I often have trouble with claimed cognate words without any references or analysis. This very often applies also to Wiktionary.
E.g. even the given example - I would find it believable that "muscle" cognates with Latvian "miesa", Russian "мяса" and English "meat", but "mouse" seems sketchy.
I had a teacher in high school, specializing in ancient latin and greek, who told me about the musculus --> little mice connection. He was also a very strong Austrian -- to demonstrate the reason, he pulled up his shirt sleeve, made a classic bicep curl motion, and rotated his fist rapidly from inside to outside facing. It was pretty shocking to see the 'little mice' running under the 'covers' of the skin of his arm -- since then, I've never had trouble believing this particular hypothesis.
To add on this, musculus is not only an entirely regular diminutive derivation of mouse (mus + -culus) it's even the same pattern as is used for testicle: it's a small witness (testis, see also latinate English words like testify) of manhood.
testis (n.) (plural testes), 1704, from Latin testis "testicle," usually regarded as a special application of testis "witness" (see testament), presumably because it "bears witness to male virility" [Barnhart]. Stories that trace the use of the Latin word to some supposed swearing-in ceremony are modern and groundless.
Not really involved in the field, but this makes me wonder, would training deep neural nets first on root languages such as Latin and ancient Greece improve speed of learning subsequent languages that arose after? Essentially, make the steps that need to be learned between languages smaller, helping to bridge larger translation tasks between modern languages more quickly.
I can't help getting annoyed at anything showing "unexpected etymology" that ignores key elements of the structure of words. Fledermaus isn't a cognate with mouse, it's that the German word for bat is "flying mouse" (compare English flying fox for a type of big fruit bat).
This reminded me, there is more recent discussion than I was previously aware for the etymology of kobieta (Polish word for woman), an outlier in Slavic for a word that naively seems would be ancient.
> The information is for the most part mined from Wiktionary.
It's not a popular opinion here to criticize a star of the open Internet, but Wiktionary is not a reliable source of information (unless I misunderstand its provenance - it is crowd-sourced?). I love crowd-sourcing, but not for factual research.
And this is how misinformation spreads - now someone builds another thing on top of Wiktionary.
The core of the problem, however, is that the reliable sources of etymology, such as the Oxford English Dictionary, are not open and free. You can't just build a visualization of the relationships. Shame on the scholars for hiding the most precious, valuable treasures of civilization behind walls, when the miraculous opportunity finally came to share them with the world globally and freely (i.e., the Internet).
In my experience wiktionary is a pretty great+reliable source for word etymology. I've corrected a few things, but generally it gets it right faar more often than wrong, is good about citing sources, and it has an active + helpful community. It is pretty reliable in my experience for English/German/Latin/Chinese (in order of quantity of experience).
Growing up in my dictionaries etymology always stopped at Greek/Latin/Old English, which is a shame in some ways. Having easy access to wiktionary to plumb the depths further back to older reconstructed languages is a treat :)
> In my experience wiktionary is a pretty great+reliable source for word etymology. I've corrected a few things, but generally it gets it right faar more often than wrong ...
Serious question: How do you know if it's right or wrong?
How do I know the Oxford English Dictionary is accurate? I trust it because experts trust it, because experts write it, and because it's had over a century to mature and it has retained its reputation for that long. Also, they show and cite actual quotes.
> good about citing sources
Ever check cites on Wikipedia? Many of them do not at all support what is written in the article.
> Serious question: How do you know if it's right or wrong?
I don't know much linguistics, but sometimes the sound-changes involved seem plausible to me which leads me to trust it ("yeah that makes sense").
When I do follow the references they normally check out. In some cases where there weren't references I did some sleuthing myself and provided them (things checked out).
In another case it seemed that two words should be marked as related, so I asked on the etymology scriptorium ( https://en.wiktionary.org/wiki/Wiktionary:Etymology_scriptor... is their etymology messageboard - I find it highly entertaining to browse through people figuring out a random assortment of etymologies in the same place), and someone took the time to explain to me (to my satisfaction) that no it was just by random chance.
A few other times when I've found problems (last one was some vowel-length inconsistencies for a Latin word entry) I've asked about them, had them confirmed as problems, and fixed them (in the case mentioned, it was reverting someone else's erroneous change).
Also, some people I know who are a lot more experienced with linguistics [still amateurs, but...very highly skilled amateurs] than I am contribute to it.
All of these experiences lead me to have a high level of regard for wiktionary. As said, I'm not very experienced with linguistics, but I haven't gotten the sense that the people running the shop are anything other than competent.
>Ever check cites on Wikipedia? Many of them do not at all support what is written in the article
Dictionary citations of the sort you get on wiktionary tend to be less open to interpretation.
> someone took the time to explain to me (to my satisfaction)
> I've asked about them, had them confirmed as problems, and fixed them
Based on that, Wiktionary's standard of accuracy is what will be accepted by a non-expert. That's a pretty low standard; that's people talking in a college dorm room or in a bar. It's nothing personal - my intuitions on etymology are no better.
That's also how misinformation is created and spreads; it's right out of the textbook. Research shows that our intuitions about what's true are terrible if we don't have expertise in the issue, and that is how we are most easily fooled.
"The first principle is that you must not fool yourself -- and you are the easiest person to fool." - Richard Feynman
Yeah as a non-expert I can only vouch so far. I can say that the time I did my deepest dive (about the various Latin words that look like 'pila' if you ignore vowel length, and their etymology - https://en.wiktionary.org/wiki/pila#Latin), using wiktionary as a starting point for trawling 'real' scholarly references I presented the work to my Latin teacher, a university professor of Ancient Greek, who said it checked out (though, he's a classicist and not a general linguist, so maybe that doesn't count. And maybe he was just humouring me).
What we need is a linguist to get on here and give their take, and tell us how it fares compared to the professional resources.
> What we need is a linguist to get on here and give their take, and tell us how it fares compared to the professional resources.
Agreed. BTW, regardless of Wiktionary, I love the research - it's such creative, intriguing, stimulating work to explore those things. When people react, as (IMHO) they are conditioned to, to intellectual things with fear, hesitation and/or negativity, I think 'ugh, you are missing so much, the most beautiful things in this universe and many others.'
And see my other comment about the OED. It's the best tool for it.
(To reiterate: I'm not very learned in this domain. This is just an amateur impression).
For English it's pretty solid from what I've seen, and the way it presents etymologies as coherently written readable articles is more accessible than Wiktionary, which is a lot rawer. It's nice that it goes back before written sources to reconstructions of older languages (which OED doesn't do IIRC, but Wiktionary does). On the Proto-Indo-European language front Wiktionary is slightly more luxuriant in this regard because you can search through non-English languages as well and explore etymologies a bit more freely because of this.
It seems to lacks citations as to where the info comes from, which is a bit unfortunate, but I guess it's part of its friendly vibe? But it also obscures how people know this stuff, and makes it harder to fact-check. (I'd be surprised if they didn't have the references stored somewhere that's not getting published - I imagine they're something you'd want to keep track of as you're writing the articles).
[I don't use Etymonline very much because I'm mostly looking at relating etymologies between words of different (Indo-) European languages I'm learning/know right now, rather than plumbing the origins of single English words.]
In fact, the basic sources used by etymonline.com are mentioned on the site's home page, which also links to this more extensive (complete?) list of primary and secondary sources: https://www.etymonline.com/columns/post/sources
A few of the individual entries do reference one or more of these sources by name; for example, the entry for "better" says "...Boutkan finds no good IE etymology", where "Boutkan" is one of the entries under "Other Sources".
Why is not reliable? The same has been said about Wikipedia for long time, but for some reason Wiktionary still does not get the same level of trust as Wikipedia. A lot of the etymologies on Wiktionary come from reputable sources such as the mentioned OED. In some cases there might be multiple conflicting sources and theories, and such complex cases are likely misrepresented by the automatic extraction tools.
> Not reuses wholesale, but cites as a source of information. [Which is what you want, right?]
Yes! Thanks.
As an aside: I splurged on an OED subscription, which isn't cheap (they had a sale recently, maybe still going on, but usually it's something like $300/yr). If you care about concepts and ideas, I can't recommend it enough: It's a dictionary of every concept to which anyone has ever assigned a word (or term) in English, and in minutes you can see the concept from every perspective it's been seen, in every time and place, and the OED bring that together with the primary sources - the actual (brief) quotes where the person introduced the term. IME, it's also the best place to start for scientific and mathematical terms. If you look up "relativity" you get original quotes from Maxwell, Poincare, Einstein, etc., as well as several paragraphs succinctly and clearly defining the Special Theory, General Theory, etc.
Agreed; from my time with using it in University OED's early usage sources are A+++. Definitely blows wiktionary's out of the water (in as much as you can blow something out of the water that doesn't exist to begin with).
> Wiktionary reuses OED content? I'm surprised the OED allows that.
Facts (etymologies) are not copyrightable. The exact same text can't be used of course, but it can be rephrased, or to some extend quoted under "Fair use".
28 comments
[ 3.0 ms ] story [ 65.7 ms ] threadE.g. even the given example - I would find it believable that "muscle" cognates with Latvian "miesa", Russian "мяса" and English "meat", but "mouse" seems sketchy.
testis (n.) (plural testes), 1704, from Latin testis "testicle," usually regarded as a special application of testis "witness" (see testament), presumably because it "bears witness to male virility" [Barnhart]. Stories that trace the use of the Latin word to some supposed swearing-in ceremony are modern and groundless.
https://www.etymonline.com/word/testis
Has a good description of this derivation.
I was thinking of deploying it somewhere but never got around to it.
https://forum.wordreference.com/threads/all-slavic-kobieta-w...
It's not a popular opinion here to criticize a star of the open Internet, but Wiktionary is not a reliable source of information (unless I misunderstand its provenance - it is crowd-sourced?). I love crowd-sourcing, but not for factual research.
And this is how misinformation spreads - now someone builds another thing on top of Wiktionary.
The core of the problem, however, is that the reliable sources of etymology, such as the Oxford English Dictionary, are not open and free. You can't just build a visualization of the relationships. Shame on the scholars for hiding the most precious, valuable treasures of civilization behind walls, when the miraculous opportunity finally came to share them with the world globally and freely (i.e., the Internet).
Growing up in my dictionaries etymology always stopped at Greek/Latin/Old English, which is a shame in some ways. Having easy access to wiktionary to plumb the depths further back to older reconstructed languages is a treat :)
Also many etymological resources are open/free - http://pielexicon.hum.helsinki.fi/ for instance.
[disclaimer: I can't compare it to closed resources, and I don't work professionally with this data.]
Serious question: How do you know if it's right or wrong?
How do I know the Oxford English Dictionary is accurate? I trust it because experts trust it, because experts write it, and because it's had over a century to mature and it has retained its reputation for that long. Also, they show and cite actual quotes.
> good about citing sources
Ever check cites on Wikipedia? Many of them do not at all support what is written in the article.
I don't know much linguistics, but sometimes the sound-changes involved seem plausible to me which leads me to trust it ("yeah that makes sense").
When I do follow the references they normally check out. In some cases where there weren't references I did some sleuthing myself and provided them (things checked out).
In another case it seemed that two words should be marked as related, so I asked on the etymology scriptorium ( https://en.wiktionary.org/wiki/Wiktionary:Etymology_scriptor... is their etymology messageboard - I find it highly entertaining to browse through people figuring out a random assortment of etymologies in the same place), and someone took the time to explain to me (to my satisfaction) that no it was just by random chance.
A few other times when I've found problems (last one was some vowel-length inconsistencies for a Latin word entry) I've asked about them, had them confirmed as problems, and fixed them (in the case mentioned, it was reverting someone else's erroneous change).
Also, some people I know who are a lot more experienced with linguistics [still amateurs, but...very highly skilled amateurs] than I am contribute to it.
All of these experiences lead me to have a high level of regard for wiktionary. As said, I'm not very experienced with linguistics, but I haven't gotten the sense that the people running the shop are anything other than competent.
>Ever check cites on Wikipedia? Many of them do not at all support what is written in the article
Dictionary citations of the sort you get on wiktionary tend to be less open to interpretation.
> seem plausible to me
> someone took the time to explain to me (to my satisfaction)
> I've asked about them, had them confirmed as problems, and fixed them
Based on that, Wiktionary's standard of accuracy is what will be accepted by a non-expert. That's a pretty low standard; that's people talking in a college dorm room or in a bar. It's nothing personal - my intuitions on etymology are no better.
That's also how misinformation is created and spreads; it's right out of the textbook. Research shows that our intuitions about what's true are terrible if we don't have expertise in the issue, and that is how we are most easily fooled.
"The first principle is that you must not fool yourself -- and you are the easiest person to fool." - Richard Feynman
What we need is a linguist to get on here and give their take, and tell us how it fares compared to the professional resources.
Agreed. BTW, regardless of Wiktionary, I love the research - it's such creative, intriguing, stimulating work to explore those things. When people react, as (IMHO) they are conditioned to, to intellectual things with fear, hesitation and/or negativity, I think 'ugh, you are missing so much, the most beautiful things in this universe and many others.'
And see my other comment about the OED. It's the best tool for it.
For English it's pretty solid from what I've seen, and the way it presents etymologies as coherently written readable articles is more accessible than Wiktionary, which is a lot rawer. It's nice that it goes back before written sources to reconstructions of older languages (which OED doesn't do IIRC, but Wiktionary does). On the Proto-Indo-European language front Wiktionary is slightly more luxuriant in this regard because you can search through non-English languages as well and explore etymologies a bit more freely because of this.
It seems to lacks citations as to where the info comes from, which is a bit unfortunate, but I guess it's part of its friendly vibe? But it also obscures how people know this stuff, and makes it harder to fact-check. (I'd be surprised if they didn't have the references stored somewhere that's not getting published - I imagine they're something you'd want to keep track of as you're writing the articles).
[I don't use Etymonline very much because I'm mostly looking at relating etymologies between words of different (Indo-) European languages I'm learning/know right now, rather than plumbing the origins of single English words.]
A few of the individual entries do reference one or more of these sources by name; for example, the entry for "better" says "...Boutkan finds no good IE etymology", where "Boutkan" is one of the entries under "Other Sources".
I don't think Wikipedia is reliable at all. I can't speak for others.
> A lot of the etymologies on Wiktionary come from reputable sources such as the mentioned OED.
Wiktionary reuses OED content? I'm surprised the OED allows that.
Not reuses wholesale, but cites as a source of information. [Which is what you want, right?]
Yes! Thanks.
As an aside: I splurged on an OED subscription, which isn't cheap (they had a sale recently, maybe still going on, but usually it's something like $300/yr). If you care about concepts and ideas, I can't recommend it enough: It's a dictionary of every concept to which anyone has ever assigned a word (or term) in English, and in minutes you can see the concept from every perspective it's been seen, in every time and place, and the OED bring that together with the primary sources - the actual (brief) quotes where the person introduced the term. IME, it's also the best place to start for scientific and mathematical terms. If you look up "relativity" you get original quotes from Maxwell, Poincare, Einstein, etc., as well as several paragraphs succinctly and clearly defining the Special Theory, General Theory, etc.
Facts (etymologies) are not copyrightable. The exact same text can't be used of course, but it can be rephrased, or to some extend quoted under "Fair use".
dog: http://www.lexvo.com/info/eng/dog
bird: http://www.lexvo.com/info/eng/bird
which can only be traced backed to Anglo-Saxon and have no deeper relatives in any known language.