Comparative-historical linguistics became an academic pursuit in the 19th century. At first, it was letters that were compared, but finally, with the development of phonology, it was sounds (phones and phonemes) that were. Let's compare words from the following IE languages: Sanskrit, Avestan, (Classical) Greek, Latin, Gothic, English, Lithuanian, and Russian. These languages are all related to one another. Let's take the words for son, sheep, smoke, wolf, winter, and earth. Some words have not survived in all of these languages; in such a case, I've used a double dash to indicate this, and then in parentheses I have placed the most common word that replaced it.

PIE *suh_xnus
Skt sūnú
Av hūnuš
Gk υιυς (huius)
L -- (filius)
Goth sunus
Eng son
Lith sūnùs
Rus сын (syn)

PIE *h₂ovis
Skt ávi 'ewe'
Av -- (maēša)
Gk οις 'sheep'
L ovis
Goth avistr
Eng ewe 'female sheep'
Lith avìs
Rus овца (ovtsa)

PIE *dhuh₂mos
Skt dhūmá
Av --
Gk θυμός (thūmos)
L fūmus
Goth dauns 'odor'
Eng -- (smoke)
Lith dūmas
Rus дым (dym)

PIE *wḷkwos
Skt vrka
Av vəhrka
Gk λυκος (lukos)
L lupus
Goth wulfs
Eng wolf
Lith vìlkas
Rus волк (volk)

PIE *ĝheim-
Skt himá 'snow'
Av zyam
Gk χειμα (kheima)
L hiems
Goth -- (wintrus)
Eng -- (winter)
Lith žiemà
Rus зима (zima)

PIE *dheĝhom- (older reconstruction *ĝðem-)
Skt kshās
Av zā, zəm
Gk χθών (khthōn)
L humus 'soil' (terra 'land', tellus 'earth', cf. also homo 'man')
Goth -- (airþa, land)
Eng goma 'man' (earth, land, soil)
Lith žẽmė
Rus земля (zemlja)

[1. Edited to fix formatting. 2. Fixed related words in *dheĝhom- section for latin and English.]

This message has been edited. Last edited by: zmježd,

I notice that some of the PIE examples contain the number 2 (such as *h2ovis). What does that stand for? Or am I just seeing that as an artefact of my browser's terrible rendering of non-ASCII characters?

I am at work and have to use IE5.5 still.

number 2

They are an artefact of my never having been able to get subscripts to show up properly on this site. There are some phonemes called laryngeals in PIE, discovered by Saussure but not really proven until Hittite (and other Anatolian languages were discoverd and deciphered). There are 3 or 4 variants, and subscripts are used to denote them, while linguists try to sort out what they may have sounded like. h1, h2, h3, h4, and hx (as in the reconstructed word for 'son') when it is not even known which of the variants they might have been. I think goofy has been successful posting subscripts, maybe he could send me a PM and I can edit my opening post.

The code for subscripts is {sub}2{/sub} (replace the curly brackets with square brackets), thus: H₂O.

Go to Tools > Help, then UBB Code. Oddly, there doesn't seem to be a UBB code for superscripts.

Go to Tools > Help, then UBB Code.

Thanks, arnie; that's worked. I swear I tried it a while back and it didn't, but anyway, I've gone back and edited my OP. (Also replaced /g'/ with /ĝ/ and /l/ with /ḷ/.

quote:

PIE *dheĝhom- (older reconstruction *ĝðem-)
Skt kshās
Av zā, zəm
Gk χθών (khthōn)
L -- (terra 'land', tellus 'earth', humus 'soil')
Goth -- (airþa, land)
Eng -- (earth, land, soil)
Lith žẽmė
Rus земля (zemlja)

The English derivative was Old English guma "man", and it survives in bridegroom, from earlier brydegome. The r was inserted by association with the unrelated word groom.

Latin humus 'soil' is a derivative, at least according to this.

goma [...] humus

Right, you are, goofy. My bad. I've edited the section. Also, Latin homo 'man' whence English human.

zmježd's examples demonstrate a problem with the technique of claiming that two languages are related because they share a lot of words that have a similar sound and meaning - sometimes called mass comparison. Often you can't tell that words are related by superficially comparing them, because related words, like the words in zmježd's examples, aren't phonologically similar at all.

I saw that z posted this article on another word board, and I found it fascinating. Giving z full credit for finding it, here it is.