Language Change and Historical Linguistics: Crash Course Linguistics #13
Hi, I'm Taylor and welcome to Crash Course Linguistics!
Language change is a constant.
It's why I don't speak the same way as Shakespeare and he didn't write the same way Chaucer did,
and he didn't write the Canterbury Tales in exactly the same language as Beowulf.
Each generation takes a language and makes it their own, pushing it in new directions.
Today's small changes can be tomorrow's new dialects,
and those dialects may one day be so different we'll think of them as distinct languages.
While no one can predict the future of a language,
we can look backward in time and piece together the path that a language has taken.
This is a field known as historical linguistics.
To understand the historical path a language has taken,
we need to know how language change happens in the first place.
Language change occurs across all levels of language, including sounds, words, and grammar.
At the end of the Old English period in the 1100s, English speakers dropped a lot of suffixes,
changing the language's grammar.
Then, between 1400 and 1600, English went through a major change in the pronunciation of vowels, known as the Great English Vowel Shift.
So while languages can change in multiple ways, they don't always change in all areas at once.
As anyone who has misheard a song lyric knows, sometimes we invent words that were never there in the first place.
Many newer speakers of a language can mistake one word for something else, or reanalyze the language.
For example, English ‘apron' used to be ‘napron', but people heard ‘a napron' as ‘an apron'.
Now, apron is the standard English word.
Sometimes innovation happens because people need the language to work better for them.
Take English third-person pronouns.
In Old English, they were "hē," "hēo," "hit," and "hīe."
But these words sound pretty similar, so speakers started to differentiate them in Middle English.
"Hēo" changed to "she", "hit" to 'it" and "hīe" became an entirely new pronoun borrowed from Old Norse, "they".
English speakers started using "they" with nonspecific, singular referents like "someone" or "anyone" in the fourteenth century.
It's something both Chaucer and Shakespeare did.
The use of singular "they" for a specific person is more recent.
In the eighteenth century, singular "they" was used to anonymize a person,
and in recent decades it's become the pronoun for some trans, nonbinary, and genderqueer people.
Although there have always been people who don't identify with a binary gender,
to get into all of the ways they've expressed that through language would be a whole other video!
However, not all changes have a meaningful explanation.
Some varieties of Spanish spoken in Spain still maintain a difference between casa 'house' and caza 'hunt.'
Meanwhile, other Spanish varieties, especially those in South America, pronounce both words with an /s/ as casa.
There's no real reason for this change, though.
Language change is typically incremental, but sometimes there's a situation that is perfect for the creation of an entirely new language.
Let's head to the Thought Bubble to look at one example.
In Nicaragua, before the 1970s, there wasn't a Deaf community.
There were deaf people, sure, but because of a lack of structural support they stayed at home in their villages, and didn't share a common language.
Instead, they used various ad hoc gestures to communicate with their hearing family and friends, known as home sign.
This changed in 1977, when Nicaragua established the first school for the deaf and brought together deaf children of a wide age range from villages all over the country.
The curriculum of the school didn't try to teach the kids a signed language,
instead focusing on spoken Spanish, lipreading, and some fingerspelling of Spanish words.
But the students were signing with each other anyway,
and they wove together bits of their individual home signs into a richer, more complicated system, something closer to a language.
What's interesting is what happened after that.
The next generation of deaf children who arrived at the school were exposed to this intermediate communication system and took it a step further.
They made it into a full-fledged language, one that had a complex and systematic grammar like any other language.
The older generation of children had used pointing sometimes,
but the younger generation used pointing more systematically to track repeated mentions of the same person.
Pointing acted like a pronoun, according to linguist Ann Senghas.
Linguists around the world were fascinated.
Here was an opportunity to study how a language could evolve from scratch in real time!
It seems to take exposure to something linguistic enough,
like the intermediate combination of homesigns, at a young age and as part of a community.
Children's brains will sort out that input into language.
In this case, the students created what's now known as ISN, Idioma de Señas de Nicaragua,
the official Nicaraguan Sign Language.
Thanks, Thought Bubble!
Nicaraguan Sign Language is a unique example.
A more common, observable type of language change occurs because of language contact,
which is when languages in the same area influence each other.
It's very common for languages to come in contact with other languages.
In fact, multilingualism has been the norm in human society.
When languages are in contact, there are broadly two possible ways they can influence each other.
First, languages can become more like each other, that is, they can converge.
Click consonants are a typical feature of languages of the Khoesan family in southern Africa.
There are a number of Bantu languages in southwestern Zambia that also have clicks,
even though this isn't a typical feature in Bantu languages.
The clicks in these Zambian Bantu languages are a result of long term contact between Khoesan and Bantu languages in Zambia.
Or second, languages can diverge and become less similar.
On the northernmost island of the Pacific nation of Vanuatu there are 17 different languages.
These languages are all part of the Oceanic language family and have lots in common in terms of their grammar, but have all diverged in their vocabulary.
In this context, language contact has made speakers willingly diverge as they change words to be more distinct from each other.
If two or more languages aren't understandable to each other, we say they're mutually unintelligible.
When you put speakers of mutually unintelligible languages together,
they'll use the elements of their languages to create some basic way of communicating, generally known as a pidgin.
If the pidgin is spoken for long enough in a community where children are learning it,
they'll use the same skills as the Nicaraguan deaf children did to expand it into a language.
These languages are known as creoles.
There are many across the world, often with names that reflect these origins,
such as Hatian Creole, Kriol in Australia, and Tok Pisin in Papua New Guinea, which means ‘talk pidgin'.
When talking about creole languages, we should note the power asymmetry between the people and languages involved.
The language contact was usually the result of Western colonization,
during which colonists exploited peoples who spoke other mother tongues.
This power imbalance explains why a creole language is often made up of more words from the dominant language.
But creoles are languages in their own right.
Just like French isn't "bad Latin" for being descended from it.
And many creole languages have gained recognition and support.
For example, Tok Pisin and Haitian Creole are both official national languages of their respective countries.
When it comes to studying language change, we've been lucky enough to have audio recordings of some major languages and their changes for over a century now!
But for anything older than that, we have to rely on the writing system for evidence,
and hope that it reflects major features of how the language is spoken.
The good news is, for a language like Tibetan, or English for that matter,
there are over a thousand years of written examples that show us the path the language took.
Research that looks at the change that happens to a language over time with this kind of evidence is called diachronic analysis.
Meanwhile, studying variation across dialects or languages at the same point in time is known as synchronic analysis.
English has ‘father', Dutch ‘vader,' Icelandic ‘faðir,' and German ‘vater'.
A couple of centuries ago, linguists using synchronic analysis realised that similarities like this were too much to be a coincidence.
Words across different languages that share a common origin, like the different words for “father,” are known as cognates.
Historical linguists now know that languages like Dutch, English, and German have lots of easy-to-see cognates because their paths diverged only in the last couple thousand years.
To prove a relationship between languages, historical linguists make giant spreadsheets of words that might be related and look for systematic patterns across the whole.
We can reconstruct what the common ancestor word could have sounded like.
Something like *fadēr.
Because it's the oldest version of a Germanic language we can reconstruct,
it's known as Proto-Germanic.
By the way, historical linguists use that asterisk to signal that it's still a hypothesis.
We can go back even further, analyzing Proto-Germanic *fadēr, Latin ‘pater', and Sanskrit ‘pitr.'
Sanskrit is the classical language that gave rise to modern languages like Hindi and Nepali.
While the Germanic languages all have /f/ or /v/ at the start,
these more historically distant languages start with the /p/ sound.
Since the oldest languages start with /p/, historical linguists hypothesize that “father”
must have begun with /p/ and later changed to /f/ in the newer Germanic languages.
We see this in other cognates too, like Latin ‘pod' and English ‘foot.'
Or Latin ‘planus' for “something flat,” and English “field.”
These and hundreds of other cognates prove that the /p/-/f/ swap is a pattern, not a fluke,
so there must be a common ancestor among these languages.
If there were only a few similar words, it might be a case of one language borrowing from another, or simply a coincidence.
This process of methodically piecing together a probable common ancestor language from the existing records is called comparative reconstruction.
But historical linguists don't stop there.
We need to find further similarities among other sounds to establish a wider pattern.
Linguists then combine this with information about the grammar and evidence from other fields like history, archaeology,
and even genetics, to conclude that the languages themselves are related.
Comparing unrelated languages is interesting too, because it can tell us things about language in general, but this is referred to as language typology rather than historical linguistics.
It takes thousands and thousands of cells in a table containing words like "father" and "pater” and “vater" to reconstruct a hypothesized ancestor language, called a proto-language.
Centuries of dedicated and meticulous work to find cognates and understand language change backwards has led to the reconstruction of the Proto-Indo-European language.
It was probably spoken around six thousand years ago by a group of nomadic pastoralists somewhere east of Europe,
and evolved into many of the languages of Europe and the Indian subcontinent.
However, everything we know about their language and culture is only an educated guess.
We can only guess what it would've sounded like, because it was before there were written records,
and it's not like audio recordings existed thousands of years ago!
Other large language families that have been reconstructed around the world include:
Proto-Semitic, the ancestor language of Semitic languages like Arabic, Amharic, and Hebrew;
Proto-Algonquian, the ancestor language of Algonquian languages like Cree, Ojibwe, and Massachusett;
Proto-Austronesian, the ancestor language of Austronesian languages like Javanese, Tagalog and Malagasy;
Proto-Pama-Nyungan, the ancestor language of Pama-Nyungan languages like Yolŋu, Kaurna and Dharug;
and Proto-Bantu, the ancestor language of Bantu languages including Swahili, Zulu, and Shona.
Sometimes, even after years of careful study, a language doesn't have any evidence of being related to any other language, near or far.
These languages are known as isolates, and examples include Basque in Spain and France, Ainu in Japan, and Korean.
They might have originated independently, like Nicaraguan Sign Language,
or we may have lost all record of their relatives.
Regardless, at some point we can't go back any further.
Just like the real horizon, the path a language has taken is no longer visible to us.
But there are fewer than 100 known isolates, and many of the world's 7000 languages can be grouped into larger families using methods like comparative reconstruction.
In the next video, we'll learn more about how there came to be so many languages around the world.
Thanks for watching this episode of Crash Course Linguistics.
If you want to help keep all Crash Course free for everybody, forever, you can join our community on Patreon.