Talk:False cognate
![]() | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||
|
|
|
Polish: Kim , Turkish: Kim (English: Who)
[edit]Is this a false cognate? 1907AbsoluTurk (talk) 12:48, 2 August 2011 (UTC)
- Likely, pronouns are rarely borrowed. 惑乱 Wakuran (talk) 00:47, 13 June 2012 (UTC)
- Clarifying, Polish is a Slavic and an Indo-European language, while Turkish is a Turkic and possibly an Altaic language, and there's no mainstream propositions that clearly establish a link between the two families.
If you adher to the Nostratic hypothesis, I guess they would be considered distant cognates. 惑乱 Wakuran (talk) 10:53, 21 June 2012 (UTC) - Polish kim is instrumental or locative, though. So more like "whom". The correct translation of Polish kim into Turkish would thus be kimde (locative sense) or kimle (comitative/instrumental sense).
- By the way, there's also Sanskrit kim "what", whose etymology is a bit unclear. --Florian Blaschke (talk) 16:48, 9 February 2014 (UTC)
- All the Indo-European etymologies come to the PIE kwom which meant "whom" (accusative of kwos, "who"). This is very likely to be a cognate with Turkish because Turkic belongs to Altaic, a branch of Eurasiatic languages, to which PIE also belongs and the pronouns are considered the most conservative words. Moreover, the interrogative pronoun ki "what" has been already reconstructed for Eurasiatic. --Reciprocist (talk) 15:39, 31 March 2014 (UTC)
- PIE and Altaic are only related on the Nostratic theory, which is not widely accepted outside Russia. --Sir Myles na Gopaleen (the da) (talk) 16:14, 31 March 2014 (UTC)
- Nostratic is a wider groupping, but I was referring to Eurasiatic languages which are a narrower part of Nostratic.--Reciprocist (talk) 18:54, 31 March 2014 (UTC)
- And even that is not part of the general consensus of linguists. --Sir Myles na Gopaleen (the da) (talk) 14:40, 1 April 2014 (UTC)
- Nostratic is a wider groupping, but I was referring to Eurasiatic languages which are a narrower part of Nostratic.--Reciprocist (talk) 18:54, 31 March 2014 (UTC)
- PIE and Altaic are only related on the Nostratic theory, which is not widely accepted outside Russia. --Sir Myles na Gopaleen (the da) (talk) 16:14, 31 March 2014 (UTC)
- All the Indo-European etymologies come to the PIE kwom which meant "whom" (accusative of kwos, "who"). This is very likely to be a cognate with Turkish because Turkic belongs to Altaic, a branch of Eurasiatic languages, to which PIE also belongs and the pronouns are considered the most conservative words. Moreover, the interrogative pronoun ki "what" has been already reconstructed for Eurasiatic. --Reciprocist (talk) 15:39, 31 March 2014 (UTC)
- Clarifying, Polish is a Slavic and an Indo-European language, while Turkish is a Turkic and possibly an Altaic language, and there's no mainstream propositions that clearly establish a link between the two families.
The English words "march", "marshal" and "martial"
[edit]These three words, which I've for a long time suspected of being cognates (they sound very similarly and all are military-related), apparently derive from three different sources ("march" is from PIE *mereg "edge, boundary", "marshal" is from PIE *markos "horse" (well that's what Wiktionary says), and "martial" is from the Latin god Mars<*Mavors, which almost certainly isn't from either of the above). Do you think that this triple (or at least part of it) is significant enough to be included in the article? — Preceding unsigned comment added by 85.141.164.64 (talk) 21:34, 12 July 2012 (UTC)
English mysterious, Hebrew "mistori" (מסתורי)
[edit]The entry for מסתורין ("mistorin" - mystery) in the Hebrew Wiktionary reads roughly:
...
Etymology:
1. While the root of the word מסתורין is ס-ת-ר, it's source is the Greek word μυστηριον (Mystirion), which means a secret, secret rituals. Many other languages derive their "mystery" from the same source.
2. In the Talmud, the word is usually written, as transliterated Greek, with the letter teth (ט) as מסטורין, but also with tav, signaling its connection to the Hebrew root [My comment: The Hebrew root סתר is associated with, among other things, hiding and obscurity as in the words להסתיר (to hide) or נסתר (hidden, obscure)]
3. Other Hebrew words were also renovated based on the meaning of the Hebrew root and similarity in sound to a foreign word (גאון (gaon) - genious, מסכה (masekha) - mask and more)
...
I'm not a linguist myself, and I'm not sure what's the best thing to do here: on the one hand, the words are false cognates, in the sense that the meaning of the Hebrew word is derived from the Hebrew root which doesn't share a root (as far as I know) with the Greek word.
On the other hand, this is a much weaker form of false cognates, as the Greek word only resembled a previously nonexistent structure of the root סתר (s.t.r), and was essentially borrowed and fitted to look like a Hebrew word.
Any ideas? — Preceding unsigned comment added by 109.65.201.192 (talk) 22:44, 13 November 2012 (UTC)
- I'm not a linguist either, but I'd consider them true cognates. The potential existed in the Hebrew language to form the word מסתורין, but it was not formed until outside influence from Greek (from what I gather). The fact that it fit into a pre-existing root itself a false cognate of μυστηριον is of no consequence I think. So Hebrew did indeed borrow the word from Greek, making it a cognate with English "mystery". 70.29.63.164 (talk) 23:01, 9 July 2013 (UTC)
- In other words, it's really a form of phono-semantic matching. --Sir Myles na Gopaleen (the da) (talk) 14:42, 1 April 2014 (UTC)
Mina = "I"
[edit]Estonian mina/Finnish minä (I), and Zulu mina (I); also in Turkish men/ben (I), in English me/my/mine ,etc. a Black African Aurignacian word??? (so NOT a false cognate!) Böri (talk) 13:53, 4 March 2013 (UTC)
- There's no good evidence that these words have a common origin. The connection with Aurignacian culture is highly speculative. garik (talk) 15:23, 4 March 2013 (UTC)
- Estonian sina/Finnish sinä (singular you) and Turkish sen (singular you)
Both minä - men / ben and sinä - sen are resembles a lot. There are also minun - menim, sinun - senin, hän - an etc. --88.251.4.159 (talk) 04:11, 1 January 2014 (UTC)
Moving the list
[edit]If I were to follow the reasoning of this revert strictly, I'd have to remove the entire list of examples because almost none are reliably sourced. However, per WP:PRESERVE, it would be wise to find a better place on another Wikimedia project to put these false cognates. In this archived section, I suggested making it a Wiktionary appendix, but it never got anywhere. — Preceding unsigned comment added by Damian Yerrick (talk • contribs)
- So even if off Wikimedia, how would we go about finding a better place for this list? --Damian Yerrick (talk) 16:55, 30 August 2015 (UTC)
Some of them are NOT false cognates!
[edit]Some words came from the Ice Age. Böri (talk) 11:32, 2 November 2013 (UTC)
Issue of some words actually being related cognates
[edit]Aren't some of the English words and Latin (or Latin's direct descendant languages like French, Spanish, or other Romance languages) words nearly cognates by definition as 30% of English being derived from Latin words with another 30% being derived from a Norman/French background which in itself could be attributed to a Latin background? Not saying any words here are wrong, but some might warrant looking over especially when dealing with a progenitor language or two tongues that happen to be cousins. — Preceding unsigned comment added by 97.85.39.11 (talk) 07:53, 5 January 2014 (UTC)
- If roots in two Indo-European languages are shown through sound change laws to come from different reconstructed proto-Indo-European roots, a false cognate is more likely than a loan. Case in point: German German haben (have) looks like Latin habere (have) but is related to Latin capere (take). In this case, it is recommended to try to cite a PIE dictionary. --Damian Yerrick (talk) 17:01, 30 August 2015 (UTC)
Borrowed cognates
[edit]I wonder how to handle such cases as English name and Malay nama. The Malay word is a borrowing from Sanskrit nāman, which is an actual cognate of English name. If an example like this doesn't count, we're getting into trouble with a case such as Hungarian tíz, which is considered an early borrowing from Iranian – compare especially Ossetic dæs, a true cognate of Spanish diez. Counting borrowed cognates would obviously be impracticable and lead to an endless list full of uninteresting cases – but sometimes a borrowing is only suspected, not certain, for example in some of the Semitic/Indo-European equations, like in the numeral "seven". Is the possibility of a borrowing strong enough in the case of Hungarian tíz? My impression is that the hypothesis is pretty established, but it's not the most obvious case (and even obvious seeming cases such as Mbabaran dog can mislead). Where should the cut-off be? This one is a real headache. --Florian Blaschke (talk) 13:41, 9 February 2014 (UTC)
- I'd say they should be considered true cognates. They both can trace their roots back to the same word (namely PIE hnomn). If you aren't going to consider borrowed words cognates you might as well consider English "cream" and French "creme" to be false cognates. 218.214.92.142 (talk) 09:21, 31 May 2021 (UTC)
- This doesn't answer the question what to do if a borrowing is not completely certain (as in the case of the numerals mentioned). --Florian Blaschke (talk) 22:20, 6 September 2023 (UTC)
A group at Facebook
[edit]False cognates, true friends I record how false cognates sound by speech synthesizers. --Rostofanych (talk) 07:50, 17 February 2014 (UTC)
The definition of False Cognate is wrong.
[edit]I think that the definition of False Cognate is wrong in this article. This is what I found searching the internet, and please, if you found a credible source showing I'm wrong please respond because I'm going crazy with this.
The article states: False cognates are pairs of words in the same or different languages that are similar in form and meaning but have different roots.
While a more correct definition would be: False cognates are pairs of words in the same or different languages that are similar in form but have different roots and meaning.
And there's also a lot of confusion about the difference of False Cognate and False Friend even with dictionaries like Macmillan Dictionary using both terms as the same with the definition of "a word in a language that looks or sounds similar to a word in another language but means something different". [macmillion ref - Edited by user:IronMaidenRocks] and this article that is hosted by Brown, but I don't know if it's a article from them or they are just hosting.
The difference is that although both False Cognate and False Friend are similar words with different meaning, False Cognate don't have common origin while False Friend could have. It's like... all False Cognate are False Friend but not all False Friend are False Cognate. The problem is that my research in English either came with articles saying they are the same, or articles quoting the difference given in this article from wikipedia, witch does not have a source for the definition given, like this one from Princeton, that is just a copy of the wikipedia page.
One example of False Friend would be the word Fabric in English with the word Fábrica in Portuguese that means Factory, they have the same root, the word Fabrica in Latin but different meaning.
One example of False Cognate would be the word Cute in English with the word Cute in Portuguese that means Skin, they are the same word with different roots, Cute in English being the Latin word Acutus while the Portuguese Cute being the Latin word Cutis.
The articles that I found supporting my claim is not in English, being the most comprehensive one this article in Portuguese. But since this terms were created in French, with the meaning being the same in almost every language, I don't think would be a problem.
Mateusmat (talk) 13:01, 22 September 2014 (UTC)
- A definition of "false cognate" that does not even require a similarity in meaning would be ridiculously broad. Words that sound similar but mean something totally different are ubiquitous and therefore completely uninteresting. --Florian Blaschke (talk) 21:46, 21 February 2015 (UTC)
- A term for this is already defined as a faux ami or false friend and this article links to the page as something not to be confused with. Perhaps it should define a false friend at the top as well so it's easier to know what the confusion means. Kamusistiĉo (talk) 00:28, 22 December 2015 (UTC)
Hebrew and Persian "din" are true cognates?
[edit]I frankly think these are true cognates. "Din" in Persian appears to be a loan-word from Arabic (which is linguistically related to Hebrew). In Arabic, "din" does mean "religion". While in Hebrew "din" (דין) is used to refer to all kinds of laws (dinei nezikin=torts, din plili=criminal law, dinei mekarkein=real estate law, dinei kinyan ruchani=IP law, etc.), a "beit din" (בית דין) is specifically a court of religious law, and a dayan" (דיין) a religious judge. A secular court is called a "beit mishpat" and a judge there a "shofet" (originally meaning 'arbiter'). — Preceding unsigned comment added by 132.76.204.177 (talk) 07:38, 4 January 2015 (UTC)
- No they are not cognates. Persian "din" comes from Avestan "daena", which means "the faith" and is sometimes personified as a minor deity in Zoroastrianism. This word antedates the Arab/Muslim conquest of Iran by several centuries.
- On the other hand, the Hebrew din comes from a verbal root DYN, to judge or decide, and has no religious connotations whatever. The distinction between "din" and "mishpat" in modern Hebrew was coined to reflect Israeli constitutional practice (because of the secular nature of the state). It cannot be used to explain etymologies of two millennia or more earlier. Even in the modern religious context, "the din" means "the state of the law on religious issue X" rather than "religion" as such: the word for religion in general is "dat" (ironically, itself borrowed from the Persian word for law).
- Arabic "din" meaning religion or faith is, I think, borrowed from Persian and is not related to the Hebrew.--Sir Myles na Gopaleen (the da) (talk) 10:54, 5 January 2015 (UTC)
List needs pruning
[edit]There are a number of items for which to say that they have similar pronunciation or similar meaning is stretching things a bit. For instance
- English fruit and Hebrew perot
- English honest, Japanese honne
- English hut, Ukrainian hata
- English road, French route
- Italian donna, Japanese onna
Some may well be either true cognates or simple borrowings, sometimes from a third language, and need references
- Gascon babau, Romanian babau, Italian (dialectal) babau
- English sow, German Sau
- English observer, Russian obozrevatel
Some are just not relevant here since the pronunciations or meanings are too different, e.g.
- Dzongkha Druk (dragon) and English drake
- Greenlandic tallimat and Tagalog lima (both "five")
The mama / papa type cognates which are listed under Tamil and Korean, (with a few others) are so general that they should be put in a little group of their own. Or left out altogether since they are described plainly in the article.
Imc (talk) 20:28, 7 May 2015 (UTC)
- Except that English "drake" is derived from Latin draco and did originally mean a dragon: hence Tolkien's use of the word "cold-drake". ==Sir Myles na Gopaleen (the da) (talk) 09:19, 8 May 2015 (UTC)
- For sow and Sau, check Wiktionary; weird as it may sound, they are not the same word originally, sugu in Old English, sū in Old High German. (I was surprised myself, that's why I added it.) English observer and Russian obozrevátel’ (related to obozrénije, which has no v) are completely unconnected. I don't see any plausible way to link the babau words to each other via borrowing (given the geography) or inheritance (they shouldn't sound identical then), except that the Ukrainian one may be borrowed from Romanian or vice versa; I suspect onomatopoeia anyway. As for the first group, that's a fairly subjective judgement call to make; I see far worse examples in the list, for example, how interesting is it to have a word for and that consists of a single common vowel? As for the third group, the similarity is quite striking in the first case, as argued, and in the second you may be forgiven to analyse (well, arbitrarily chop up, as hacks doing long-range comparisons are wont to do) tallimat as ta(l)-lima-t – if you don't know anything about Eskimo languages. Only the choice of Tagalog is a bit arbitrary; one could actually compare Proto-Eskimo *taɫiman and Proto-Austronesian *lima to make the comparison a bit more impressive. --Florian Blaschke (talk) 04:39, 1 July 2015 (UTC)
- Adding Wiktionary links for some of the otherwise uncited cases might be beneficial. --Trɔpʏliʊm • blah 13:48, 1 July 2015 (UTC)
- Another funny similarity: Middle Persian spāš is glossed as "space, atmosphere" in MacKenzie's Pahlavi dictionary, p. 76. Too bad the list is gone; I enjoyed it and adding examples as I encounter them. --Florian Blaschke (talk) 16:51, 29 October 2015 (UTC)
- A less obscure example that has really had people scratch their heads is English chick and Spanish chica, both of which can refer to a young woman (especially one considered attractive or cute) but have completely unrelated origins (although the possibility that chica has reinforced chick in the "young woman, girl" sense – which is already attested in Shakespeare, so has an independent origin – in AAVE/American slang cannot be excluded and sounds entirely reasonable to me). Maybe this example can be sourced.
- As for Spanish chico "small", Wiktionary derives it from Latin ciccum "trifle", but can the match with Basque txiki "small" (which I always assumed was the origin; according to Trask, txiki is native and derived with expressive palatalisation from tiki "small") really be completely accidental? Considering that chico is rather colloquial, as well, the derivation seems perfect and more plausible than the ciccum one to me, which suffers from an unexplained phonetic mismatch: why ch-? Either way, we have yet another false cognate on our hands. --Florian Blaschke (talk) 03:11, 8 November 2015 (UTC)
- Kroonen in his Etymological Dictionary of Proto-Germanic (s. v. *bōnjan-) mentions Middle Dutch boen "fit, good, pretty" (presumably from Proto-Germanic *bōna- < Proto-Indo-European *bʰoh₂-no-), which recalls Latin bonus, but cannot be related because the Latin adjective descends from Old Latin duenos. --Florian Blaschke (talk) 17:17, 6 December 2015 (UTC)
- For a neat (and related) example "in the wild" (two Hawaiian words, actually), see Talk:Pono (word) § Latin?. --Florian Blaschke (talk) 22:13, 6 September 2023 (UTC)
Removed section on controversial hypotheses in historical linguistics
[edit]Some historical linguists presume that all languages go back to a single common ancestor. Therefore, a pair of words whose earlier forms are distinct, yet similar, as far back as they have been traced, could in theory have come from a common root in an even earlier language, making them real cognates. The further back in time language reconstruction efforts go, however, the less confidence there can be in the outcome. Attempts at such reconstructions typically rely on just such pairings of superficially similar words, but the connections proposed by these theories tend to be conjectural, failing to document significant patterns of linguistic change. Under the disputed Nostratic theory and similar theories such as that of monogenesis, some of these examples would indeed be distantly related cognates, but the evidence for reclassifying them as such is insufficient. (Alternatively, apparent cognates in Eurasian language families far removed from each other could also be early loanwords, compare Wanderwort.) The Nostratic hypothesis is however based on the comparative method, unlike some other superfamily hypotheses.
There might be a slight risk of controversy, here, but I have removed the preceding section, which comes across as the result of an edit war between editors who disagreed about controversial hypotheses in historical linguistics. Controversies in historical linguistics are interesting, but this doesn't seem the right article for discussing them. Removing the silliness, all this passage is really saying is, "There can sometimes be disagreement over whether or not a pair of words really are false cognates," and that seems too obvious to even mention. However, if real controversies over supposed cognates have taken place in the literature of linguistics, it would be good to discuss them in the article. - Oliver P. (talk) 01:31, 11 August 2015 (UTC)
Sau/sow
[edit]The article claims these to be false cognates, because they do not derive from the same Proto-Germanic root. Is that really enough? How do we know they don't derive from the same PIE root? – SmiddleTC@ 13:15, 12 August 2015 (UTC)
- Yes, this is a bad example: su and sugu have the same ancestor. Garik (talk) 13:56, 12 August 2015 (UTC)
- They're different words even on the Proto-Indo-European level, so it's not a bad example. --Florian Blaschke (talk) 22:31, 6 September 2023 (UTC)
Every listed word should have a source
[edit]Lots of unsubstantiated claims on this page. One of the first words I researched was not a false cognate. Shouldn't be too hard to check dictionaries for source. --IronMaidenRocks (talk) 00:37, 16 August 2015 (UTC)
- Are you talking about whole and holism? Why do you think they're not false cognates? Garik (talk) 19:25, 18 August 2015 (UTC)
- Upon further inspection, I misinterpreted this dictionary source (note how it links 'hol' to 'whole' without definition, and I neglected to look up the root of 'whole' from this assumption): http://www.oxforddictionaries.com/us/definition/american_english/holism. Your macmillion source is completely different from the article's definition. Macmillan seems to be wrong, though (as usual). See: False friend. 'False cognate' seems to be parlance used in linguistics and language training, from what I've learned on Google Scholar. If the definition of a 'false cognate' here is wrong, one should edit the whole article to correct this instead of nitpicking a tiny and trivial exclusion on a list. Edit by user:Tropylium (which sidestepped protocols on reverting edits which are the subject of an active discussion), was sourceless and only seemed to preserve the arbitrary list instead of improving it. Careful of doing that in the future. That said, this list is quite messy and leaves no real room for sources or explanation without becoming even more of an eyesore. I suppose we should just leave the fact checking up to our language hobbyist overlords, though.
- Let's look at 'loon' and 'lunatic'. Oxford says 'loon' probably comes from lunatic. Collins says the word origin of 'loon' is obscure. American heritage says 'loon' probably comes from the shrill cry of the bird by the same name, but that 'loon' was probably influenced by 'lunatic'. So, from several sources we can say this is uncertain. Do I believe the dictionaries, or the unsourced Wikipedia article? Well, I'm really not sure if I'm dealing with a Wikipedia article or a buzzfeed post (some of the article's only sources are from blog posts, by the way). Sloppy research work, perhaps, but I'd say more like overreliance on the personal lecture-based knowledge of the article's writers. Either way, I don't know if this is worth the time to bring to Wikipedia standards. --IronMaidenRocks (talk) 10:43, 28 August 2015 (UTC)
Original Research and Unreliable Source Tags
[edit]Article gives no sources for most 'false cognates' and even the definition of such seems to be WP:OR. I'm assuming that most of these examples cannot be reliably sourced. I'm not even really sure the term as used here exists outside wikipedia and a few blog posts which use the term similarly. Also, several of the sourced 'false cognates' are from blogs. Please find better sources according to wp: reliable sources. I really have no idea how anyone here thinks they can add any text Wikipedia without a reliable source. Next time someone says 'don't trust Wikipedia', your behaviour is part of the reason.
- What type of sources do you believe are needed here? It's probably easy to procure sources demonstrating for two similar words that they have dissimilar origins (e.g. Wiktionary links would likely work — as you suggest above, detailed footnotes would be overkill); sources that specifically compare them as an example of a false cognate are going to be more difficult to find, but I don't share your skepticism that it will be impossible. We can give a shot at sourcing a couple ones and see where we get with them.
- For an example, here are some results for dork, dung, egg and eye:
- dork ~ дурак: can be found brought up at e.g. a LanguageLog comment, and a a Finnish online forum; neither will work as a source, but this suggests that a reliable source might be findable. On the other hand, this also suggests that the comparison is also easily "OR-able".
- dung ~ ttong: sourcable; presented as an example of an accidental correspondence in Samuel E. Martin (1966), Lexical Evidence Relating Korean to Japanese, Language 42:2.
- egg ~ eggah: offhand I can find no sources discussing this pair, though cooking and onomastics sites are bringing up lots of interference.
- eye ~ ‘ayin: a well-known accidental resemblance, mentioned e.g. by Mark Liberman as an example of "crank etymology" in a critique of "Edenics".
- So this quick survey suggests that perhaps on the order of about half the list might be sourcable to some degree. --Trɔpʏliʊm • blah 17:02, 28 August 2015 (UTC)
- Its not about the list, despite the fact that the list is way too big and nobody bothered to source any of their additions. Its the fact that the primary reason for the article is not sourced at all (the definition of 'false cognate' and its distinction from 'false friend'). If there is no source for this, there can be no article. Its simply the rules. And, indeed, we are discussing and listing something which was likely almost wholly conceived Wikipedia in one of the many examples of rogue Wikipedia users making their own embellishments on information. Blogs and other illegitimate sources quote the article, then it hits the media, then the colleges, then back to Wikipedia. But no legitimate source seems to be considering the same concept Wikipedia is.
- For an example, here are some results for dork, dung, egg and eye:
- The only source we have says a false friend and a false cognate are the same thing. We need to assume thats the truth until its proven otherwise. --IronMaidenRocks (talk) 19:42, 28 August 2015 (UTC)
This article has been around since 2001 and has never had any legitimate sources. If a source for at least the definition is not provided I will nominate this article for deletion, as at this point I believe it to be a creation of a sectarian Wikipedia community with little use outside some academic/linguistics parlance. --IronMaidenRocks (talk) 15:16, 28 August 2015 (UTC)
- Take a second look at the Sources section; at the bare minimum, at least Campbell and de la Fuente are entirely legit sources that demonstrate that the topic of this article is a legit concept.
- It is quite possible though that "false cognate" is not a particularly commonly used name, though. I have perhaps more frequently seen "lookalike". A quick lookaround on Google Scholar moreover suggests that "false cognate" is especially in language-acquisition studies often used simply as a synonym for false friend. --Trɔpʏliʊm • blah 16:24, 28 August 2015 (UTC)
- The common name is 'false friend', which this article says (which seems to be incorrect) is not the same thing. Unless we have a definition, we must assume that either any mention of 'false cognate' is a synonymous phrase with 'false friend', or that the subject is not important enough to have been reasonably documented in academia.. without a source. For a random example, academics in the field of tacos might mention the 'taco equilibrium' in their work from time to time, there might be some loose cohesion as to what that is when it appears on one or two papers, but unless its defined and talked about, its not really noteworthy enough to have an article on. Not one this large, comprehensive, or assumptious. The fact that people are adding to this list like its their job, violating regulations, does not bode well. It hints to me this is some weird trend for linguistics, like they're exploring a new field, perhaps. --IronMaidenRocks (talk) 19:35, 28 August 2015 (UTC)
- You ought to be looking at the definition presented, not the title. The idea that words in different languages can look similar without being related at all is well-documented. Here are a couple of quotes —
- Lyle Campbell (2013:255), Historical Linguistics: An Introduction:
- Chance (accident) is another possible explanation of similarities among compared languages, and it needs to be avoided in questions of deep family relationships. Conventional wisdom holds that 5–6 per cent of the vocabulary of any two compared languages may be accidentally similar.
- James A. Matisoff (1990), On Megalocomparison; Language 66:1:
- When the number of languages being considered is large, when their relationship (if any) is remote, and the criteria for sound correspondences are lax, it is not very hard to find 'phono-semantic lookalikes' - forms which more or less resemble each other both in sound and in meaning.
- Juha Janhunen (2009), Proto-Uralic — What, where and when; Mémoires de la Société Finno-Ougrienne 258:
- As is well known, the comparative method has effective tools for distinguishing between divergence and convergence, as well as between cognates and accidental lookalikes.
- Michael Weiss (2014), The Comparative Method, in Routledge Handbook of Historical Linguistics:
- If we compare languages known not to be related we may find individual cases of close phonetic and semantic matches, for example, Hawaiian mahina ‘month’ and Hindi mahinaː ‘month’ or Scottish Gaelic bò /boː/ ‘cow’ and Vietnamese bò /bɔ˨/ ‘cow’. These are not, of course, evidence for previously unnoticed Hawaiio-Hindi or Gaelo-Vietnamese phyla but mere chance occurrences (…)
- These should do for starters for defining the topic. You seem to be right in pointing out that "false cognate" is not a standard name for it; but to suggest that it was outright invented by "rogue Wikipedians" holds no water. --Trɔpʏliʊm • blah 22:02, 28 August 2015 (UTC)
- I don't believe the term was invented by Wikipedia, it appears often as a synonym for 'false friend' in many linguistics papers. The use of 'false cognate' as the exact term which describes the situation in your sources seems to be extrapolation and development independent of scholarly input. If they aren't using sources, they're rogue Wikipedia users, defying the most basic rule of our process. Besides that, they're going way overboard on OR by comparing words on their own and (even though accurate for the Wikipedia invented definition) drawing conclusions is original research in its own right. Unless someone says in an accredited source 'helios and tacos are false cognates', its simply not something we can put in an article about 'false cognates'. If you look in the archives, you see phrasing like "we're pretty sure [two discussed words] are false cognate" where 'we' appears to be 'Wikipedia editor consensus'.
- We do not define the topic. Academia does. If Academia has not defined the term - if Academia does not discuss the term - there is no topic. What you have in these 'starter' sources is something on which to extrapolate. That's a methodology for a different website; a forum or blog. --IronMaidenRocks (talk) 07:02, 29 August 2015 (UTC)
- Update: user:Ehrenkater removed citation tags and replaced them with arbitrary unsourced description. I've reverted to previous by User:Damian Yerrick --IronMaidenRocks (talk) 14:06, 1 September 2015 (UTC)
- Nope, no citation tags were removed ----Ehrenkater (talk) 14:35, 1 September 2015 (UTC)
- That's funny, because the history says you did. Link --IronMaidenRocks (talk) 14:45, 1 September 2015 (UTC)
- "False cognate" certainly is a real term in academia, and properly distinct from "false friend". The problem is twofold. The first is that scholars who know what a cognate is tend not to need to refer to false cognates directly, even less define the term explicitly. If you're doing linguistic reconstruction it's important to identify genuine cognates and, by implication, distinguish them from false ones. But that's a fairly elementary distinction, so the emphasis in foundational texts is to define "cognate", with the assumption that the term "false cognate" can be understood based on that. In Baldi's Linguistic Change and Reconstruction Methodology for example, Zorc refers to "negative evidence, which is sociolinguistic in nature, i.e., leading to a false cognate because the form is borrowed" (p. 176). The term is really of more use outside academia, where amateur etymologists come up with all sorts of nonsense that academics are occasionally called upon to dispel. The second problem is a handful of writers (some of them academics) who apparently don't know what "cognate" means and thus use "false cognate" in an unhelpful way as a mere synonym of "false friend" (a usage that is sadly common enough that it should probably be mentioned in this article). Chamizo-Domínguez, on page 166 of Semantics and Pragmatics of False Friends helpfully draws a clear distinction: "false cognates are only a special kind of false friends, the one I call chance false friends ... by contrast semantic false friends are true cognates because they derive from the same word." Garik (talk) 13:33, 8 September 2015 (UTC)
- If its so elementary that no one has ever talked about or defined it, why are we having this discussion? Let's just make an article on "bad ideas". We'll loosely define what a 'bad idea' is, with no source, and then make an arbitrary list of all the bad ideas that have ever happened. That's my argument anyway. Your final source looks good enough to add to the article as a definition, though. I don't have that book and that page isn't available on Google Scholar. Do you mind sourcing it in the article for us? --IronMaidenRocks (talk) 21:11, 9 September 2015 (UTC)
- "False cognate" certainly is a real term in academia, and properly distinct from "false friend". The problem is twofold. The first is that scholars who know what a cognate is tend not to need to refer to false cognates directly, even less define the term explicitly. If you're doing linguistic reconstruction it's important to identify genuine cognates and, by implication, distinguish them from false ones. But that's a fairly elementary distinction, so the emphasis in foundational texts is to define "cognate", with the assumption that the term "false cognate" can be understood based on that. In Baldi's Linguistic Change and Reconstruction Methodology for example, Zorc refers to "negative evidence, which is sociolinguistic in nature, i.e., leading to a false cognate because the form is borrowed" (p. 176). The term is really of more use outside academia, where amateur etymologists come up with all sorts of nonsense that academics are occasionally called upon to dispel. The second problem is a handful of writers (some of them academics) who apparently don't know what "cognate" means and thus use "false cognate" in an unhelpful way as a mere synonym of "false friend" (a usage that is sadly common enough that it should probably be mentioned in this article). Chamizo-Domínguez, on page 166 of Semantics and Pragmatics of False Friends helpfully draws a clear distinction: "false cognates are only a special kind of false friends, the one I call chance false friends ... by contrast semantic false friends are true cognates because they derive from the same word." Garik (talk) 13:33, 8 September 2015 (UTC)
- That's funny, because the history says you did. Link --IronMaidenRocks (talk) 14:45, 1 September 2015 (UTC)
- Nope, no citation tags were removed ----Ehrenkater (talk) 14:35, 1 September 2015 (UTC)
- Update: user:Ehrenkater removed citation tags and replaced them with arbitrary unsourced description. I've reverted to previous by User:Damian Yerrick --IronMaidenRocks (talk) 14:06, 1 September 2015 (UTC)
- The common name is 'false friend', which this article says (which seems to be incorrect) is not the same thing. Unless we have a definition, we must assume that either any mention of 'false cognate' is a synonymous phrase with 'false friend', or that the subject is not important enough to have been reasonably documented in academia.. without a source. For a random example, academics in the field of tacos might mention the 'taco equilibrium' in their work from time to time, there might be some loose cohesion as to what that is when it appears on one or two papers, but unless its defined and talked about, its not really noteworthy enough to have an article on. Not one this large, comprehensive, or assumptious. The fact that people are adding to this list like its their job, violating regulations, does not bode well. It hints to me this is some weird trend for linguistics, like they're exploring a new field, perhaps. --IronMaidenRocks (talk) 19:35, 28 August 2015 (UTC)
koma
[edit]In many Bantu languages, e.g. Kikuyu "koma" means "to sleep". Is this a false cognate with old Greek "koma"? — Preceding unsigned comment added by 93.223.164.85 (talk) 23:55, 4 September 2015 (UTC)
- Probably. Given the discussion sections above though, a more relevant question would however be "is this original research?" --Trɔpʏliʊm • blah 14:18, 5 September 2015 (UTC)
'Take your list to another website'
[edit]Is there a website or forum where all this material could be moved, cataloged, and extrapolated upon? The material you're all collecting is interesting, but it just doesn't belong here. --IronMaidenRocks (talk) 14:45, 6 September 2015 (UTC)
- I took the liberty of moving your post to its own section, so it doesn't get lost in the discussion about koma. I've long thought that there are far too many examples in this article. If you look through the archives you'll see that it's been discussed quite a few times. There used to be a note—in fact I wrote it—encouraging users to think hard about whether new examples were really necessary before adding them. Of course no one took a blind bit of notice, and now the note seems to have disappeared anyway.
- If you'd like to create a new article called "List of false cognates" or something like that, then great. Be bold. Garik (talk) 18:47, 6 September 2015 (UTC)
- Absolutely not. This article is probably getting deleted, why would we make a fork for it? --IronMaidenRocks (talk) 15:16, 7 September 2015 (UTC)
- Why should the article be deleted? It's hardly one of Wikipedia's best articles, but there should be an article about false cognates. The solution is to improve it (which should involve removing most of the examples), not delete it. Garik (talk) 13:06, 8 September 2015 (UTC)
- Its not an appropriate Wikipedia article. The concept seems to be an invention or extrapolation by users. Its been three weeks and nobody can find any source for a definition. Edit: I see we've found a good source, though. Nevermind all that. --IronMaidenRocks (talk) 21:05, 9 September 2015 (UTC)
- Would the concept be more appropriate for a Wiktionary appendix, like its appendix on Proto-Indo-European reconstructions? You might want to engage Wiktionary editors with respect to finding a new home for the list. --Damian Yerrick (talk) 00:37, 15 October 2015 (UTC)
- I'm not a Wiktionary user, but it doesn't seem to be heavily moderated over there. You might recover the list from history and give it a try. --IronMaidenRocks (talk) 05:08, 15 October 2015 (UTC)
- Old discussion, but just to be clear: this is most certainly an appropriate article for Wikipedia, as it easily passes WP:GNG for notability, with hundreds, or perhaps thousands, of references in academic books, and in reputable academic journals. Mathglot (talk) 09:41, 3 January 2019 (UTC)
- Would the concept be more appropriate for a Wiktionary appendix, like its appendix on Proto-Indo-European reconstructions? You might want to engage Wiktionary editors with respect to finding a new home for the list. --Damian Yerrick (talk) 00:37, 15 October 2015 (UTC)
- Its not an appropriate Wikipedia article. The concept seems to be an invention or extrapolation by users. Its been three weeks and nobody can find any source for a definition. Edit: I see we've found a good source, though. Nevermind all that. --IronMaidenRocks (talk) 21:05, 9 September 2015 (UTC)
- Why should the article be deleted? It's hardly one of Wikipedia's best articles, but there should be an article about false cognates. The solution is to improve it (which should involve removing most of the examples), not delete it. Garik (talk) 13:06, 8 September 2015 (UTC)
Example list viability?
[edit]Considering what Garik said above, how much of the list is actually possible to source? Are we ever going to find academics or other reputable sources saying 'yes, this and this are false cognates' or 'these words are thought to have the same origin, but do not'? Like, I always thought 'sole' and 'soul' had similar origins, but that's not the case. How can I establish that many people also think that way, enough for the reference to such a thing be notable? I can't without a good source. It's undue weight and original research otherwise, which of course, we've said over and over in these last few topics. --IronMaidenRocks (talk) 21:23, 9 September 2015 (UTC)
- False cognates are unrelated words with similar forms and meanings. To show that words are unrelated, cite their etymologies, as I have done in a few entries such as that for Dutch maar and Italian ma that go back to Wiktionary's PIE appendix. The original research issue I'll take up elsewhere. --Damian Yerrick (talk) 00:27, 15 October 2015 (UTC)
Specific source needs work
[edit]http://epubs.surrey.ac.uk/1326/1/fulltext.pdf This is used as the ref. for about five different words, but is an entire chapter from a book photocopied. The article's reference has no notation as to where the contextual material can be found. The source itself contains various notes in pencil, but doesn't say who the notes are from as far as I can tell. That's quite problematic if any of these pencil notes are being used as authoritative material. I don't know if that's the case, though. I'm not sure what tag to use for when an article's reference needs additional notation. If anyone knows this material, please add on what page and paragraph the contextual information occurs for each instance the source is used. If this material can be found in a more internet accessible format, please link it. --IronMaidenRocks (talk) 21:36, 9 September 2015 (UTC)
- The Archi words appear on p. 260 (p. 20 by the pdf's internal numbering). No etymological discussion is attached though; so yes, this seems like a case of confused sourcing, if not original research in disguise altogether. --Trɔpʏliʊm • blah 11:59, 14 October 2015 (UTC)
Pretend, a false friend of prétendre?
[edit]I was really confused by the claim that pretend is a false friend of prétendre. I'm a native French speaker in Canada and when I hear the word "prétendre", the first meaning I think of is "to claim as if something is true when it is not true". There is no other word to convey that meaning in French as far as I know.
This is also the main meaning of "pretend". Are words really false friends when they have the same principal meaning but also have other meanings? This is very different from "library" (bibliothèque) and "librairie" (bookstore), for example. It's very frequent for cognates in French and English to essentially have the same meaning but to also have nuances in their meaning depending on the context; I doubt we can claim that all these words are "false friends". — Preceding unsigned comment added by 24.212.252.210 (talk • contribs)
- The difference is that, in European French, "prétendre" means to make a claim, whether true or false. English still has that meaning when we speak of pretenders to the throne, and in the word "pretension". But the normal English meaning of "pretend" would be rendered in French by "feindre".
- I don't know about Canadian usage. It could well be influenced by English: just as in Italy "fattoria" is a farm but in American Italian it is a factory. --Sir Myles na Gopaleen (the da) (talk) 10:16, 30 September 2015 (UTC)
- Does anyone have a source?--IronMaidenRocks (talk) 07:40, 14 October 2015 (UTC)
Deleted examples
[edit]Per tag from January, I've deleted the unsourced/unchallenged examples. Please find sources if you wish to add content to the article. --IronMaidenRocks (talk) 07:50, 14 October 2015 (UTC)
- You appear to have overshot and deleted also all but one of the about a dozen of what seem like reliably sourced examples, though. Mistakes can happen I'm sure; but if you had any concerns about these cases similar to the Archi ones, please actually bring them up too. --Trɔpʏliʊm • blah 11:55, 14 October 2015 (UTC)
- One of the sources just says 'Oxford dictionary'. That's not really acceptable. Some of these sources need the ISBN codes, attempts to link them to online material, etc. I was hasty in deleting some of these, but so many of the sources were just abstract notes, my frustration got the better of me. --IronMaidenRocks (talk) 04:52, 15 October 2015 (UTC)
Some Words on the List Are Actual Cognates
[edit]The list of supposed false cognates needs some serious revision. For example, French "feu" and German "Feuer" (not to mention "fire") are all from the same Proto-Germanic root (in French of course through Frankish), and ultimately from the same Indo-European root. Special care must be taken when dealing with Indo-European languages, even if they are quite different from each other, because there are many, many cognate terms among them, even when the languages would seem disparate to someone not familiar with the historical diffusion of Indo-European languages (eg Farsi/Persian, Hittite, Hindi/Sanskrit, Germanic languages, Lithuanian, Greek, and Romance languages). A true false cognate (a funny term, I know) would be something like sheriff/sharif, in which each language is in a different family and is known not to have borrowed the term from the other. I would love to revise the list myself, but I'm at best a dabbler when it comes to comparative historical linguistics, so it's something that should be undertaken by someone with some expertise in the field.
Mpaniello (talk) 22:46, 4 May 2017 (UTC)Mpaniello
- Are you sure that feu is a Germanic borrowing? I always believed that it came from Latin focus, a hearth, like Italian fuoco and Spanish fuego. --Sir Myles na Gopaleen (the da) (talk)
- Mpaniello: A good warning to keep in mind, but perhaps not in this particular case:
- If there are other words on the list that you suspect are actual cognates, please list them here. Mathglot (talk) 10:45, 5 May 2018 (UTC)
"Day"
[edit]Booth "day", "dia", etc., derive from "dyau"--MiguelMadeira (talk) 16:32, 23 June 2017 (UTC)
No synthesis of published material
[edit]Numerous word pairs in the Examples section are going to have to come out, even some of those having two citations. This is because they violate the core principle of no original research. WP:SYNTH says: Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources. An example of this is the following:
Note that neither of the two references say anything about the word pair being false cognates; they merely list the definition of each one. (It's even worse than that in this example, as the references point to Wiktionary, and Wiktionary is not a reliable source.) In fact, the only "source" for this being a false cognate pair, is the editor who decided they sounded alike, believed they were false cognates, and listed them here, relying on two definitions from two different sources to support that decision. That is a textbook example of synthesis, and is prohibited.
Any word pair in the list that has a reference that does not say anything about the pair being false cognates is not verified by that source, and may be challenged or removed. Any word pair having two references, where each reference is merely a definition of one of the two terms, is WP:SYNTH and may be challenged and removed. Putting it another way: it's not the job of editors to find words in different dictionaries (or different parts of the same dictionary) that have different etymological derivations, and then list them here when the dictionary or other source never mentioned the pair as false cognates; that's the very definition of original research. Instead, find one source that lists both words as false cognates of each other, and cite it. Mathglot (talk) 01:44, 17 May 2018 (UTC)
An example of the latter, is this entry:
This is not WP:SYNTH, because the single source mentions that the two are false cognates. Mathglot (talk) 07:52, 8 November 2018 (UTC)
References
- ^ "ako - Wiktionary". en.m.wiktionary.org. Retrieved 2018-03-16.
- ^ "ego - Wiktionary". en.m.wiktionary.org. Retrieved 2018-03-16.
- ^ Schomer, Karine; McLeod, W. H. (1987). The Sants: Studies in a Devotional Tradition of India. Delhi: Motilal Banarsidass Publ. p. 3. ISBN 978-81-208-0277-3. OCLC 879218858. Retrieved 7 November 2018.
Thus conceptually as well as etymologically, it differs considerably from the false cognate 'saint' which is often used to translate it. Like 'saint', 'sant' has also taken on the more general eithical meaning of the 'good person' whose life is a spiritual and moral exemplar, and is therefore attached to a wide variety of gurus, 'holy men', and other religous teachers.
"mama" for "mother"
[edit]Would the fact that "mama" for "mother" has popped up independently in many languages mean that the "mama" words in the different languages are false cognates? The reason for this is not coincidental, but because the /m/ sound is easy for small children to pronounce. /b/, /p/, /d/ and /t/ are too, which is why words for "father" starting with those sounds are common in languages. 2600:1700:E660:9D60:907A:B04A:159B:F587 (talk) 16:41, 19 July 2019 (UTC)
- Although it's a fascinating topic, it's not for us, as Wikipedia editors, to speculate on such things. Just follow verifiability policy, namely: if a reliable source says they are false cognates, then you can add the word group, citing that source. Mathglot (talk) 00:10, 25 July 2019 (UTC)
Sourcing claims of word pairs as false cognates
[edit]There are numerous word pairs in the article that are either not sourced, or improperly sourced, and I'm going to start removing them, per WP:V. Here's how to avoid that:
When adding word pairs to the lists, don't simply add random similar-sounding word pairs that sound similar and have different derivations; that's not enough. It's not sufficient that they have independent reliably-sourced etymologies back to two different PIE (or other) roots; the key question is, who is claiming that the pair are false cognates: 1) a published, independent, secondary, reliable source, or 2) a Wikipedia editor based on their analysis of word origins? If it's the former, it's okay to add it; if it's the latter, it is not.
This article is not an indiscriminate collection of everything that interested editors can turn up by hunting down pairs that look similar but have different derivations. The topic of False cognates is analyzed in countless published, reliable, secondary sources, and many, many reliable examples can be found and cited. There is no reason for us, as Wikipedia editors, to add examples of our own discovery, just because our research shows they have different derivations. Moreover, that would be WP:SYNTH and is prohibited. For further details on this point, please see the discussion #No synthesis of published material above.
Tl;dr: if you want to add a word pair to the article as "false cognates", find just one source that says they are false cognates, and cite it. If you can't find one, then don't add it. Mathglot (talk) 00:20, 25 July 2019 (UTC)
- You are absolutely right in saying that the false cognate claims here need reliable sources, but I think there are a few additional things worth considering: WP:SYNTH prohibits drawing a conclusion from multiple sources and this rule was created to prevent users from claiming that their own (often controversial) interpretations (WP:POV) are sourced solely because they are based on reliable sources. However, etymologies are generally not susceptible to controversy or diverging interpretation, so if two words have similar meaning and pronunciation and their different etymologies are supported by good sources, they are worth mentioning even if no source explicitly claims they are false cognates. Of course such source would be preferred, but what I'm saying is that it doesn't need to be a rule, or at least that etymological references suffice while someone doesn't find a much better reference. Remember WP:5P5 and that the presence of these entries may even encourage other users to add better sources they may happen to know about. As you said, there are many of such sources; they only need to be found.
- If you still want to strictly abide by what our policies say, I think some bits of WP:OR are noteworthy:
The phrase "original research" (OR) is used on Wikipedia to refer to material (...) for which no reliable, published sources exist.
By "exists", the community means that the reliable source must have been published and still exist—somewhere in the world, in any language, whether or not it is reachable online—even if no source is currently named in the article. Articles that currently name zero references of any type may be fully compliant with this policy—so long as there is a reasonable expectation that every bit of material is supported by a published, reliable source.
- As you said, papers and other material about false cognates are numerous. For example, the only source relating to Latin duo/Malay dua I could find only said that the Malay term is not a borrowing from French or Italian, but it would seem quite unreasonable if no material has ever termed such striking coincidence a "false cognate" somewhere. For other examples for which not a single source is found, {{citation needed}} can be used in the meanwhile. Of course, if the statement stays completely unsourced for a long time, it can and must be removed.
- Finally, please remember that all of the above is just my opinion on the subject, which I'm not trying to impose on anyone. - Alumnum (talk) 14:31, 25 July 2019 (UTC)
- I think your comments about OR, and completely unsourced articles are fine. I have a comment about this statement:
Of course, if the statement stays completely unsourced for a long time, it can and must be removed.
- While I agree, I think that it doesn't go far enough. The Verifiability policy has this to say about it:
- All quotations, and any material whose verifiability has been challenged or is likely to be challenged, must include an inline citation that directly supports the material. Any material that needs a source but does not have one may be removed.
- This means, that any editor that isn't satisfied with the sourcing on a statement may simply remove it. It isn't actually necessary to add a {{citation needed}} template at all, although it might be a courtesy in certain situations, and I might do so, depending on how long ago the statement in question was placed. If it was just added, and especially if it was just added by a new editor, or if the article is controversial, I'm more likely to place the template. Failing that, removing unsourced material is both perfectly acceptable, and specifically in accordance with Wikipedia's written policy, regardless how long the statement has been present, or whether it was tagged or not as needing a source. The burden of proof is not on the challenger to tag, be patient, and monitor for changes; the burden is on the person who added new material, to source it. Mathglot (talk) 09:28, 7 November 2019 (UTC)
- I think your comments about OR, and completely unsourced articles are fine. I have a comment about this statement:
Unreliable sources
[edit]I'm challenging the source Mozeson (2000) as WP:FRINGE; any claims in the article sourced *only* to this book need to be removed. Mozeson has completely abdicated any tools of linguistic analysis, and bases his conclusions about language and word origins on the Bible. Here is Mozeson from the Foreword:
Put away your dictionaries – with their charming old-fashioned myths of standardized spelling and pronounciation, [sic] with their superstitious, tribal need to create a new language ("Indo-European") out of ignorance of the ancient one. Take out instead, your Bible – and a pair of human ears. ... I began from the Biblical given that Hebrew is the Mother tongue (Genesis, chapter 11). ... More English words can be clearly linked to Hebrew, than to Latin, Greek, or French.
This is the fringe of the fringe; he's in cloud-cuckoo land. As an example: on page 66, he gives the origin of the English word Each as coming from Hebrew Ekh-ud (Genesis 1:5), Earth as coming from Hebrew Eretz (Genesis 1:1), Egret as coming from Hebrew Ug-oor (Isaiah 38:14). Standard dictionaries source these words to proto-Indo-European roots: līk-, er-3, and ker-2).
Per the guideline at WP:FRINGE, no assertions of fact by Mozeson should be reported in this article, and he should be considered unreliable for any claims about language. Mathglot (talk) 18:06, 4 September 2019 (UTC)
Notice to the "day - diary" example
[edit]Day and diary are claimed not to be related, because they are derived from different PIE roots, one meaning 'to burn', the other meaning 'heaven'. Semantically it's not a problem, we could say, since in Hungarian the word ég has these two meanings: 'to burn', and 'sky'. Another coincidence, that it is claimed by linguists to be a coincidental homonymy of two words of different origins. — Preceding unsigned comment added by Istenaldja (talk • contribs) 01:59, 24 September 2019 (UTC)
Arabic و (wa) and Vietnamese và
[edit]Those words have similar pronunciation and have the same meaning "and", but I'm sure they have a different origin, hence those words can be categorized as false cognate as well.Ekirahardian (talk) 18:11, 5 October 2019 (UTC)
Examples
[edit]I'd only keep those examples where a non-expert might reasonably suspect cognate status.
- "ask" in Jaqaru (South America) and English - no reasonable connection
- "to be" in Gbaya (Africa) and English - a very basic word, so unlikely to be borrowed in colonial period. Also not very interesting (a very short word).
- "dung" in English and Korean, "eye" in Greek and Malay - same
- English "sheriff", Arabic "sharif" - meanings not very close
Some examples that should be kept:
- "bad" in Persian and English - both Indo-European
- "dog" in Mbabaram (Australia) and English - one might suspect colonial borrowing; also kind of a famous example
User38453838 (talk) 10:13, 4 December 2019 (UTC)
- Removed the above entries and a few more per WP:BOLD. Removed English tiny / Yana (California) tʼinii (thin space), since the meanings aren't that close. Kept Inuktitut kayak / Turkish kayik, since the first is an international word. User38453838 (talk) 10:28, 4 December 2019 (UTC)
- You're missing the point I fear: a false cognate is not defined to be a word that might be suspected to be cognate. Any word pair that shows accidental resemblance qualifies. The point of examples like English / Jaqaru is to illustrate that coincidences can happen (which is a surprizingly hard sell to some people). --Trɔpʏliʊm • blah 11:13, 4 December 2019 (UTC)
- I get the point, but the ones I removed were either really short words (I guess there is a false cognate in some language for nearly every simple monosyllabic word (e.g CVC, CV, VC), especially if some sound alternation is allowed), or either the meanings or the pronunciations were not that close. There are a lot more "dramatic" false cognates, such that some I added, and they'd get lost among the less interesting ones. User38453838 (talk) 08:34, 9 December 2019 (UTC)
User38453838 said,
I'd only keep those examples where a non-expert might reasonably suspect cognate status.
Yes, that sounds like a reasonable condition for inclusion in this article. But it isn't a sufficient condition. The other one, is that a reliable source documents the fact that the words are false cognates.
There is a SYNTH trap lying in wait here, whereby an eager and well-meaning editor finds two words that any reasonable person would agree look alike, then finds the words in two dictionaries (or other reliable sources) as having unrelated origins, and adds the word pair here, including citations to the two sources. This fails WP:SYNTH and must be removed. The reason is, that the editor is substituting their own judgment about what is a false cognate, over a statement in some published source that says that they are; the editor merely deduces that the words are false cognates, based on material they combined from two different sources, where neither one of the sources listed both together and claimed that they are false cognates. This is the very definition of WP:SYNTH, which is a type of original reasearch.
As you make additions to the article, please ensure that your source lists both words, and identifies them as false cognates. Having two different sources that identify the origin of just one of the words each, is not sufficient. Mathglot (talk) 11:18, 9 December 2019 (UTC)
- From an inclusionist perspective, I myself would allow very explicit examples in languages that have been in contact and have many cognates otherwise to be cited just by two dictionary entries. But I understand the synthesis argument too, and WP:NOT probably has an appropriate entry.
Would be nice to have a page for a list of false cognates on some other wiki. User38453838 (talk) 07:40, 11 December 2019 (UTC)
- From an inclusionist perspective, I myself would allow very explicit examples in languages that have been in contact and have many cognates otherwise to be cited just by two dictionary entries. But I understand the synthesis argument too, and WP:NOT probably has an appropriate entry.
I also removed the sharif/sheriff main image, as I had removed the example. Also, the image caption stated that the meanings are different, which counts *against* considering them false cognates. User38453838 (talk) 07:47, 11 December 2019 (UTC)
Proposals for examples
[edit]Could someone add Finnish poika and English boy? I don't have time to add references. Would be good to also mention Swedish pojke alongside the Finnish (Germanic language like English). Poika could be referenced from Finnish etymological dictionary, which can be found online AFAIK (the Estonian one definitely can, it mentions "poika" under the cognate "poeg"). I'm Estonian myself, but I suggest adding the Finnish entry since it's the more widely known language. (For Estonian, the derived word "poiss" sounds closer to the English than "poeg"; I don't think two actual cognates need to be listed as examples with the same false cognate though). For "boy", note that Wiktionary and Etymonline propose different etymologies.
More suggestions, can be sourced from e.g dictionaries:
- Estonian kalamari (roe/caviar), from "kala" (fish) + "mari" (berry), vs English calamari (another seafood)
- Estonian seitse, Japanese shichi (seven). The Indo-European similar in some languages (e.g Spanish siete), but there's a hypothesis that it could be a loan from IE to Uralic (see this [Wiktinary entry: https://en.wiktionary.org/wiki/Reconstruction:Proto-Uralic/%C5%9B%C3%A4j%C4%87em]).
Also, I suggest moving the haben/habere example into the English table and list English "have" alongside the German. Same with Feuer/feu.
User38453838 (talk) 12:07, 9 December 2019 (UTC)
- The boy/poika case is most likely sourcable, Google Scholar suggests that Libermann (2000) or Sayers (2012) might be worth checking if they discuss the case, in case anyone has access to these.
- The 'seven' words are however a good example of a word that is not a clear false cognate: they've been proposed to be actually cognate within variants of Ural-Altaic, which is disputed only as a theory in whole, not because there is anything individually fundamentally wrong in a comparison like this. In general, the bare fact that two similar words occur in "unrelated" languages does not a false cognate make.
- Feuer / feu would not be as good an example if including English — although I've not seen a source that spells this out in complete foolproof detail, the point is that they both share besides the meaning also the spelling feu(-), which English doesn't.
- Kalamari also sounds like a poor example: they're very close in shape, but only vaguely in meaning.
--Trɔpʏliʊm • blah 17:21, 9 December 2019 (UTC)
- I feel the word calamari being long, and pronounced basically the same, speaks in favour of including it. Couldn't find a source though, hoped to find some trivia section in an Estonian newspaper mentioning it but a quick Google search failed to find anything. User38453838 (talk) 07:55, 11 December 2019 (UTC)
- @Tropylium: Given that Japanese shichi does not go back to Proto-Japonic, but is a borrowing from Middle Chinese *tsʰit, ultimately from Proto-Sino-Tibetan *s-nis, the comparison between Estonian and Japanese is most definitely a false cognate. --Florian Blaschke (talk) 21:39, 6 September 2023 (UTC)
More proposals:
- Finnish/Estonian "ei" and Swedish/Danish "ej" ("not"). The Icelandic spelling is literally "ei", but Swedish/Danish is more interesting due to the extensive historical interaction.
- Also Swedish "nej" and "ej" are apparently unrelated - comes as a surprise to me!
- Pardon my language, but: English "shit" and Estonian "sitt". (The Finnish cognate is uncommon/dialectal according to Wikipedia). Might have trouble finding sources though. User38453838 (talk) 08:04, 11 December 2019 (UTC)
- shit (Proto-Germanic *skītaz ~ *skitaz) vs. sitt (Proto-Finnic *sitta) are at least proposed to be related (as a loanword in Finnic), but this is not well received in literature. The discussion would be sourceable, but debated etymologies like these are probably not the best examples to cover here — unless perhaps to illustrate that it is not straightforward to tell what is or is not a false etymology. --Trɔpʏliʊm • blah 16:18, 26 April 2023 (UTC)
Italian ciao and Vietnamese chào seems to be a good example. --Florian Blaschke (talk) 00:13, 30 March 2023 (UTC)
Mama and papa
[edit]An interesting discourse, but entirely unsourced:
Section Mama and papa type (less first sentence) from rev 931880783 of the article
|
---|
The striking cross-linguistical similarities between these terms are thought to result from the nature of language acquisition. According to Jakobson (1962) , these words are the first word-like sounds made by babbling babies; and parents tend to associate the first sound babies make with themselves and to employ them subsequently as part of their baby-talk lexicon. Thus, there is no need to ascribe the similarities to common ancestry. This hypothesis is supported by the fact that these terms are built up from speech sounds that are easy to produce (nasals like [m] or [n], typically for "mother" words, or plosives like [p], [b], [t], [d], typically for "father" words, along with the low vowel [a]). However, variants occur; for example, in Old Japanese, the word for "mother" was papa, and in Slavic languages, baba is a common nickname for "grandmother", as in Baba Yaga and babushka. In Georgian, the usual pattern (nasal for "mother", plosive for "father") is inverted: the word for "father" is mama, and the word for "mother" is deda. |
Preserving here, as a starting point for finding references, or modifying it, as appropriate. Mathglot (talk) 22:31, 21 December 2019 (UTC)
The title should be “False cognates”
[edit]The title should be “False cognates” (in the plural form). Compare it with “False friends”. SuzieMillen (talk) 12:47, 28 July 2020 (UTC)
Japanese "namae"
[edit]Can somebody please add English "name" & Japanese "namae"? I would but I'm not familiar with Wikipedia editing. — Preceding unsigned comment added by 70.29.174.201 (talk) 13:27, 18 November 2020 (UTC)
Mistake in a bibliographical reference
[edit]There is a mistake in this bibliographical reference:
de la Fuente, José Andrés Alonso (2010). "Urban legends: Turkish kayık 'boat' | "Eskimo" Qayaq 'Kayak'" (PDF). Studia Linguistica Universitatis Iagellonicae Cracoviensis. Retrieved 2015-03-06.
The author's given names are Jose Andres (I've omitted the diacritics) and Alonso is part of his family name: Alonso de la Fuente.S. Valkemirer (talk) 13:22, 3 January 2021 (UTC)
Is this a false cognate?
[edit]There are some more word pairs that I think are false cognates, such as Vietnamese mới (From Proto-Vietic *ɓəːjʔ) and Thai ใหม่ (From Proto-Tai *ʰmɤːlᴮ), both mean "new" in English. Do you think that these two words are considered false cognates?
Quang, Bùi Huy (talk) 13:36, 11 February 2021 (UTC)
- They may be, but if you add them to the article without providing a valid source, that would be original research and someone could delete them. This happened to myself a while ago. - Munmula (talk), second account of Alumnum 08:48, 12 February 2021 (UTC)
Thank you Quang, Bùi Huy (talk) 02:12, 13 February 2021 (UTC)
Also, are "Gwen" in English and "Quang" in Vietnamese false cognates? Quang, Bùi Huy (talk) 13:46, 14 August 2021 (UTC)
- Do Gwen and Quang have similar meanings? Kinship would be farfetched. Gwen is a shortening of various Welsh names. —Tamfang (talk) 02:52, 17 May 2023 (UTC)
School and Scholar
[edit]Are these two false cognates? For practically my entire life I've believed that scholar was derived from the word school, but it turns out they have wildly different etymologies (School is Germanic, Scholar is Greek, and ultimately come from a different PIE root). Do you think they are similar enough to be considered false cognates? 218.214.92.142 (talk) — Preceding undated comment added 09:05, 31 May 2021 (UTC)
- It's not about whether we think they're similar enough. It's whether a published book or academic article thinks they are similar enough. When you find one that does, write up a citation to the source you found, and include it with your edit. Thanks, Mathglot (talk) 09:30, 31 May 2021 (UTC)
- They both go back to Latin schola (ultimately from Ancient Greek σχολή skholḗ), so I don't know what you're talking about. --Florian Blaschke (talk) 00:10, 30 March 2023 (UTC)
English 'meal' and Korean '밀'(mil)
[edit]In English, the word 'meal' can refer to the flour of some grain, mainly wheat. In Korean, the word '밀'/mil/ means wheat. 𝒞𝒽ℯℯ𝓈ℯ𝒹ℴℊ (talk) 11:53, 29 March 2024 (UTC)
Start a discussion about improving the False cognate page
Talk pages are where people discuss how to make content on Wikipedia the best that it can be. You can use this page to start a discussion with others about how to improve the "False cognate" page.