Talk:Cognate

From Wikipedia, the free encyclopedia

Are the writing of hanzi, kanji, hanja considered cogna.

It has been calculated that if one takes a word from a language, there's a 40% ....experiment. Perhaps that number comes from some theoritical derivations, but how could a reader know?

If we cannot support the claim with actual factual or published research, I think that it's better to remove the fact, even if it sounds so interesting, atractive and beautiful as this one.

-Germán 2004-02-16 09:42

In the following article, available at:

http://www.zompist.com/chance.htm

the author (Mark Rosenfelder) states:

"We haven't worked the numbers. Even trained linguists, though they know that random matches will occur, generally can't say how many."

That after a long discussion on random resemblance.

I'd like to change the article to take out the number '40%' and replace it by something like 'high number/more than expected' and somehow summarize the Ronsenfelder's article in a new article and point to it. Would it be correct?

What do you think?

-grh 12:02, 2004 Feb 20 (UTC)

  • That phrase should be removed, if nothing else because it does not say what is meant by "roughly similar". The examples given -- "over" and "a'var" (do they mean the same thing?), "dog" and "dog" (ditto?) -- suggest that "similar" means "same consonants, similar vowels in similar places. If that is the case, then the figure is obvious much, mcuh less than 40%. Even between, say, German and French.

Contents

[edit] Origins, roots and etymologies

A word's etymology is more than its origin. It includes its history. Cognates are words in two languages that have origins in a shared ancestor. This makes their origin the same, although their etymologies will be very different. The word "etymology" is loosely used commonly for "origin" but I don't think we should encourage it. Furthermore, root, particularly as concerns Indo-European, has a specific meaning in linguistics. A root is a core morpheme, particularly in verbs (so that "denk-" is the root of the German verb "denken" and "think" the root of its cognate in English). In IE studies roots are indeed original "words" but they are taken to be the original morphemes, rather than inflected words as such. This is a useful enough distinction to be worth keeping IMO, so I have edited the page to reflect it.Dr Zen 04:01, 17 Nov 2004 (UTC)

Also, two words derived from the same IE root are called cognates only if they belong to two different languages. Same-language cognates (such as English grave and grief, royal and regal) are more properly called doublets. Psp 04:55, 30 Jan 2005 (UTC)
What about words such as "thatch" and "deck", coming from the same root, but where the wrd "deck" is borrowd from Dutch?

[edit] Milk

Afaik, "Milk" and "Leche" aren't cognates, but derive from two different IE words for milk. Also, russian Moloko might be a germanic borrowing.

Moloko come from a IE root, it is no more a "germanic borrowing" in russian than it's would be a "slavic borrowing" in german. As for the 'moloko' vowel change, see Old East Slavic language, protoSlavic Melko or Mleko -> Moloko.
You're absolutely right. I hadn't read that far. "Milk" is thought to come from a root meaning "wiping" or similar. "Milky Way" is a straight translation of the Latin "via lactea", not a parallel formation.Dr Zen 05:05, 17 Nov 2004 (UTC)

"Some philologists" trace a common IE root? Who? Can you give sources rather than weasel words? This is simply not true. They derive from very different roots.Dr Zen 08:14, 18 Nov 2004 (UTC)

Also "galaktos" means "of the milk" in Greek, so one probably ought not to say that it "signifies" milk. I have re-edited the para in question. Please substantiate the part about the common IE origin before putting it back in.Dr Zen 08:19, 18 Nov 2004 (UTC)

This guy derives them from the same root but he doesn't describe how. How he gets "gala" from "mlagHt-" only he knows. It goes without saying that he'd need to show other words that have shown similar soundshifts (which I can't see that he does). Note this discussion, which clearly indicates that two unrelated IE nouns are involved.Dr Zen 08:33, 18 Nov 2004 (UTC)

Preobrazhensky's Etymological Dictionary of the Russian Language gives an overview of various theories, including the common derivation theory. I'm happy if mention of the common derivation theory is left out, but in the interests of helpfulness, it really ought to be mentioned somewhere what galaktos, lait and leche roughly mean for the benefit of the general reader. Man vyi 11:54, Nov 18, 2004 (UTC)
Maybe it's a particularly Slavic theory? If you really like the idea, perhaps it should be there. Maybe you could put "Preobrazhensky suggests that all of these words could ultimately be cognates" or something similar. I'll leave it to your judgement. I've put an explanation of the words' meaning. No problem at all if you want to word it differently.Dr Zen 12:38, 18 Nov 2004 (UTC)

[edit] 40%

The article stated "It has been calculated that if one takes a word from a language, there's a 40% chance that one will find a word with roughly similar sound and meaning in another random, non-related language."

Many people have doubted this claim, and it has not been sourced. I've removed it. If anyone can find a source, please put it back in. – Quadell (talk) (sleuth) 14:22, Feb 9, 2005 (UTC)

Latin cotidianum, French quotidien, Irish cotiantach??

Isn't the French word an evolution from Latin, anyway?

Some other false cognates from German are "Gift" which ironcially means poison, "Wiener" which means a person from Vienna or even "Hamburger" which can mean someone from Hamburg, Germany.

I removed these examples as they're not false cognates, but rather false friends. According to the German dictionary Der Sprach Brockhaus, "Gift" in Old High German meant what it does in English. "Hamburger" and "Wiener" are borrowings, and in German are simply adjectives which can denote anything from the respective cities (whether or not these are the true origins of the foods). --Damezi 14:20, 22 Mar 2005 (UTC)

I've tried to explain the difference between false friends and false cognates. In short: if two words look or sound similar but have different meanings, they're false friends - but they can still be related. However, they're false cognates only if they have different origins - like English "have" and Latin "habere". --Damezi 00:53, 25 Mar 2005 (UTC)

[edit] Just a funny pair

These two are "false friends" through another language --

  • "Minne" (in Swedish) = memories
  • "Minne" (in Sicilian) = mammaries

Both are pronounced the same.

Neither cognates nor truly false friends, but somehow they seem related to me. Steve Rapaport 17:24, 6 Apr 2005 (UTC)

Ok... =P Actually, the Swedish word is singular, "memories" would be "minnen". Nevertheless, it's still funny...

Certain mammaries can cause memories, especially Swedish ones...--Gohiking 18:48, 5 November 2006 (UTC)

[edit] Definition

Since English capture derives from the Latin word captura, captus by way of Old French, I do not consider that a cognate: cognates to me are words that independently derive from the same Proto-Indo-European root (in this case, PIE *kap, 'to grasp'): borrowed words don't count as cognates in my book. Alexander 007 10:57, 24 Apr 2005 (UTC)

I don't know, I have seen the words being used as a catch-all phrase for all kinds of words that are somehow related, either through PIE roots or borrowings. For instance, I found this example on Swedish language, "Finland-Swedish has a set of separate terms that are close cognates of their Finnish counterparts, chiefly terms of law and government." and I have trouble understanding completely what it should mean, since Finnish and Swedish are unrelated languages with most basic vocabulary being completely different (although Finnish has borrowed thousands of Words from Swedish or Scandinavian during the ages). I believe that these words referred probably either are calques or loanwords from Finnish. I will make a note on the discussion page.
We seem to be dealing with two separate usages. The article, however, is concerned with the technical definition, which should be as I described it. English night and Lithuanian naktis are cognates, independently derived from a Proto-Indo-European root. But English capture is not cognate to Latin captura, because capture is directly derived from Latin captura by way of Old French. And this is certain, of course, because initial PIE k became h in the Germanic languages and PIE p usually became f. So PIE *kap- could not have given rise to anything like captura in the Germanic languages, but it did in Latin. The other type of usage of the term cognate can be explained, of course, as long as they are not confused. Alexander 007 05:15, 1 November 2005 (UTC)

The word cognate itself is a borrowing from Latin, and means "related". That can include a parent-child relationship as well as the relationship of cousins (unlike the German word normally used to translate cognate: urverwandt, where the ur- prefix stressses that the words are not just related, but are related in the distant past). But obviously it is cousins which are more interesting here: we have good terminology for loanwords and other forms of direct derivation, but this is the only term we have for the relationship of words in parallel branches, and there is a lot to be said for using linguistic terms in their narrowest senses where possible, for the sake of clearer differentiation. How about using "related forms" as the broad catch-all term and keeping "cognate" for the narrower meaning? --Doric Loon 00:54, 6 November 2005 (UTC)

[edit] Question

Question: Why not put (things like) this in the general linguistics category too? —The preceding unsigned comment was added by 145.99.239.173 (talk • contribs) 01:38, 11 January 2006 (UTC).

Answer: Category:Historical linguistics is a subcategory of Category:Linguistics, so that would be redundant. Wiki policy is to simplify the cats. Alexander 007 01:42, 11 January 2006 (UTC)

[edit] Superscript help

Hey, how or where do u get the superscript characters to show aspiration? I can read them as superscript when reading the article but they appear as squares in the edit script. And if I try to copypaste them from the article, they are pasted as squares. YoungSpinoza 20:35, 28 January 2006 (UTC)

[edit] Old English/Old Norse mutually intelligible?

Not pertaining to the subject of the article, but is it true that these two languages were mutually intelligible at the time of the Viking invasions of Britain? Considering the possible timeframe of the divergence of North and Western Germanic language groups around the beginning of the CE or earlier, this gives almost a thousand years for the OE and ON to diverge. Descendants of Latin and proto-Slavic have certainly become NOT mutually intelligible within a thousand years, but perhaps language change accelerated with the development of society?

Summary: words are borrowed from languages whether the borrower and the donor are mutually intelligible or not (cf. the example of words in English borrowed from French Norman.) I will investigate this further and if I don't find evidence on OE/ON mutual intelligibility, excise the parenthesized clause. Zapiens 17:17, 20 February 2006 (UTC)

[edit] New stuff

The new material I removed probably belongs in the article, but not in the lead where it was placed. It would go further below in the article. Alexander 007 19:24, 26 February 2006 (UTC)

[edit] Embarazada

Contributors to this article may be interested to know that the article on Embarazada (the Spanish for "pregnant", a false friend cognate with "embarrassed") is currently up for deletion at Wikipedia:Articles_for_deletion/Embarazada. Flapdragon 10:27, 12 April 2006 (UTC)

[edit]  ???

What is the difference between etymon [[1]], cognate, Root (linguistics) ??? and should we say "cognate" [[2]] or "cognate word" [[3]]. Sorry to insist, i saw there was already an explanation, but i was not totaly convinced.

If you think of it in terms of genealogy, etymons are like grandparents and cognates are like cousins. They both embody historical relationships. On the other hand, roots -- at least as the term is used in the linked article -- are synchronic rather than historical/diachronic; thus, "run" is (sort of) the root of "running," etc. They embody only a relationship within the language at a specific point in time. As far as I can tell, "cognate" can be either an adjective or a noun, so in most cases there is no need to say "cognate noun." Perhaps the article should deal more extensively with these matters? -- Visviva 15:06, 5 November 2006 (UTC)

[edit] about cognate

the latin word was cognatus but the word is composed by cum + natus the particle co is unexistent in ancient latin and the g on cognatus and the u becoming o are normal linguistic transformations when a word is composed by 2 or more words....can anyone control this and make the necessary corrections? i wouldnt do that because i don't read or study latin from 15 years ...... (the meaning of the word is "born togheter, born at the same time") thank you 151.80.3.184 22:08, 8 May 2007 (UTC)

See recent change. Note that "co-" is the equivalent of cum (originally quum) and is used in a prefix form. BTW, I think you mean classical Latin, not ancient. •Jim62sch• 20:47, 11 May 2007 (UTC)

[edit] Vandalism

The addition of this text

   a cognate is when your coc gets stuck in the nate and that hurts.

entered on

   16:13, 26 August 2007

is trash. I leave it to a more experienced Wikipedia editor to click (undo).


144.226.230.36 22:10, 26 August 2007 (UTC)