Cognates
They are pair or set of words descended from a common ancestor. These are
not just words that look like each other, but are historically related words.
I found out "coffee" is not a congnate of "kaffe", or "cafe" or "kahawa" ! If a proto "h" becomes "s" in a daughter language, all instances have to follow the same rule.
Below in this document, I have listed some already established rules for cognate reconstruction.
Sanskrit* | European language |
pitR | father (english) |
mAtR | mater (greek), mother (english) |
bhrAtR | brother (english) |
dEva | deus (latin), devine (english) |
jnyA (know) | jignos (latin or greek?), znat (russian), know (english) |
asti | est (latin), ist (german), is (english) |
nAma | nom (french), name (english), nAmo (old high german) |
shatam | satem (old persian), centum (latin), cent (english) |
vid (to know) | vidEo (latin), wit (english), eidos (greek), idea (english) |
ganika (woman - not very respectful) | gynaik (woman in greek), gynaecology (english) |
manu | man (english) |
upari | uber (gernam), super (latin) |
svarga | svarog (heaven in old slavic languages) |
pAda | pod (greek), fotu (gothic), foot (english) |
madhu | mel (latin), mead (old english) |
aham, me | me (latin), me (english) |
bhAra | bear (english) |
nIDaM (nest) | nId (french) |
sarpa | serpent |
viraH (man) | wir (latin), were (english) |
danta | dent (latin, english) |
nAva | navigate, navy |
agni | ignis (latin), ogni (slavic), ugnis (lithuanian) |
parjanya (thunder god) | perun (latvian), perkunas (lithuanian), fyorgyn (icelandic) |
* - I have tried to write all Sanskrit words here phonetically; pretty
similar to ITRANS encoding.
Some Rules
I started collecting this so that I can understand
Avestan - the language of the Rakshasas. I occasionally try to read some
old Avestan text at www.avesta.org