In getting ready for my sessions on Sanskrit at the upcoming ISCA conference, I am in the process of totally overpreparing. I mean, all I can hope to accomplish is to create enough interest that there will be sufficient momentum for people to get motivated to continue with the language--a quick introduction to the script and some fundamentals of sentence construction. Nevertheless, I'm acting as though I'm about to take a Sanskrit language exam for grad school. My nose is forever stuck in the Sanskrit textbook, grammar, or reading selections. Reading a full grammar of any language is a lot like studying 1 & 2 Chronicles; there's a lot of repetition of seemingly irrelevant information, but when you start to read it with an eye out towards little nuggets, it becomes worthwhile.

Anyway, yesterday (Tuesday) I reminded myself of the numbers from 1 to 10, which was not exactly exploring the depth of the language. However, that little venture led me to think of the various names for numbers in other IndoEuropean (IE) languages. If you need a quick refresher on the Indo-European language family, you may want to go back and look at the blog entry of February 7 of this year and the chart that went with it.

Now here is a little table of the numbers 1 through 10, with the first column giving us our familiar "Arabic" numerals. Of course, they're not Arabic in origin, but Indian. The South-Asian cultures not only had the decimal system and corresponding numeric signs (as opposed to using letters for numbers), they devised one of the most important contributions to intellectual culture in human history, the integer zero (0) as place holder. That innovation made it possible to distinguish between such numbers as 11 and 101, among other things. European cultures learned this system of numerical notation from the Arabs towards the end of the middle ages and quickly adopted it. Until then, all arithmetic and other mathematics  had to be done in Europe by means of the "Roman numerals," e.g. i, ii, v, vi, lix, etc. They used a lot of prepared charts for common arithmetical problems because the Roman numerals did not lend themselves to quick calculations, though an abacus helped. Paradoxically, we refer to our system of writing as "Roman," but when it come to writing numbers, fortunately we do not use Roman numerals. We use the "Arabic" ones, which are modifications of the Indic ones.

	Sanskrit			Greek	Latin	English	German	Cognates in English
NO.	#	Word	Translit
1			eka	ena	unum	one	eins	each? equal?
2			dva	dio	duum	two	zwei	duo, duet
3			tri	tria	trium	three	drei	triangle
4			chatur	tessera	quattuor	four	vier	quarter, tetra--
5			pancha	pende	qinque	five	fünf	pentagon
6			shash	eksi	sex	six	sechs	hexagram
7			sapta	epta	septem	seven	sieben	September
8			ashta	okto	octo	eight	acht	octagon
9			nava	enea	nova	nine	neun	November
10			dasha	deka	deccem	ten	zehn	decalog

So, what do we see on this table? Obviously, we see the usual resemblances one is bound to expect whenever someone makes a comparative table of IE languages. No oohs or aahs are expected at this time. Unless, of course, you've never encountered the fact that this family of languages stretched from Western Europe all the way to Bangladesh, and that the resemblances are still noticeable. The cohesion becomes even more remarkable once you become acquainted with the patterns in language transformations when people groups migrated and became isolated from each other.

There is an interesting example of such a linguistic change in the Bible. The context is, of course, Hebrew, a member of the Semitic language family. In Judges 12:4-6 we read of the uncomfortable encounter between the judge Jephthah, leader of the  Gileadites, who lived east of the Jordan river and the Ephraimites. Jephthah had defeated the Ammonites without having waited for the ever-ambitious Ephraimites to join the venture, which really angered the Ephraimites. So, since they couldn't kill any Ammonites any more, they decided to satisfy their blood lust by crossing over the Jordan and attacking Jephthah's army. This was not a good idea; the Ephraimites not only lost the battle, but later that evening the Gileadites kept them from crossing back to the western side of the Jordan and heading home. Everyone spoke Hebrew, but different dialects had already developed. Whenever someone approached the river, the Gileadites challenged them to say "shibboleth," but the Ephraimites were not able to pronounce the word properly, and could only come up with "sibboleth." On this basis they were identified as Ephraimites and killed. It was a rough night to be an Ephraimite, but--more importantly for us right now--it illustrates how quickly language differences show up in separated populations.

When we look at the way in which a language family develops, we can expect the words most closely associated with basic human existence to have engrained themselves in earliest times and to be carried along with the migrating tribes to their furthest extensions. As cultures change, new words are coined, grammars take different directions, and idioms pile up, but everyday words, such as basic counting numbers, are going to be there at the bedrock of the various languages.

*Interestingly, the word for "one" shows a lot of variation, and I couldn't come up with any obvious cognates in English. So I just totally guessed that maybe our words "equal" and "each" could manifest some of that resemblance.  But this is absolute guesswork. So, please disregard that field. If it serves a purpose then it is to give an example of the kind of hypothesizing that made early nineteenth-century philology a wide open playground for any number of scholars' imaginations.  If my guess were to turn out to be correct, it would still be purely luck. So, please take a magic marker and cross out that field on your computer screen.

The resemblance between the Sanskrit ashta and the German acht  is really fascinating. Ignore the "a" at the end of the Sanskrit term, which constitutes a unique wrinkle of Sanskrit: Unless otherwise specified a consonant is always followed by an "a." Then, there is a theory that the sibilant "sh" (; ) developed as a smoother version of the "ch"  sound (as in "chat") represented by the letter "c" (c). So, if we were to posit a single antecedent (such as proto-IE) for the two languages, it could be something like act or even the German acht  exactly.  Let me hasten to assure you, though, that things aren't that simple. It's just weird, if you know what I mean, to see these two languages, as widely separated as they are in their geographic and cultural settings, have such similarity in some words. But please, if necessary, take another look at the chart of languages to make sure you don't lose the appropriate sense of confusion.

Aside from such speculations, as it stands, we do see here a case in point of "Grimm's Law," one of the early principles of scientific linguistics, formulated by Jacob Grimm, who is popularly known for the collection of fairy tales accumulated together with his brother Wilhelm. According to this "rule", there is a regular shift in the types of consonants in certain words as IE languages changed over time. Hang in here with me for a second. A common expression in language studies is that consonants serve as "stops" between vowels. In the history of IE, early forms of words contained an abundance of aspirated stops, which means consonants that use a bit of extra breath than others. In Sanskrit, a relatively early IE language, about 40 percent of regular consonants are classed as "aspirated." This certainly is the case for the sht combination in Sanskrit ashta.  Though not consisting of officially designated aspirated consonants, you expel a bit of air as you say the word. Then, in later developments, a lot of these stops became hard, as seen in the Latin octo.  Finally, as the German branch of IE languages sprouted, the hard stops often turned into fricatives, "scratching" sounds, such as the stereotypical German "ch" as manifested in the German acht.  You can see another, truncated example with the change from "p" to "f" in the number five.

Another phenomenon that appears in this table is also often included under the heading of Grimm's Law. Look at the numbers 2 and 10. In Sanskrit, Greek, and Latin the words for them begin with the letter (or the sound of) "d." Then, as the Germanic branch split off the hypothetical main line, these and many similar words went into two opposite directions. Heading towards the "Anglo" twig, the stop hardened and the corresponding words now started with a "t," as in two and ten.  On the other hand, on the "High German" twig, the "d" is replaced by a "z," which receives a short pronunciation, similar to, say, the "ts" sound in words like "tsunami," namely zwei and zehn. There are many such correlations between English and German words. Other examples include tame and zahm, team and Zaum, and sometimes they are slightly hidden, e.g. between ('tween) and zwischen.

There was a physicist early in the twentieth century who was known to respond to his students' remarks in class by asking, "Are you saying this simply because it is true, or are you trying to make a point?" That reminds me of a similarly annoying Chemistry professor I once had, but that'll have to wait. In this case, there is no further big point I know to make. It's like showing someone a beautiful temple and saying, "Isn't this a gorgeous piece of architecture?" I find these relationships to be intriguing, but to make any further special point out of it would surely be gilding the lily for those who see them the same way. Those who don't find them interesting probably won't go into a state of fascination just because I should make a declaration of enthusiasm.

Though, really, just between us, don't you think this is incredibly cool?

Very Fascinating. I harbor a secret love of philology and Grimm's law. Had I the time and talent, I think I would pick it up as a study. As it is, I sit on the sidelines, pick up what I can and enjoy the interplay in different languages and their history.