a number of significant overlaps
between use of eh in NZE and use of e in Scots. E can be used to confirm shared background knowledge, which matches
its usage in NZE, where eh
acknowledges the shared understanding between speakers. For example:
‘we know him quite well by now, e?
Furthermore, both eh and e can also be used
as a positive politeness feature to make a statement, opinion, or request less
sharp and more polite. For example;
‘Put it down there, e’ ‘I like Sambuca, e’
However Scots e is also noticeable in question and answer sentences, unlike NZE. For
‘he’s coming, e?’ ‘he isnae coming, e?’
We can see here that Scots e performs a number of functions, some
of which have significant similarities with eh
in NZE, and some which differ. So, if NZE eh
possibly comes from e, how did it
get into the accent?
Scottish e contributed to the rise of eh
in New Zealand English through process of new
dialect formation. Historical dialect formation is (often) the result of a
number of different dialects being brought into close proximity with one
another in unique, isolated circumstances. Through various processes these form
a new dialect. These processes have been categorized into five distinct periods
by Peter Trudgill. Initially there is reduction
and accommodation between the
different dialects; the most dialectal features are discarded and ‘half-way’ features
are frequently chosen. The next two steps involve further levelling (so removing the strongest dialectal features) and modification through speaker convergence
(speakers adapt their speech to make themselves more comprehensible). During
this process one feature is chosen and becomes standardised; in this case it was eh rather than other tags that was chosen as the agreement marker.
The final components to dialect formation are focussing and adoption by
the wider community. These last steps are still ongoing today; use of eh is led by the youth in the NZE
One of the great things about the New
Zealand dialect is that we actually have recordings from the very first British
settlers setting foot on New Zealand soil, right up until present day NZE.
These recordings, stored in what is known as the ONZE (Origins of New Zealand
English) corpus (https://www.canterbury.ac.nz/nzilbb/research/onze/), have
allowed researchers to see (or rather hear) these processes of dialect
formation in action. In the corpus, we found that use of eh was significantly higher in the region of Otago, which historically
saw a high concentration of Scottish settlers. Unlike the rest of New Zealand,
the dialect from this local area has a number of Scottish-inspired features,
including Scots vocabulary items and rhoticity. Furthermore, speakers with Scottish
parents showed greater usage of eh, regardless
of where they had settled in New Zealand. Small numbers of e were in fact present in the first wave of recordings (1860-1900),
but this becomes gradually replaced by eh
after 1900. So here we can see the stages of dialect formation taking off; initially
e is present in the dialect, but with
reduction, accommodation, and levelling,eh was chosen and has become widely
adopted into everyday NZE during the last fifty years. However, this might not
be the whole story.
Whilst it seems likely that eh came into NZE from Scots and pre-colonial
varieties of English, the difference in pronunciation between the two is more
difficult to account for. However, there is some precedent for minority
language influence on New Zealand eh;
various studies have found that Maori speakers, particularly males, were the most
frequent users of eh. The particle eh is very similar both in pronunciation
and function to the Maori tag particle nē
(pronounced [næe]. It is possible that once eh
was adopted by Maori speakers if would have been influenced by nē to produce a form similar in phonetic
quality. The functions of eh also
appear to have expanded, again through influence from nē.
This change in turn possibly influenced young
Pakeha (non-Maori) speakers, who have shown increasing use of eh by from around 1940 onwards. This gives
us the particular ‘ay-ye’ pronunciation that is now in wide circulation, as
well as the new meanings associated with eh.
We can see this change happening shortly after increasing numbers of Maori were
migrating to the cities in search of work, bringing them into greater contact with
Pakeha speakers. The New Zealand Government also practiced a policy of ‘pepper
potting’- the scattering of individual Maori families among Pakeha neighbours,
in an effort to prevent the Maori community from clustering together in the
cities. This naturally brought the two speaker groups into closer contact with
one another, allowing for cross-dialectal influence.
So it appears that eh came initially from
Scots and influenced the New Zealand English dialect. It was chosen as the
invariant tag of choice, and was in use within the post-colonial population in
New Zealand. This tag was then adopted by Maori speakers acquiring English and
influenced by their own particular tag particle, nē. The pronunciation changed, as well the particular uses of eh. This new form of the variant was
then adopted by younger, Pakeha speakers, and is now spreading through the
society, led by the youth.
what about Canadian eh?
Again, there are similar possible links
between the Scots e and Canadian eh. In 1851-61 there were several waves
of British settlers to Canada, especially Scots and Irish immigrants as part of
a concerted effort by the British government to populate Canada. In 1901-11
another wave of British migrants settled in Canada, particularly Scottish. In
the unsettled areas of Ottawa Valley, the colonial lineage of Scottish and
Irish accents remains to this day and can still be heard in the speech of some
local speakers in the Ottawa basin.
So, it seems that eh could have spread via Scottish immigration during the colonial
period. It concurrently underwent linguistic changes through new dialect formation to produce the form
that has surfaced in several colonial countries over time. Both the New Zealand
and Canadian dialects have developed their own version of eh, but it seems that the roots of this particle in both dialects
stems from the same source; Scots. Pretty cool, eh?
I teach fifth-grade Latin, and recently we were discussing the pronunciation of the Latin digraph and diphthong <ae>. One of my bright young scholars asked if the Latin letter was written with “one of those connected a-e thingies.”
My Anglo-Saxonist heart soared. That “connected a-e thingy” is <æ>, a symbol called by the Anglo-Saxons aesc, like an ash tree. With the adoption of the Latin alphabet, <æ> inherited all the rights and responsibilities of ᚫ, a rune of the same name in the Old English fuþorc. It was pronounced [æ]1, like in, well, ‘ash’.
My Latin class and I had to plow ahead with the nominative plural, but in the back of my mind, I kept mulling it over: Where did my beloved aesc come from, and why isn’t it all over the Classic Latin texts I read?
As with so many questions linguistic, the answer lies in human laziness. Since man started putting pen to paper (stylus to papyrus, wax, clay, &c.2), we’ve been conjoining letters to cut corners and save time and space. Cursive is one thing, but typographic ligatures are little clumps of two or three letters written as a single symbol. An example of a well-known ligature that grew up to be a letter in its own right is <w>, which as the name implies, began life as a double <u>.
There are copious examples of ligatures dating all the way back to Sumerian, but we’re investigating <æ>, and for that we have to look to medieval scribes. It’s as simple as you might imagine: Whether for speed or aesthetics, medieval scribes took <a> and <e> and wrote them as one. In Latin, it made no nevermind whether you used the ligature or wrote <ae>.3 In fact, as Latin pronunciation changed throughout the Middle Ages, the spelling was sometimes reduced to merely <e>. (Thus, we modernly tend to write “medieval” rather than “mediæval”.)
Old English wasn’t the only language to promote this particular ligature to a letter. Today, it can still be found in languages like Icelandic and Norwegian.
In Modern English, aesc has been relegated to the status of relic. It gets trotted out when calligraphers and designers want to make something look fancy or antiquated, but otherwise, it’s just some letter that we used to know.4
1 It becomes fairly obvious where linguists found the symbol to represent this sound in IPA. 2 I would just like to share that the ampersand or “and sign” (&) began life as a ligature of <et>. “Et” is “and” in Latin. I can’t even. 3 As far as Classical Latin goes, the Romans themselves and modern editors use distinct <ae> much more often than not. 4 Alas for me! I suppose I’ll just have to stick to doodling aesc in various margins.