The First Words

I’ve been on the road for most of this week, but I want to post at least a short note on a report that adds a bit more evidence to last week’s post that wondered if the earliest languages were much more stable than current ones. While traveling I read Mark Pagel’s  paper in  Nature Reviews: Genetics (“Human language as a culturally transmitted replicator,” abstract here).  He discusses the most stable words in languages and not surprisingly they are also the most basic words.  Qualities that make them stable include: frequency of use, no need to borrow an alternate term from another language, and their references are so basic there is no need to revise them when cultural shifts occur. Thus, even without the arguments discussed in last week’s post we should expect the early language to be more stable than languages today. 

Also valuable in the article is reference to what many of the first 200 words might have been. Morris Swadish compiled in the 1950s. I’ve heard vaguely of the list, always with a dubious tone of voice, but Pagel persuades me that it is a valuable heuristic device. In plain English, it provides a rule of thumb that can get people started thinking about a particular language. And now that I have read through the list there is plenty of food for thought. One thing that seems likely is that most of the first 200 words ever used are indeed on it. Snow might have come later, but and, at, and there… how long can any speaker get along without them? The full list:

all, and, animal, ashes, at.

back, bad, bark of a tree, because, belly, big, bird, to bite, black, blood, to blow (wind), bone, to breathe, to burn (as: it burned up; NOT he burned it up).

child, cloud, cold weather, to come, to count, to cut.

day (as a phenomenon, not as a measure), to die, to dig, dirty, dog, to drink, dry (as a state, not as a verb), dull (blade, not a situation), dust.

ear, earth (soil, not the planet), to eat, egg, eye.

to fall (to drop), far, fat (thing, not person), father, to fear, feather, few, to fight, fire, fish, five, to float, to flow, flower, to fly, fog, foot, four, to freeze, fruit.

to give, good, grass, green, guts.

hair, hand, he, head, to hear, heart, heavy, here, to hit, to hold (by hand, not delay), to hunt (game), husband.

I, ice, if, in.

to kill, to know (facts, not locations).

lake, to laugh, leaf, left (hand), leg, to lie (on ground, not prevaricate), to live, liver, long, louse.

man (a male), many, meat (flesh), moon, mother, mountain, mouth.

name, narrow, near, neck, new, night, nose, not.

old, one, other.

person, to play, to pull, to push.

to rain, red, right (correct), right (hand), river, road, root, rope, rotten (log), rub.

salt, sand, to say, scratch (itch), sea, to see, seed, to sew, sharp (blade), short, to sing, to sit, skin (human), sky, to sleep, small, to smell (sense an odor, not to stink), smoke, smooth, snake, snow, some, to spit, to split, to squeeze, to stab, to stand, star, stick (of wood), stone, straight, to such, sun, to swell, to swim.

tail, that, there, they, thick, think, to think, this, thou (you singular), three, to throw, to tie, tongue, toot (front, incisor), tree, to turn (veer, not roll over), two.

to vomit.

to walk, warm (weather), to wash, water, we, wet, what, when, where, white, who, wide, wife, wind (breeze), wing, wipe, with, woman, woods, worm.

ye (you plural, year, yellow.

Comments

[Jesús Sanchis]

I agreee with you when you say: "we should expect the early language to be more stable than languages today". This is a fact which is often overlooked by many linguists, who try to establish "change rates" for prehistoric languages based on more recent phenomena, e.g, the increase in population and cultural contact due to a series of relevant technological innovations in the last few millennia.

As far as I know, the Swadesh lists are not very often used today as a means of obtaining significant inferences about the history of languages. The principle behind these lists is quite obvious, and any linguist would more or less apply some of these concepts, as they are just common sense.

I've heard about Pagel's research, and I think some part of it looks quite unrealistic. A few months ago I read a series of posts in Language Log (for example this one: http://languagelog.ldc.upenn.edu/nll/?p=1186), where some of his theories were strongly criticised, to say the least. Maybe the media impact of some of his theories somehow distorted them, but in any case I think that the type of research done by Pagel, aimed at finding 'rates' for language change, is not exactly the best way to try to understand the history of languages. The following sentence can be read in the Reading Evolutionary Biology Group web-page (see here: http://www.evolution.reading.ac.uk/LingCultEvo.html): "We are working on methods for inferring phylogenetic trees of languages and using those trees to measure rates of word evolution over time". It is clear that Mark Pagel, and probably some other people involved in this project, are not linguists. Like Noam Chomsky.

[María]

In the list there is plenty of food for thought, I agree. Thus I focus on the verb 'to say'. This verb is not at all necessary to speak, to say things about the world. It only became necessary (and only came into existence, I would suggest) when the attempt was made to tell what someone else had said. Before this, there was no verb ‘to say’, but, naturally, this did not prevent many things really being said.

[Michael]

I can't recall it exactly, but I think I've read somewhere that some word classes are more prone to change than others, even if they are just as frequently used as others. So that may account for the fact that some words that we think of as crucial might be missing in this list.

It is also interesting to compare these lists with the Natural semantic metalanguage project by Anna Wierzbicka, Cliff Godard and others. They compiled a list of universal human concepts found in all known human languages, and it overlaps to a large extent with the list given here. (see e.g. the wiki page on the topic or this article by Goddard:
http://www.degruyter.de/files/pdf/9783110188745Introduction.pdf )

As Maria already noted, it's interesting that you find verbs like to say, and to think on the list, and other mental verbs. These can also be found on the Natural Semantic Metalanguage list.

[Ornithophobe]

Prehistoric languages should have experienced faster rates of change than modern ones, simply owing to the widespread nature of modern literacy. Writing retards linguistic evolution. I find the idea of "survivals" in the lexicon utterly tantalizing, and I personally believe that they do exist, and that correspondences between superfamilies can be located. But I think the Swadesh lists were overly optimistic in the number of words that are prone to survive, and I have questions about just how much "linguistic integrity", for want of a better term, each word must possess in order to be deemed a survival. I've seen too many of the purported cognates that don't seem to have much to do with one another.

[Miles]

Michael makes a good point, but there might be two counterweights at work for and against language change. One is the fact that when language is not written down, if some people begin pronouncing words differently then after a few generations no one will remember how the words used to be pronounced and then change will take place, eventually making new words from old. Meanings change the same way and for the same reason--no dictionaries to remind anyone of what words used to mean. On the other hand, a community that speaks the same language, with no change in technology or climate, might have a conservative influence opposing radical change in the language, acting like gravity pulling the language back toward a stable template. What we obviously don't understand is how these two opposing tendencies operated throughout prehistory. Which tendency is stronger or does each operate under certain different circumstances?

[Miles]

Sorry, I was referring to Ornithophobe's point, not Michael's.