– THE LINGUIST’S CORNER –
By Dr. Jon Aske
As we have seen before in this column, languages are in a constant state of change. The change is almost always slow and imperceptible, but it is change nonetheless and it adds up after a while. Today I want to show you some salient examples of change in the sound system of English, and how this is evidenced in our erratic spelling. In many cases the spelling of English words was fixed shortly after the printing press was invented half a millennium ago and changes in pronunciation that took place afterwards are not reflected in the spelling. That, of course, is a major reason for our silent letters and crazy spelling conventions.
The history of English is usually divided into three periods. Old English is the period between the time the Angles and the Saxons, two Germanic tribes from what is now the coast of Denmark and the Netherlands, invaded what is now England when the Romans abandoned it after the Roman Empire’s collapse 1500 years ago. This period lasted until the 11th century, when England was invaded by Normans (a word derived from North man or Norseman), another Germanic people also originally from Scandinavia, but which had settled in Normandy (in the north of what is now France) starting in the late 9th century and had adopted a version of French. (By the way, the Normans, also known as Vikings, a restless people, also conquered and settled southern Italy and Sicily around the same time.) The period starting then is known as the Middle English period, which lasted until the 16th century. The period from the 16th century on is known as the Modern English period. Keep in mind that these periods are somewhat arbitrary and that language change was gradual all along.
Let’s start with an example of sound change from the early period. Have you noticed that the i vowel in the words child and children (or Christ on the one hand and Christmas or Christian on the other) are different? We call them long i and short i now, respectively. Until the 7th century the vowels were the same, but the i was shortened in certain contexts, such as when followed by 3 consonants. The short vowels after the change was pretty much like it is now, but the long one was just a bit longer in duration, not like it is now. Later changes made the long i in child and Christ different, as we will see below. Other pairs of words that evidence a shortening pattern in the same context are please (long) vs. pleasant (short) and sane (long) vs. sanity (short).
Let us look at consonants now. Do you know the difference between a shirt and a skirt? Well, I’m sure you do, but at one time there didn’t use to be any. They both referred to the same article of clothing, some kind of a long shirt. The two words come from the same original, which used to be pronounced more like skirt (in Icelandic it was skyrta, in Swedish skjorta, and in Danish and Norwegian skjorte). The word shirt and others like it which have the letters “sh” in them, originally (in the language that English comes from, Proto-Germanic) had the sounds [sk] in them instead (I will use quotation marks to indicate letters and square brackets to indicate sounds). Eventually the sound [k] in the [sk] sound sequence mutated into the sound [h], sounding something like the “sh” in mishap. This may seem like an odd change but actually the [h] sound, as it was pronounced then (more like the “j” sound in Spanish is now) is produced in a very similar fashion in the mouth. Eventually the two sounds [s+h] merged into the single sound that is nowadays represented by that combination of letters “sh”. (In French that sound is written “ch”, as in chapeau; in Spanish the sound does not exist and Spanish speakers learning English replace it with the sound written “ch”, which is why the word shampoo became champú when Spanish borrowed it from English.)
So if the word shirt comes from the word skirt, then why do we still have the word skirt in English? That’s because the word skirt came back into English at a later date from a sister language, probably Old Norse, brought by the Vikings to England after the sound change had already taken place in England. So, indeed, shirt and skirt were originally one and the same word.
And, of course, this sound change ([k] > [h]) didn’t just happen in the word skirt, but in any word that had a similar combination of sounds. So perhaps now you can tell me what the relationship is between a ship and its skipper. That’s right, a skipper is nothing but a “shipper”, a person that runs a ship. But why do we say skipper and not “shipper”? Because the word skipper was borrowed into English later on in the 14th century from yet another sister (Germanic) language, namely Dutch, where the sound change sk > sh had not taken place.
Let’s move now to the topic of silent letters, which you may have wondered about before. Where exactly do they come from? Although silent letters are silent now, they were indeed pronounced at one time. One example is found in the “sw” combination of letters (which used to be pronounced /sw/), which before some vowels (“back vowels”) become /s/, as in sword. And /mb/ becomes /m/ (the /b/ is lost) as in lamb and comb. These two changes happened between 900 and 1400AD in England.
A rather dramatic sound loss was the one which was written with the combination of letters “gh” in words like night, taught, and caught. The sound that these letters used to represent is similar to the /h/ sound in the English word hill. Actually, as we saw earlier, it was more like the sound of the letter “j” in Spanish José as pronounced in most dialects (in Caribbean Spanish and the Spanish of Southern Spain, the “j” sound is indistinguishable from the sound of the English letter “h” in the word hill).
By far the most disruptive change in the English sound system took place in the vowel system in the 15th century. The phenomenon is known as the Great Vowel Shift (GVS). As we saw, English distinguished between long and short vowels before then, but at this point, the pronunciation of all long vowels changed drastically. They were all raised (the tongue position became higher) or diphthongized (pronounced as diphthongs). For example, the word mouse was pronounced at the time as we now pronounced the word moose, with the sound of the vowel “u” in Spanish or “ou” in French. This pronunciation applied to all words with those letters, such as house and stout. Of course, before the GVS the word moose, which also contains a long vowel, was not pronounced as it is now, but more like the word most is pronounced now if you omit the “t” (or as a Spanish speaker would pronounce the word moss).
As we saw earlier, ever since the 7th century the i in child was longer than the i in children. The two vowels sounded much the same but the first one was just a bit longer lasting. Think of the difference between the vowel in leaf and the vowel in leave, which is the same but shorter in the first case than in the second. Long i became the diphthong it is today due to the GVS. Think of it this way: Until 500 years ago the word I was pronounced “ee” (that’s how the letter i is pronounced in Spanish, Italian or French).
One should keep in mind that there were sometimes exceptions to these changes and that in some cases there were later changes in the spelling of some words which hide the regularity of these changes. Also, words that entered the English language after this period didn’t undergo the change, so the i in police (new word) is not the same as in the word polite (after the shift). By the way, how do we know if a vowel was long or short? Sometimes the spelling was the same (as in child vs. children). Sometimes the letter representing the long vowel was doubled, as in meet. Sometimes a silent e in the following syllable was a clue that it was long as in like or home (an old trick English speaking children still learn in school).
Let us quickly go through the GVS changes. As we saw, a long i in words like child or my used to sound like the ee in modern English meet, but it came to be pronounced as in modern English after the shift. Long close e as in the word me or meet, used to be pronounced as the vowel in modern English mate (in other words, like Spanish e), but came to be pronounced as it is now, that is, the way long i used to be pronounced earlier. Long open e, as in the words beg or red, which used to sound more like modern bag or rad, changed their pronunciation as well. The long a in mate used to sound like the a in modern day father, but you know that it doesn’t anymore, but like the “ee” used to sound. We already saw what happened to the long u sound in mouse and house. The long close o in words like fool or boot, which sounded like modern boat, changed too. Finally the long open o in words like boat, which used to sound more like the word bought does now in dialects in which bought is pronounced differently from bot, came to have the pronunciation that boot used to have. In other words, vowel musical chairs.
Lots of other sound changes took place in English in the 1400-1600 period. The initial g in initial gn clusters was lost in words like gnome, gnat, and Gnostic. Similarly, the initial k in the initial kn cluster was lost, as in knot and know. The word knot thus became homophonous (sound the same) with the word not. The w in wr clusters was lost at this time too, and thus pairs of words such as wrap and rap became homophonous. Also, generally all double consonants became pronounced as single consonants, as in apple and common.
The next time period we are going to consider is the 17th century, up to the time when British and American English split. The sound of the letter “r” became greatly changed during this time. English “r” used to be pronounced for the most part as Spanish rolled “r” (as in rosa “rose” or carro “car” ) or as tapped “r” (as in Spanish mar “sea” or caro “expensive” ). But during this time, the sound changed to the current one. The new “r” sound had appeared first in English when the r was before other consonants, in words such as charm, only to later change before vowels too, in words such as rose. In Scottish English we still hear the tapped “r” sound (in stereotyped Scottish English pronunciation it sounds more like the rolled “r”).
Have you noticed that the vowel in the word put is different from the vowel in the word cut? This was not the case until this time. In both words the vowel used to sound like the vowel in put, but the vowel in words like cut changed to the new pronunciation pretty much whenever the vowel was not followed by a labial consonant, such as p or b. This change did not happen in northern England, which is why to this day people from northern England who speak “dialect” pronounce words like cut differently (the original way, which in this case is not the “correct” way).
Have you noticed that we do not pronounce the letter g in words like sing and singer, although we do pronounce it in the word finger? This is not obvious to most people, so you may need to do some actual listening to the words pronounced by another person. This is a change that happened around this time too. This change first happened in sing (that is, at the end of a word) and spread to the derived word singer by analogy. Since we don’t have a word fing related to finger, it did not happened in finger. Notice that it is not that the g became silent. The ng sound in the word sing is indeed different from the n sound in the word sin. Think about it.
In most dialects of modern English the words meat and meet are homonyms today. This didn’t use to be the case until this period. What was spelled “ea” had a sound that was more like the vowel in mate today, which was the sound that meet had had before the GVS. I know, it’s enough to drive you crazy. Think how foreigners learning English must feel. The same thing is true of words such as team and cream, of course. As we have said, sound changes are quite general and not tied to specific words.
Until this time starry used to rhyme with marry. It still does in Scottish English. This is consequence of a change that took place in the word star, where the vowel was lengthened due to the r being word final. The change in the pronunciation then spread to the derived word starry. This change did not take place in marry since this word is not derived from a word mar (which doesn’t exist).
The letter “a” can be pronounced in many different ways in English, viz. male, man, talk. The sound of the a in male stems from the GVS mentioned earlier. The sound of the a in talk, however, comes from this period and it reflects a change that took place before “l” in words such as talk, bald, and salt (but not if the l was followed by m, as in calm).
Other sound changes produced sound mergers, so that words that used to be pronounced differently became homophones. This is the case of what’s known as the pane-pain merger and the toe-tow merger. Unlike now, these pairs of words had different vowel sounds until one (or both) of the vowels in the pair changed.
Another 17th century change occurred with the vowels in foot and food. These vowels are different nowadays (had you noticed?) but this wasn’t the case until this time. Until then the vowel used in foot was the same as the one in food and it sounded like the vowel in food sounds now (which is how someone learning English might be tempted to pronounce it at first). This change took place in some words (foot, book, good,), but not in others. Some dialects in England are still resisting this change and pronounce foot with the vowel in food and thus the two words rhyme in those dialects.
The next period that we are going to consider reflects the split between British and American English and it is the period of the 18th and 19th centuries, approximately. The loss of the pronunciation of the consonant “r” in syllable-final position (as in car, hard, and often also in water) dates from this period. This is a standard feature of English English (English from England), but it is also found in parts of the US, such as Boston or New York English. Dialects where the r is not pronounced are called non-rhotic. This loss started as early as the 15th century in parts of England; however it didn’t become widespread until the 18th century when it became standard pronunciation of the London dialect, historically England’s most prestigious dialect. In Western England the r is still pronounced by speakers of local dialects. Parts of the US that had greatest contact with London, such as Boston and New York, adopted this change, but areas with large immigrant populations from Western England did not.
In most varieties of North American English the first vowel in the word father and in the word bother are the same. That was not always the case and it is still not the case in other dialects of English. An exception to this vowel merger can be found in varieties of English in Eastern New England (“Boston accent”) and New York-New Jersey English. How do you pronounce them?
We do not only find vowel mergers in English. We also find vowel splits. For instance, the vowels in words like bath and trap used to be the same in English. They still are in many places, but in parts of southern England, including standard or BBC English, the vowel in words like bath and pass often changed their pronunciation. This happened when the following consonant was -s, -f or -th. So in Standard British English pass does not rhyme with mass (an exception), but it does in most dialects of American English, as well as in Scottish English and the English of northern England. Curiously in parts of New England, as in the typical Boston accent, the split did take place and so the vowel in the words bath and half for many people in this area is what’s called the Broad A, very similar a to the southern England version. Fewer words in these accents have the Broad A than in British English and fewer speakers use them too. The word aunt is a word that is generally pronounced with a Broad A in New England (in other parts of the country aunt sounds just like ant).
The words whine and wine (and witch and which) are now homonyms in most varieties of English, including prestigious varieties, but this is a new development from this period. In parts of the US the “h” in one of the members of this pair can still be heard, actually sounding before, not after, the “w”, as the spelling might lead us to think.
In most American English dialects the words writer and rider are homonyms too. As you could have guessed this was not always the case, as it is still not the case in Standard British English and other varieties of English. In American (and Australian) English, a t or a d sound between two vowels, the first one of which is stressed, changes to a “flap” sound, which happens to be identical to the sound of Spanish flap r in a word like caro “expensive”. That the stress matters is nicely exemplified by the fact that it happens in motive but not in motif. This sound change also doesn’t happen when the t is followed by what is known as a syllabic n, as in the word cotton.
Mergers are not a thing of the past. They are still going on in English. In the US South what’s called the pin-pen merger is taking place right before our eyes. The vowels in these two words (and others like them) are merging before nasal consonants, such as n and m (but not before other consonants, so that pit and pet do not sound the same). The resulting vowel is closer to the vowel in pin than the vowel in pen in other dialects. You may have noticed if you have travelled in the Southern states. But not only the South. African American English also displays this merger, as do many speakers of the Midland region and, interestingly, the city of Bakersfield in California. If you want to hear about an ongoing vowel shift in American English, listen to this NPR interview with the great late linguist William Labov on NPR.
To end this trip through the sound changes of English let us discuss the letter “h”, the topic of an earlier column I wrote here dealing with Spanish. As you may have noticed the English letter “h” is pronounced in some words, such as horn, hide or hose, but not in others, as in honor or honest. Basically an h is not pronounced if it was not pronounced in the language where the word was borrowed from, which in the case of English means typically medieval French (in words borrowed from book Latin, such as hospital, the h was not pronounced in classical times, but it is in English, since the word was not learned by word of mouth). English borrowed massively from Norman French 800-1000 years ago. Interestingly, if you have been to England, you may have noticed that many speakers of non-standard English there, such as London Cockney, do not pronounce their aitches ever. They say ‘e for he, and ‘ome for home. That is also a change that is taking place in the language there. This change has not become part of the standard as so many other changes have, but it could some day. As may have become clear by now, whether a pronunciation is considered beautiful or ugly, correct or incorrect, in English as in any other language, is pretty much an accident of history. Change is pervasive in all languages. Some changes, those adopted by the powerful elites, tend to become accepted as “correct” and “beautiful”. Others, those who have not yet been adopted by those same elites, will be considered uncouth and ugly, for the same reason, even if they represent the older, original pronunciation. Nothing new here. So if you feel bad about your “Boston” accent where r’s are dropped in words like Harvard, just remember that that is the prestige pronunciation in England, where r-less pronunciation rules.
I hope you have enjoyed this short trip though the history of English pronunciation and will now understand the reason for our crazy spelling system, which still reflects old-time pronunciations or pronunciations of other dialects.