False cognates

A forum for discussing linguistics or just languages in general.
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: False cognates

Post by Salmoneus »

qwed117 wrote: 12 May 2020 03:45 The reconstruction back to Proto-Bantu would indicate that the word was used in that area around 500 AD at minimum, well before significant Indo-Arabic trade had reached that far south
Just to point out: trade between subsaharan Africa and the north (Rome, Carthage, Egypt, Persia, India) was extensive long before 500AD. [Julius Caesar, for example, had a pet giraffe].
, and additionally, at a time when Bantu speakers would be confined to the interior of the Congo Rainforest, making a borrowing unlikely.
Proto-Bantu would have been spoken thousands of years earlier, and not in the interior of the Congo Rainforest (which is, of course, about the only place where there wouldn't have been any lions).

Do we know which Bantu languages actually have the word? Also, wiktionary only gives Swahili as having it mean 'lion' - every other language I can find it in has it mean 'genet'.

That probably makaes Bantu origin more likely. However, lion > big cat > genet could conceivably be a recent shift, as the range of the lion has collapsed.

The one thing that makes the wanderwort an intriguing idea here is that if Bantu speakers DID reach east africa through the congo (no lions!) then they could conceivably have borrowed a word for 'lion' from east africans who were already dealing with lions. And the native east africans at that point would in many case have been afro-asiatic and would have migrated from further north, where they could have either produced or received a wanderwort.

But this idea is a bit of a stretch, and only works if the word in Bantu is limited to languages from the eastern branch, or languages that could have borrowed from the eastern branch.
Additionally the anusvara in Sanskrit would imply that a loanword would probably contain a nasal vowel or a velar nasal, the most common reflexes of *ṃh in Indic languages (cf. :ind: Hindi सिंघ siṅgh lion). It would additionally be unusual for the nasalized vowel to fortite to mb.
You're assuming the word would be borrowed from Sanskrit, which seems unlikely.
Alessio
sinic
sinic
Posts: 385
Joined: 03 Sep 2012 21:27
Location: Modena, Emilia-Romagna, Italy

Re: False cognates

Post by Alessio »

:fin: muuttaa /muːtːɑː/ - to change or it changes
:ita: muta /mu(ː)ta/ - it changes

The Finnish word is from muu (other, different) + the causative verbalizing suffix -ttaa, where muu is from Proto-Finno-Ugric *mu (couldn't find an etymology up to Proto-Uralic), whereas the Italian is from Latin mutat, 3SG of mutō, which is supposed to come either directly from PIE *meytH- (to exchange) or from moveō (to move), itself from PIE *mew- (to move).
:ita: :eng: [:D] | :fra: :esp: :rus: [:)] | :con: Hecathver, Hajás, Hedetsūrk, Darezh...

Tin't inameint ca tót a sàm stê żōv'n e un po' cajoun, mo s't'armâgn cajoun an vōl ménga dîr t'armâgn anc żōven...
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: False cognates

Post by DesEsseintes »

English fire and Thai ไฟ fai are, perhaps unsurprisingly, not cognate.
GrandPiano
mayan
mayan
Posts: 2080
Joined: 11 Jan 2015 23:22
Location: USA

Re: False cognates

Post by GrandPiano »

:es-pv: da "is"
:jpn: だ da "is, are, am"

:es-pv: haize "wind"
:jpn: 風 kaze "wind"

:es-pv: bi "two"
:eng: bi-, :lat: bis "twice"

:es-pv: sei "six"
:ita: sei "six", :esp: :por: seis "six"

:es-pv: urre "gold"
:lat: aurum "gold"
User avatar
qwed117
mongolian
mongolian
Posts: 4094
Joined: 20 Nov 2014 02:27

Re: False cognates

Post by qwed117 »

GrandPiano wrote: 26 May 2020 05:58 :es-pv: bi "two"
:eng: bi-, :lat: bis "twice"

:es-pv: sei "six"
:ita: sei "six", :esp: :por: seis "six"

:es-pv: urre "gold"
:lat: aurum "gold"
I'm not sure, given how little we know of proto-Vasconic, if we could make the argument that these are truly not cognates.
Spoiler:
My minicity is [http://zyphrazia.myminicity.com/xml]Zyphrazia and [http://novland.myminicity.com/xml]Novland.

Minicity has fallen :(
The SqwedgePad
User avatar
Sequor
sinic
sinic
Posts: 438
Joined: 30 Jun 2012 06:13

Re: False cognates

Post by Sequor »

qwed117 wrote: 27 May 2020 20:52I'm not sure, given how little we know of proto-Vasconic, if we could make the argument that these are truly not cognates.
I agree. I was thinking that urre could alternatively come from Latin aerem 'copper' perhaps.
hīc sunt linguificēs. hēr bēoþ tungemakeras.
User avatar
Pabappa
greek
greek
Posts: 577
Joined: 18 Nov 2017 02:41

Re: False cognates

Post by Pabappa »

Trask claims there is no connection, and that "old B" had a word urragin "silversmith", though I dont think that what he calls old B is the same as what we call old Basque, so it might not be that old.

https://www.bulgari-istoria-2010.com/Re ... echnik.pdf

edit: i think old B means medieval Bizkaian Basque.
Kavunupupis, šiŋuputata.
When I see you pointing at me, I know I'm in trouble. (Play)
GrandPiano
mayan
mayan
Posts: 2080
Joined: 11 Jan 2015 23:22
Location: USA

Re: False cognates

Post by GrandPiano »

qwed117 wrote: 27 May 2020 20:52
GrandPiano wrote: 26 May 2020 05:58 :es-pv: bi "two"
:eng: bi-, :lat: bis "twice"

:es-pv: sei "six"
:ita: sei "six", :esp: :por: seis "six"

:es-pv: urre "gold"
:lat: aurum "gold"
I'm not sure, given how little we know of proto-Vasconic, if we could make the argument that these are truly not cognates.
Apparently bi was originally biga (attested in some dialects and apparently the form reconstructed for Proto-Basque), so based on that a borrowing from Latin bis seems unlikely. It also seems odd to me that, if Basque had borrowed its word for "two" from Latin, the borrowing would be from bis "twice" and not from duo "two".

As for sei, Trask (linked by Pabappa in the previous post) says "Attempts at deriving this from Rom. have failed, since all neighbouring Rom. varieties have a final sibilant in their word for ‘six’, and hence a borrowing should have yielded a Bq. *seits or *seis, at best." A borrowing from Latin also seems unlikely to me, since the Latin word was sex.

For urre, Trask says that a connection with aurum isn't possible, as Pabappa noted. I think *auru would be the expected result if aurum had been borrowed into Basque.

If by "Proto-Vasconic" you mean the hypothetical shared ancestor of Basque and Aquitanian, then yes, we don't know much about that; however, Proto-Basque has been reconstructed, and we also know what correspondences to expect between Latin loanwords in Basque and the Latin words they came from.
User avatar
All4Ɇn
mayan
mayan
Posts: 1765
Joined: 01 Mar 2014 07:19

Re: False cognates

Post by All4Ɇn »

:hun: föld "earth, soil, field"
:eng: field

I would have guessed this was a borrowing from some variety of German but evidently not
User avatar
Xonen
moderator
moderator
Posts: 1080
Joined: 16 May 2010 00:25

Re: False cognates

Post by Xonen »

All4Ɇn wrote: 04 Jun 2020 22:26 :hun: föld "earth, soil, field"
:eng: field

I would have guessed this was a borrowing from some variety of German but evidently not
For what it's worth, Hungarian Wiktionary does suggest that it's "maybe" a borrowing from German Feld. There's also the possibility in cases like this that the association with a similar-looking foreign word has reinforced the use of the word for the same or similar meaning(s), even if it ultimately comes from different source.
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: False cognates

Post by Salmoneus »

GrandPiano wrote: 30 May 2020 03:58
If by "Proto-Vasconic" you mean the hypothetical shared ancestor of Basque and Aquitanian, then yes, we don't know much about that; however, Proto-Basque has been reconstructed, and we also know what correspondences to expect between Latin loanwords in Basque and the Latin words they came from.
However, we also have to bear in mind that Vasconic had been living alongside Indo-European languages, including undescribed Indo-European languages that may well have been closely related to Italo-Celtic, for hundreds if not thousands of years before the era of Latin-to-Proto-Basque borrowings - and that it may well have migrated alongside Indo-European from the steppe, indicating millennia more of loans, if not indeed a direct family connexion. And numerals are tricky in particular because they're often subject to irregular shifts (like dropping final consonants as a result of the onsets of the next number when counting, for instance).

[Or maybe sei vs Old Irish sé just proves the whole Milesian Hypothesis!]

All4En/Xonen: perhaps even more temptingly than 'field', there's 'fold' (PGmc *faludaz).


On which note:

English: fold (land, earth, country, archaically a nation or group - from OE folde, related to 'field')
English: fold (enclosure, field, poetically a group of animals or people - from OE fald, not related to 'field')
English: fold (valley in an undulating landscape - from OE faldan, theoretically not related to 'field', although the PIE words for 'flat' and 'folded' being almost the same seems rather suspicious)
Last edited by Salmoneus on 18 Jun 2020 16:38, edited 1 time in total.
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: False cognates

Post by Salmoneus »

French/English: timbre
French/English: tambour

and likewise:
English: timbrel - a tambourine
English: tambourine - a timbrel

The former words are from Greek 'tympanum', which has an internal etymology from 'to beat', while the latter are from the Middle Eastern word for a lute (it makes more sense when you remember that the soundboxes of most middle-easten lutes were covered in skin, not wood, and often had open backs, so were basically drums with strings stretched over them).

[however: the Greek word and instrument only occur late, and are strongly associated with imported religious practices from the middle-east, so it's not impossible that it actually is a loanword from the same source, reconfigured to match a folk etymology...]


--------------------

Latin: ordo - a band, troop or company of soldiers (among other meanings)
Italian: orda - a large army | Turkish: ordu - an army

The former word is from the same verb as English 'order', and means simply an ordering, a regimentation; the latter is ultimately either from Proto-Turkic, where it mostly means 'army', or from Mongolic, where it mostly means 'palace'. [in either case, the semantics are clearer when you realise that a Mongolian or Turkic palace was a large tent in which the army commander lived and worked, so it's a simple metaphony]
User avatar
All4Ɇn
mayan
mayan
Posts: 1765
Joined: 01 Mar 2014 07:19

Re: False cognates

Post by All4Ɇn »

Salmoneus wrote: 18 Jun 2020 16:34 Italian: orda - a large army | Turkish: ordu - an army
I found this borrowing really surprising until I realized it's just Italian's version of horde with the <h> left off due to their orthography. It's interesting how in words like this and Ungheria, Italian's principle of removing <h> when other Western European languages keep it actually moves the word closer to its etymological origin.
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: False cognates

Post by Salmoneus »

All4Ɇn wrote: 19 Jun 2020 05:01
Salmoneus wrote: 18 Jun 2020 16:34 Italian: orda - a large army | Turkish: ordu - an army
I found this borrowing really surprising until I realized it's just Italian's version of horde with the <h> left off due to their orthography. It's interesting how in words like this and Ungheria, Italian's principle of removing <h> when other Western European languages keep it actually moves the word closer to its etymological origin.
Apparently it's believed the <h> was added by the Polish for some reason - but given how quickly the word spread throughout Europe, it's possible that h-less forms could have reached Italy via a non-Polish route. Although you're right that it's more likely that the Italians just dropped the <h> again.
User avatar
Pabappa
greek
greek
Posts: 577
Joined: 18 Nov 2017 02:41

Re: False cognates

Post by Pabappa »

:eng: tantrum is not related to tantra, despite Wiktionary tracing both to languages of India. I had always assumed tantrum was a Latin word with tantra as its plural, and attributed the wide semantic gap to the >2000 years which the word has had to evolve since its supposed origin.

oh, right ... the /h/ thing reminds me also that I only learned recently that the Spanish name :esp: Horacio and its many cognates are not related to the word for prayer, oración.
Kavunupupis, šiŋuputata.
When I see you pointing at me, I know I'm in trouble. (Play)
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: False cognates

Post by Salmoneus »

A small addendum: the ordu word is apparently also one of the few words known from Ruan-ruan (/Rouran, etc).

But as Ruan-ruan may have been para-Mongolic, or (para-)Turkic, or (para-)Yeneseian, or an isolate, and may have either borrowed the word from or lent the word to Turco-Mongol, that doesn't really help anyone much. [The (other?) Turco-Mongols probably got the word from the Ruan-Ruan (along with 'khan' and 'khagan' and other political and military terms), but it's entirely possible that the Ruan-Ruan, before they were THE Ruan-Ruan and just some people, might have got it from (other?) Turco-Mongols in turn; or they my have gotten it from the Xiongnu, who in turn may have been Turkic, Mongolic, Yeneseian, Indo-European, or Other, and may in turn have gotten it from who knows where...]
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: False cognates

Post by Salmoneus »

English: bond
English: bondage

Entirely unrelated. The former is related to "band", as in a rope or cord or string. The latter is from words relating to farming, ultimately meaning roughly "residence" - related to English "neighbour" (one who farms nearby) and "Boer" (farmer of Dutch origin), and German 'bauen', 'to build'. From farming to serfdom, via Old French.

Presumably the recent shift in meaning from general 'slavery, restraint' to specific 'restraint by cords for sexual purposes' is partially coloured by a folk etymology to the word 'bond', though.
User avatar
All4Ɇn
mayan
mayan
Posts: 1765
Joined: 01 Mar 2014 07:19

Re: False cognates

Post by All4Ɇn »

Salmoneus wrote: 13 Jul 2020 15:24The latter is from words relating to farming, ultimately meaning roughly "residence" - related to English "neighbour" (one who farms nearby) and "Boer" (farmer of Dutch origin), and German 'bauen', 'to build'.
This is also the origin of -band in husband.
GrandPiano
mayan
mayan
Posts: 2080
Joined: 11 Jan 2015 23:22
Location: USA

Re: False cognates

Post by GrandPiano »

Salmoneus wrote: 18 Jun 2020 16:23However, we also have to bear in mind that Vasconic had been living alongside Indo-European languages, including undescribed Indo-European languages that may well have been closely related to Italo-Celtic, for hundreds if not thousands of years before the era of Latin-to-Proto-Basque borrowings - and that it may well have migrated alongside Indo-European from the steppe, indicating millennia more of loans, if not indeed a direct family connexion. And numerals are tricky in particular because they're often subject to irregular shifts (like dropping final consonants as a result of the onsets of the next number when counting, for instance).
Hm, that's an interesting point. I suppose it's possible that sei could be a borrowing from an ancient non-Italic IE language, but I still doubt that that's the case for bi and urre, since they only resemble Latin bis and aurum as a result of sound changes that happened in Latin (namely dw > b and z > r / V_V). And even if it's plausible on phonological and semantic grounds, sei being an IE borrowing still seems unlikely to me, since it would mean that the word for "six" was borrowed but not any of the other numbers. Though I suppose it could be plausible if the ancient Vasconic people had a superstition surrounding the number six?
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: False cognates

Post by Salmoneus »

GrandPiano wrote: 12 Aug 2020 06:07
Salmoneus wrote: 18 Jun 2020 16:23However, we also have to bear in mind that Vasconic had been living alongside Indo-European languages, including undescribed Indo-European languages that may well have been closely related to Italo-Celtic, for hundreds if not thousands of years before the era of Latin-to-Proto-Basque borrowings - and that it may well have migrated alongside Indo-European from the steppe, indicating millennia more of loans, if not indeed a direct family connexion. And numerals are tricky in particular because they're often subject to irregular shifts (like dropping final consonants as a result of the onsets of the next number when counting, for instance).
Hm, that's an interesting point. I suppose it's possible that sei could be a borrowing from an ancient non-Italic IE language, but I still doubt that that's the case for bi and urre, since they only resemble Latin bis and aurum as a result of sound changes that happened in Latin (namely dw > b and z > r / V_V). And even if it's plausible on phonological and semantic grounds, sei being an IE borrowing still seems unlikely to me, since it would mean that the word for "six" was borrowed but not any of the other numbers. Though I suppose it could be plausible if the ancient Vasconic people had a superstition surrounding the number six?
Well, and speaking not to promote a particular theory but just, as it were, to use this as an example of difficulties one can encounter...

...can you really rule out cases of similarity only due to later sound changes, when those soundchanges are themselves common, or areally common?

In the case of /z/ > /r/, for instance, we know that a few centuries after this happened in Latin in Italy, it happened again in northern Europe, in Northwest Germanic. Indeed, it's assumed to have happened independently in North and West Germanic, and possibly even several times independently in West Germanic. It's theorised that /z/ in Germanic, or at least Northwest Germanic, must have been in some way proto-rhotic, prone to rhoticising.

Is it just a coincidence that it's the branches of Germanic that were in western Europe that had this pre-rhotic /z/, when Old Latin (pre-Old Latin?) must likewise have had pre-rhotic /z/ in western Europe? Is it, we might then wonder, also a coincidence that sibilants in Iberia, a part of the Latin world under heavy Germanic influence, have always been so weird?

Let's be more specific in our hypothesis, actually. We know that across Western Europe, in the middle ages, both Germanic and Romance languages tended to have a retracted apico-alveolar /s/, as the reflex of ancestral *s - which has survived in various scattered dialects of both families. Let's assume then that the parents of these languages, Proto-Germanic and (pre-)Latin ALSO has retracted apico-alveolar /s/. [indeed, apparently this is theorised for Latin, explaining why it uses /s/ in borrowings from Hebrew: shabat>sabbath, jeshua > jesus.]

Both these families, then, seem to have two unusual traits: retracted apico-alveolar /s/; and loss of corresponding [z]. Proto-Germanic had /z/, but it became /r/ in all descendents others than those that left the Western European area very early on; pre-Latin had [z] as an allophone of /s/, but it had become /r/ already by the time of Classical Latin. Maybe these two traits are not coincidental? Because if you want a /z/ that might turn into /r/, then a retracted apico-alveolar might be exactly what you'd look for!

And if this form of sibilant was found in Germanic, Romance, and indeed Basque, but is lost (or never gained) in the Germanic that left Western Europe, maybe it was a general areal trait of Western Europe? In which case it could indeed have been found, with the resulting /z/ > /r/ shift, in other IE languages of Western Europe? Indeed, interestingly, Ogham inscriptions do indeed have two different letters for later /s/ - which doesn't mean, of course, that there was necessarily the same contrast between normal /s/ from /ts/ and retracted apico-alveolar /s/ from *s that there was in Germanic and Romance... but it would kind of fit!

And we don't even have to assume a third case of /z/ > /r/, actually. Since Basque doesn't have /z/, it could easily have borrowed a retracted apico-alveolar /z/ as /r/ itself.
Post Reply