Historical Linguistics

Part of a series on Stuff You Might Not Know About

Language families of the world (click to enlarge)

Where have all your good words gone?
Where have all your stories gone?
Where are all the pleasures from
The timbre of your tongue?
—"Where Have All Your Good Words Gone?" Laura Gibson

Like many young Americans with more intelligence than sense, I studied linguistics in college, a discipline with few applications in the post-college world of having to get a job and be a real adult. Today's post is part of my ongoing efforts to justify my decision to spend four years studying this stuff by demonstrating how totally fascinating languages can be.

Historical Linguistics was a fun class. It was taught by a grad student who I'll call Ilya. Ilya was what many linguists wish they were but can never be: a true polyglot (a fluent speaker of multiple languages; note that most linguists know a lot about language but very few speak many languages). Ilya's parents made it a goal to teach him multiple languages as he grew up; they reserved certain days of the week for different ones: on Tuesdays, for example, his parents would only speak to each other or him in French, and would only respond to him if he did the same. Whether you view this parental strategy as ingenious or cruel, it certainly bore fruit. A typical linguistics instructor during my time in college, when confronted by a student with a question requiring an on-the-spot example from a real language, would use something either from English or the one obscure language they knew a lot about, a Native American language say, often struggling for a moment or two before answering. Ilya, though, would quickly pull examples from multiple languages, often only very distantly related; a typical response might include words from German, Greek, Albanian, the Celtic languages, and something long dead—Ilya's specialized in the languages of ancient Anatolia, like Hittite and Luwian, but he could just as easily give you examples from Latin and Old Church Slavonic, among others. 

Ilya was the perfect person to learn historical linguistics from, then, because it's a discipline that frequently requires the detailed comparison of multiple languages. Originally, all linguistics was historical linguistics; the discipline was started by curious folks in the 18th century who noticed similarities among the different languages of the ancient texts they were studying, languages like Greek, Latin, and Sanskrit. The early linguists developed tools of study that allowed them to link these languages and others together into a family, with a common "ancestor" language from which all the others evolved—in this case, the common ancestor of Greek, Latin, and Sanskrit (not to mention English and many others) is called Indo-European and its descendants are known as the Indo-European languages.

Today, historical linguistics is just one branch on the big linguistics tree, one that many linguists don't consider all that important. Historical linguistics is about studying and reconstructing what languages used to sound like, whereas the most urgent activity of modern linguistics is investigating the languages we have now, especially the many, many languages that are dying, before we lose the ability to study them. Still, the discipline is important to the history of linguistics as a discipline and as part of the foundational knowledge and tools of people who want to investigate the inner workings of language.

The central tool of historical linguistics is the comparative method. The comparative method is a way of reconstructing what the words and event the individual sounds of an ancestor language were, by comparing the words of the "daughter" languages and applying some universal rules of language change. Let's take a look at some data!

(For these data, I'm going to spell things with some phonetic English spelling; importantly, "ch," "sh," and "th" will all sound like they do in English. If you're a speaker of any of these languages, I apologize. I know they don't usually spell words this way.)

	Sardinian	Italian	Romansh	French	Spanish
"100"	kentu	chento	chyent	sa	thyen
"sky"	kelu	chelo	chil	syel	thyelo
"stag"	kerbu	chervo	cherf	ser	thyerbo
"wax"	kera	chera	chaira	sir	thera

We can look at words across these five languages and notice that the words we've selected have a lot of similarities, but there's a big key difference: within a language, each of these words starts with the same sound, but across languages they start with different sounds. A historical linguist sees this, and guesses that these languages are probably related, that they share a common ancestor. The historical linguist can then ask: what sound was at the beginning of these words in the original language?

The best guess comes from a knowledge of what kinds of change tend to take place in languages. There is a language change process called "lenition" (i.e., "weakening") which takes place all the time in language. Lenition is when sounds that have more obstruction of air flow in the mouth ('strong" sounds) become sounds with less obstruction ("weak" sounds); sounds like "k" where the air is stopped completely can become sounds like "ch" where the air is stopped briefly and then allowed to flow through a constricted passage, and "ch" can become sounds like "th" and "s," where the air is never stopped at all but simply passes through a constricted space. Given that this process is extremely frequent in the world's languages, it makes for a good guess as to what happened here. The original ancestor of these languages probably had "k" at the beginning of these words; in Sardinian, the sound never changed. In the others, it probably first changed to "ch," and in French and Spanish, it changed further to "s" and "th," respectively.

In many cases, this is where historical linguistics has to stop: a hypothesis about what an ancestor language sounded like. In this case, we can go a step further, because we have information about the ancestor language: these are all Romance languages, descended from Latin, a language well attested in written records.

Wikipedia has the coolest maps, guys. (Check here for giant, more readable version.)

Scholars have long known the ancestor words for these terms; they're centum, "100;" caelum, "sky;" cervus, "stag;" and cera, "wax." We also know for reasons independent of these data that the first sound of each was originally a "k" sound, just like we predicted from the data.

With these kinds of tools, historical linguists have done pretty fascinating things. They've been able to help chart the prehistory of the peoples of the world, not only reconstructing what the languages of our ancestors sounded like, but even using vocabulary analysis and other tools to make strong guesses about where certain cultures and ethnicities originated and track their migration across the ages (here's a book featuring one example). Hand in hand with archaeology and other disciplines, historical linguistics can teach us about ourselves by connecting us to our human past, back to times before written language even existed. I think this is pretty exciting.

Historical linguistics is really, fundamentally neat. If you're interested in knowing more, you can check out Historical Linguistics by R.L. Trask and Historical Linguistics, an Introduction by Lyle Campbell. Both are excellent introductions and relatively accessible to the lay reader.

More generally, if you're interested in getting an introduction to linguistics as a whole, you could start with the book that got me hooked, Language: The Basics (warning: do not give to impressionable undergraduates whose major is undecided), or you could take a look at the excellent Language Myths for a discussion and debunking of commonly held beliefs about language.

---

Image Sources:
http://en.wikipedia.org/wiki/File:Primary_Human_Language_Families_Map.png
http://en.wikipedia.org/wiki/File:Indo-European_branches_map.png
https://upload.wikimedia.org/wikipedia/commons/3/31/Romance_20c_en.png