Jump to content

Talk:do

Page contents not supported in other languages.
Add topic
From Wiktionary, the free dictionary
Latest comment: 10 months ago by Silmethule in topic Irish Etymology 1

Doing time in jail

[edit]

So there was an RfV discussion a while ago that I missed about a definition of "do" as to be in jail; I think what that person meant is to spend in jail, as in "I did two years in Folsom." I've added that definition. Pstinchcombe 22:06, 2 June 2007 (UTC)Reply

Homonyms format

[edit]

I really think this interlaced homonyms format is unhelpful. The verb "do" has a different etymology from the spelling of the musical note. In other articles which cover multiple homonyms we have separate subarticles for each with only the language heading being shared. The current format weaves all homonyms together with a subsection for each mixed together and labeled in the same section. If it's difficult to understand my description that's a pretty good indicator that it's also difficult to undersand the article. Especially when not all sections for all homonyms even exist yet. — Hippietrail 22:11, 9 August 2005 (UTC)Reply

From RFV

[edit]

The following information has failed Wiktionary's verification process.

Failure to be verified means that insufficient eligible citations of this usage have been found, and the entry therefore does not meet Wiktionary inclusion criteria at the present time. We have archived here the disputed information, the verification discussion, and any documentation gathered so far, pending further evidence.
Do not re-add this information to the article without also submitting proof that it meets Wiktionary's criteria for inclusion.


do

[edit]

For these senses:

RFV failed. Senses (re)moved. —RuakhTALK 04:38, 31 May 2007 (UTC)Reply

On second thought, more accurate would be fixed; it seems that the contributor put the right info at the wrong place. —RuakhTALK 18:27, 2 June 2007 (UTC)Reply

This entry has survived Wiktionary's verification process.

Please do not re-nominate for verification without comprehensive reasons for doing so.


do

[edit]

3 rfv-senses tagged since July 2007. May never have been at RfV. Worth a look, but hard to cite. How will we ever cite this kind of polysemic entry without our own corpora and a system for annotating it? DCDuring TALK 19:13, 11 May 2008 (UTC)Reply

In areas like this, IMO humility is the best policy. If a (verifiable) sense is recognized as distinct in respected dictionaries, it behooves us to keep it. If not, it should only be kept if it is clearly not covered by an existing sense (and verifiable).
In this regard I note that the attack, travel, and cook senses are all recognized as separate senses in Longman's DCE, as well as being fairly easy to cite. IMO that should be good enough, barring any strong counterarguments; or at the very least they should be on RFD instead. ("Attack" is actually given by the DCE as a specifically British sense; I'm not sure what to make of that.)
On the other hand, the "be well/take care of" sense is absent from the dictionaries I have on hand. Also, the examples that were given clearly didn't fit that sense (I've moved them to more suitable locations.) It could very well be legitimate, although I'm having a hard time imagining how it could ever be used to mean "take care of" in contemporary English without causing horrific misunderstanding. -- Visviva 15:21, 13 May 2008 (UTC)Reply
I was just the messenger, but think that RfV tags inserted in an entry are worthy of some kind of consideration. I'd like to be able to cite these things. I usually lack the Sitzfleisch.
My MW3 has the travel and cook senses (and several more that we don't have, though my working memory is not capacious enough to really tell for sure in an efficient way).
MW3 does not have the attack sense.
MW3 actually has a sense "provide", saying that is used with "well". Their usage example is from w:Arnold Bennett speaking of "lunchers doing themselves ... well" at a restaurant. DCDuring TALK 16:46, 13 May 2008 (UTC)Reply
Definitely; I didn't mean to be dismissive of the tags. I do suspect, however, that the tagger really meant to tag these as {{rfd-redundant}} rather than {{rfv-sense}}. There is not much question that all of these (with the possible exception of the "take care of" sense) are in clearly widespread use. Three cites can certainly be given for each, but that seems like a bit of overkill.
Still, to bring the spirit of verification to this issue, I've started a general data page at User:Visviva/do. It's interesting to compare the treatments given in various sources. Thinking of moving it to Appendix:Dictionary treatments of do or something like that (though that table needs some cleanup first). -- Visviva 15:05, 18 May 2008 (UTC)Reply


Requested entry

[edit]

Requested entry: do the dishes. In this sense, "do" means "wash" or "clean." This is an unusual idiomatic usage that should be covered. 24.93.170.200 07:42, 15 October 2007 (UTC)Reply

The sense in "I think I can do with some extra sleep" doesn't seem to be covered. \Mike 20:23, 12 May 2010 (UTC)Reply
Thanks. It needs to be covered in some way. I will make an entry at do with and insert it as a "derived term". We can then figure out how else to treat it here. DCDuring TALK 22:01, 12 May 2010 (UTC)Reply
Please take a look at do with. DCDuring TALK 22:37, 12 May 2010 (UTC)Reply

Dictionary notes

[edit]

See Appendix:Dictionary notes/do. DCDuring TALK 15:11, 6 July 2009 (UTC)Reply

Latin dô entry

[edit]

The conjugation table for the Latin dô entry currently lists it as first conjugation. However, dô is a irregular verb which has a conjugation that closely resembles the first conjugation. My Pocket Oxford Latin Dictionary says: "In the present stem form of dô, the vowel after 'd' is shortened if it is 'a', except in the 2sg. indicative and imperative (i.e. dâs and dâ respectively). I will be very pleased and grateful if someone will create a {{la-conj-irr-do}} template for this entry. (I use circumflexes here instead of macrons here to denote long vowels) VNNS 08:25, 19 March 2010 (UTC)Reply

Nevermind, I'm doing it myself. VNNS 08:47, 19 March 2010 (UTC)Reply

Czech diti (“to hide”) ???

[edit]

In English -- Etymology 1 -- cognates (need to press "show" to be shown) I see the following

 Czech diti (to hide)

This is to be corrected or removed. I am native Czech speaker (non-linguist) and that does not make sense to me.

Related might be:

d{hooked-e}lat (dělat) -- to do

Similar but, I think, unrelated is:

d(long-i}ti (díti) -- to say (in old Czech)

90.180.192.165 23:48, 4 December 2011 (UTC)Reply

Removed. —Stephen (Talk) 15:55, 5 December 2011 (UTC)Reply

Do (noun) can mean shit or faeces

[edit]

As in "doggy do", or "animal do", or place the word "human" or any animal species before "do". http://www.thefreedictionary.com/doggy do Regards JohnI (talk) 17:23, 10 February 2015 (UTC)Reply

Alternative spelling of doo? — Ungoliant (falai) 17:24, 10 February 2015 (UTC)Reply
That was a quick response! It could be that you are correct as it does rhyme with "poo". I had always thought that "doggy do" had more to do with the "doing", but I may be incorrect in my assumption. JohnI (talk) 17:34, 10 February 2015 (UTC)Reply

RFC discussion: January 2011–November 2017

[edit]

The following discussion has been moved from Wiktionary:Requests for cleanup (permalink).

This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.


The order of the senses and those of the translation tables do not match, which makes the page very confusing. -- Prince Kassad 09:57, 27 January 2011 (UTC)Reply


Irish and Old Irish "do"

[edit]

Irish "do", etymology 3 (your), links to the Old Irish preposition do (to, for), and hence to Proto-Celtic *tu. However, I think it should link to Old Irish *tú, Proto-Celtic *tū. Is that right? Marcas.oduinn (talk) 18:00, 24 May 2018 (UTC)Reply

'Does that ...' pronounced [tsaet ...]

[edit]

Here are two examples thereof

min 09:02: Does that feel o.k., comfortable? https://www.youtube.com/watch?v=wdLB3udjOR0&feature=youtu.be

min 1:02 : Does that sound good? https://www.youtube.com/watch?v=wdLB3udjOR0&feature=youtu.be --Backinstadiums (talk) 17:28, 4 October 2019 (UTC)Reply

auxiliary do-support: do murder

[edit]

The Cambridge Grammar of the English Language, page 98, reads

Auxiliary do was used more widely in earlier stages of the language, and in certain genres one comes across archaic uses that go beyond our do- support constructions – e.g. in legal language (on or about the 14 th day of June, 1997, he did murder one James Robinson)

What type do-support does "did murder" represent? --Backinstadiums (talk) 16:33, 6 October 2019 (UTC)Reply

I don't know that it's a named "type". It's just being used in a way we wouldn't use it today. Equinox 20:44, 19 March 2020 (UTC)Reply

din't

[edit]

OED: din't, dint (dɪnt) contracted colloq. form of didn't, did not --Backinstadiums (talk) 20:31, 19 March 2020 (UTC)Reply

Done Done Equinox 20:45, 19 March 2020 (UTC)Reply

Irish Etymology 1

[edit]

There seems to be disagreement over the origins of the particle "do" (past tense marker). If I'm interpreting the entry correctly, eDIL states that OI "no-" became Modern "do" in its entry on "no-" (http://www.dil.ie/33202) and makes no mention of a Modern descendant in its entry on "ro-/ru-" (http://www.dil.ie/35364). It should go without saying that eDIL is not the ultimate authority on the subject, but does anyone have a source on the "ro-" etymology that's any better than a claim from eDIL? Cernacas (talk) 19:06, 17 November 2020 (UTC)Reply

@Cernacas: DIL says that about “regularly in impf. ind., impf. subj. and cond.” (no preterite and perfect – later past tense continues perfect) and also it doesn’t state that this use of do continues no (ie. that it is actually derived from it) but rather that do is used instead in later language. Very similar thing is said under ro:

In mod. Ir. it [ie. ro] is displaced by do except after a neg. or a syntactical preverb: do ghabh, nír ghabh, gach ar ghabh sé.

We should probably mention both. See also Seán Ó Catháin (1933), Some Studies in the Development from Middle to Modern Irish, Based on the Annals of Ulster, section “The transition ro > do”, pp. 14–20 where he talks about “Confusion between do, ro and no” // Silmeth @talk 14:43, 29 January 2024 (UTC)Reply
@Cernacas: I’ve edited the etymology to include no, and added sources. // Silmeth @talk 16:44, 29 January 2024 (UTC)Reply

do so

[edit]

A: "Honey, you're fine. You don't have a fever." B: "I do so have a fever! --Backinstadiums (talk) 08:55, 19 April 2021 (UTC)Reply

How do we 'sposed to keep peace?

[edit]

What meaning of do is used here? and of supposed? --Backinstadiums (talk) 20:34, 10 May 2021 (UTC)Reply

@Backinstadiums: The "supposed" is be supposed to. It means "how could anyone expect us to...; how is it possible for us to...". The do is ungrammatical and non-standard; I've never seen that before; feels like a mixture of question structures like "how do I do (something)" and "how am I supposed to do (something)". Equinox 20:47, 10 May 2021 (UTC)Reply

Do/Tun/Doen as a West Germanic auxiliary verb

[edit]

The use of "do" as an auxiliary verb as in English is not as uncommon as suggested here. In at least some German dialects like Palatine German -Die deet net laut genuch rufe. = She did not call out loud enough.- and Viennese German -Tuasd eh schaun? = Do you (really) look?- it is usual as well. You can read about this in the German Wikipedia entries. In the Dutch language 'doen' is used as a nonspecific replacement for any other verb, in short sentences or statements and then sometimes the infinitive of the actual specific verb is added again. I am therefore extremely sure that the use of "do" as an auxiliary verb definitely is a West-Germanic phenomenon. Amand Keultjes (talk) 15:52, 5 June 2021 (UTC)Reply

copular verb

[edit]

Is it a link verb in That talk show host just doesn't do subtle  ? --Backinstadiums (talk) 09:30, 10 August 2021 (UTC)Reply

RFV discussion: September–October 2021

[edit]

The following information has failed Wiktionary's verification process (permalink).

Failure to be verified means that insufficient eligible citations of this usage have been found, and the entry therefore does not meet Wiktionary inclusion criteria at the present time. We have archived here the disputed information, the verification discussion, and any documentation gathered so far, pending further evidence.
Do not re-add this information to the article without also submitting proof that it meets Wiktionary's criteria for inclusion.


do

Rfv-sense: A deed; an act. - apparently there's a Walter Scott quote. Good luck searching for this, BTW... Roger the Rodger (talk) 15:25, 30 September 2021 (UTC)Reply

You don't have to RFV everything you can't personally find. Do you really doubt this sense actually exists? Equinox 19:05, 30 September 2021 (UTC)Reply
I don't have to do anything here, Eq. And yes, I really doubt this sense actually exists Roger the Rodger (talk) 12:51, 1 October 2021 (UTC)Reply
No Scott quote indicated in the OED. By the way, @Roger the Rodger, sometimes you might be able to pinpoint these supposed quotes by looking up the OED – the links to the volumes are at {{R:Oxford English Dictionary}}. — SGconlaw (talk) 20:03, 1 October 2021 (UTC)Reply
Yeah, but searching the OED is lame - you can't just type in the word and go there, you have to flick through the pages, like in an actual paper book. Roger the Rodger (talk) 20:08, 1 October 2021 (UTC)Reply
@Roger the Rodger: SGconlaw (talk) 20:47, 1 October 2021 (UTC)Reply

RFV-failed Kiwima (talk) 20:20, 31 October 2021 (UTC)Reply

To drive at a certain speed?

[edit]

What sense of do is being used in the sentence "he was doing 70 mph", meaning "he was driving 70 mph"? Do we need a new (sub)sense to cover this? 98.170.164.88 07:48, 28 August 2022 (UTC)Reply

Latin compounds of dare

[edit]
This discussion was born in a User Talk page after I added some derived terms to the Latin verb and referenced these derived terms back in their pages (abdo, addo, condo, perdo, reddo, indo, subdo). As this is the most appropriate place I move it here, hoping to make it accessible to other editors too. --Grufo (talk) 23:54, 27 January 2024 (UTC)Reply
Discussion moved from User talk:PUC#Latin compounds of dare.

Hi PUC,

Verbs like addere, subdere, etc. are regular compounds of dare, with regular vowel change due to Latin apophony of unstressed vowels. The apparent conjugation change (from -are to -ere) is due to the fact that dare is not a first-conjugation verb, but an athematic verb (the infinitive is simply -re). You should revert your changes.

P.S. Here if you want a source. --Grufo (talk) 16:54, 26 January 2024 (UTC)Reply

No. Go read de Vaan. PUC17:10, 26 January 2024 (UTC)Reply
You are making a big mistake, this is something well known (e.g. p. 123 et seqq. – p. 113 displayed – of this PDF, or p. 15 of this other one – and the sources are really many). The reason why de Vaan does not mention these compounds is that de Vaan only mentions the words for which a lemma is presented in the dictionary, but in his dictionary there is no lemma for addere, no lemma for indere, no lemma for subdere, and so on. Once again, please revert your edits. --Grufo (talk) 20:23, 26 January 2024 (UTC)Reply
P.S. I see now that de Vaan mentions a hypothetical -do verb used only for compounds. But this does not make much sense, because using dare for compounds creates exactly the regular forms that we find (addere, subdere, etc.). --Grufo (talk) 20:37, 26 January 2024 (UTC)Reply
Orbilat is not a scholarly source; de Vaan very much is.
Were it not for the latter, it still would have not been difficult to point out that the semantics of abdō (put away), condō (put together), indō (put on), obdō (put against), and subdō (put under) match those of *dʰeh₁- (put) rather than *deh₃- (give).
As for 'regularity', I should like to ask whether there exists any example of prefixation with ad- or sub- changing a verb's conjugation class, apart from (supposedly) the - verbs themselves. Nicodene (talk) 00:54, 27 January 2024 (UTC)Reply
@Nicodene: See above: there is no conjugation change, it is an athematic verb and the conjugation is in -re. Ferre is a good example: exactly like the suffix of adferre is preserved, the suffix of addere is preserved. However, differently from the root fer-, which ends with a consonant, the root of dare ends with a short vowel (da-). This vowel follows the rules of Latin apophony in compound verbs. Note that de Vaan himself lists reddere among the compounds of dare.
As for de Vaan, I am not doubting that when dare is used in compounds it can preserve another meaning later lost in the simple verb, that of “to put” (this is mentioned also in some of my sources). What I am saying is that this does not transform it into another verb: it is the very same identical verb. It is also very possible this single verb is the result of the fusion of two different PIE roots, but those had morphologically merged already when dare was used for creating compounds, because we are not able to distinguish the morphology of reddere from that of condere, and if we were asked today to create a compound of dare we would create verbs exactly like addere and condere. --Grufo (talk) 01:16, 27 January 2024 (UTC)Reply
That is simply incorrect. You need look no further than the research paper that you yourself have cited, which gives as a first-conjugation verb, an irregular one to be sure, and the - verbs as third-conjugation and regular (pp. 113–114). The question I have posed remains unanswered.
That aside, glancing through the paper makes it clear that you've not actually read it; just a bit further down (p. 119) one finds:
"Other third conjugation compounds in °, °didī, °ditum are most likely not compounds of dare/*deh₃- at all but of the root *dʰeh₁- ‘put’ [...] abdere ‘conceal’, addere ‘add’, condere ‘establish’, indere ‘apply’, perdere ‘destroy’ and subdere ‘place under’, EDLIL, s.v. -, -dere."
ETA: Even supposing we accept your view that belongs to no conjugation at all, the fact remains that a switch from 'athematic' to 'third-conjugation regular' is a change in conjugation type. The question, once again, remains.
Nicodene (talk) 01:32, 27 January 2024 (UTC)Reply
@Nicodene: … which gives dō as a first-conjugation verb It got normalized, by analogy, in some forms, to first-conjunction verbs, but saying that dare is a first-conjugation verb is simply incorrect: it is an athematic verb. You can find this information in practically all sources that discuss Latin athematic verbs.
The question I have posed remains unanswered. I would say that the answer to it remained unread. Q: “Any example of prefixation with ad- or sub- changing a verb's conjugation class, apart from (supposedly) the - verbs themselves”? (You specifically asked for compounds with ad- and sub-, but I don't see how the preposition chosen can make a difference, so I ignored that part.) A: “de Vaan himself lists reddere among the compounds of dare [not of -dere]”. The long answer is this: Latin has only these athematic verbs (as far as I know): esse, velle, ferre, ire, edere, dare; of these, the ones whose root ends with a vowels are ire, edere and dare. The vowel in ire is long, so the rules of Latin apophony say that it should remain the same (otherwise it would become -ere in compounds). The root's vowel in edere instead is short, but the apophonic change has already happened, because it is disyllabic. You can see that apophony is fully at work in edere in subjunctive forms like edim. And in any case, the rules of Latin apophony say that -ere becomes -ere (i.e. it stays the same). So there is only dare left. Strictly speaking, I remind you that there is no change of conjugation from -are to -ere in reddere or addere, there is a permanence of the same conjugation in -re.
makes it clear that you've not actually read it I believe, once again, that you don't read my answers. I specifically wrote: this is mentioned also in some of my sources. However I had also written this, which you also seem to ignore: is … the result of the fusion of two different PIE roots, but those had morphologically merged already when dare was used for creating compounds, because we are not able to distinguish the morphology of reddere from that of condere. --Grufo (talk) 03:03, 27 January 2024 (UTC)Reply
saying that dare is a first-conjugation verb is simply incorrect
'First-conjugation irregular' is the assignment made in the paper that you yourself cited; cf. also any number of Latin grammars such as Allen & Greenough (page 67) or Wheelock (page 6).
In any case that is immaterial to the question. It has already been pointed out that your 'athematic' → 'third-conjugation regular' is a change in conjugation type regardless.
I don't see how the preposition chosen can make a difference, so I ignored that part
'Prefix' not 'preposition', but quite right aside from that. If anything a full list of prefixes would make the point even more glaringly clear.
Note that de Vaan himself lists reddere among the compounds of dare
The question clearly asks for an example that is not a - verb.
no change of conjugation from -are to -ere in reddere or addere
If dare :: reddere does not represent a difference in conjugation type, you are either claiming that reddere is athematic or that dare is third-conjugation regular. Which is it going to be?
The long answer is this [...] there is only dare left
In other words the answer to the question is thus far an uncontested 'no'.
I believe, once again, that you don't read my answers. I specifically wrote: this is mentioned also in some of my sources [...]
Your very first comment claims 'verbs like addere, subdere, etc. are regular compounds of dare' while your second comment attempts to support said claim with a research paper that flat-out contradicts it a mere six pages later. That is either lying or remarkable negligence.
Likewise your claim that 'de Vaan does not mention these compounds' when in fact both addere and subdere are cited where PUC linked.
Likewise your claim that 'de Vaan only mentions the words for which a lemma is presented in the dictionary'.
There are further issues at hand, incidentally:
and if we were asked today to create a compound of dare we would create verbs exactly like addere and condere
1) If the synchronic rule is as you imply then you explain neither circumdare nor interdare.
2) You do not explain why con- dare 'to give' would synchronically make a verb meaning 'put together, found, establish'.
Nicodene (talk) 04:20, 27 January 2024 (UTC)Reply
P.S. I forgot to address this point: the fact remains that a switch from 'athematic' to 'third-conjugation regular' is a change in conjugation type. None of these compounds switch to a regular third conjugation. If you look at their perfect tense you will see that they all preserve the root doubling that is characteristic of de-di (so ad-di-di, red-di-di, con-di-di, etc.). This is highly irregular for a third-conjugation verb and almost a marker of all the compound verbs from dare. You can see the rich group of words that come from the same root of dare listed in Halsey's An Etymology of Latin and Greek (p. 95). --Grufo (talk) 03:32, 27 January 2024 (UTC)Reply
Once again you have avoided the actual question. The conjugation type of dare is clearly not that of condere; cf. for the sake of an example the subjunctive forms dēs (first-conjugation type) versus condās (third-conjugation type). Can you cite a single example that is not a -dō verb where prefixation has such an effect?
I could also add that your above claim is contradicted again by your own source (p. 114): 'which synchronically belong to the third conjugation and are regularly inflected'. The point being I suppose that they do form a consistent type.
In any case, what precisely is the point of citing Halsey? A Latin dictionary from the 1800's is not even remotely a counterpoint to modern linguists reconstructing the development from Proto-Indo-European to Latin. Admittedly it isn't as absurd as, say, citing a random .com site against de Vaan. Nicodene (talk) 04:45, 27 January 2024 (UTC)Reply
'First-conjugation irregular' is the assignment made in the paper that you yourself cited
I have nothing against that wording per se and I would probably use it if had to write a vocabulary. I would consider the wording incorrect however if I had to write about conjugations, because dare is an athematic verb. The thing however can be argued, because the first and third conjugations did have some normative powers in these verbs, and if you say “first conjugation irregular” you could mean anything, even one single “regular” suffix in the entire conjugation and all the rest irregular. It is quite an immaterial question indeed. So context does matter, and the fact that dare looks somehow like a first-conjugation verb makes the life of Latin teachers simpler. That, however, does not change what dare really is: an athematic verb.
you are either claiming that reddere is athematic or that dare is third-conjugation regular.
I don't understand why you keep ignoring the answers that I give. Of course reddere is athematic and not at all “third-conjugation regular”. An athematic verb never belongs to any of the four conjugations (although it can be partially attracted by those in some cases – see das but reddis, for instance, or des but reddas as you mentioned – see below).
In other words the answer to the question is thus far an uncontested 'no'.
Verbs don't usually change their conjugation. This include also dare, which does not changes its conjugation in compounds either (because, once again, its true class is in -re, like ferre, and definitely and literally not in -āre). So no; and the “No” extends to dare as well.
Your very first comment claims 'verbs like addere, subdere, etc. are regular compounds of dare
Indeed they are (almost) regular compounds of dare. In the sense that the result of these compounds is more or less what we expect it to be. They are not however regular verbs. Regular compound ≠ regular verb. A different example to clarify: adsum is a regular compound of sum (i.e. it is exactly how we expect it to be); whereas possum is an irregular compound of sum (see the suppletive perfect potui); both, however, are irregular verbs.
Likewise your claim that 'de Vaan does not mention these compounds' when in fact both addere and subdere are cited where PUC linked. Likewise your claim that 'de Vaan only mentions the words for which a lemma is presented in the dictionary'.
Again you don't read what I write. Why do you think there is a “post scriptum” in my first message? What concept do you think that “post scriptum” expressed?
If the synchronic rule is as you imply then you explain neither circumdare nor interdare.
Have you ever paid attention to the vowel in circumdăre and interdăre? Have you noticed anything strange? How would you pronounce these out loud? The reason why these are not *circumdere and *interdere is the same reason that prevents benefacio from being beneficio: they are later compounds, created when Latin apophony stopped being productive.
for the sake of an example the subjunctive forms dēs (first-conjugation type) versus condās (third-conjugation type). Can you cite a single example that is not a -dō verb where prefixation has such an effect?
There are no verbs that behave like dare, so no. It's like if in a discussion about sum I asked you to find another verb that behave like sum. There is none. To have the same behaviour we would need an athematic monosyllabic verb with a root that ends with a short vowel different from e. I do not know any. Do you know any other verb in -ăre? Also -ĭre, -ŏre and -ŭre would do. But not -ĕre. The reason why you see des but reddas is that the first and third conjugations did normalize some parts these two verbs. They however never fully became first- and third-conjugation verbs.
You do not explain why con- dare 'to give' would synchronically make a verb meaning 'put together, found, establish'.
You have just given me a fantastic argument (I hadn't thought about circumdare). Circumdare shows you that the verb that meant “give” and the verb that meant “put” had already fully merged when dare was used for compounds. Speakers had no memory of two verbs. Only one, dare, existed, and this meant both “give” (reddere, indere, interdare, etc.) and “put” (obdere, condere, circumdare, etc.). Then the non-compound verb lost the second meaning. With circumdare you have basically solved the case.
--Grufo (talk) 05:54, 27 January 2024 (UTC)Reply
Let's return for a moment to Allen & Greenough, page 85 of which you cited regarding athematicity.
On page 121 of the same work we find all of the following:
  • dāre listed as a first-conjugation irregular
  • condere et al. listed as third-conjugation
  • derivation of condere et al. not from dare but rather from 'the root DHA put' (=*dʰeh₁-)
For the second and third points, see also page 72.
Could I go on to point out that even the part you cited clearly explains how thematic verbs work with the example leg-i-tis, yet you could not apply same reasoning to redditis?
Could I also point out that this:
The reason why [circum-dare and inter-dare] are not *circumdere and *interdere is the same reason that prevents benefacio from being beneficio: they are later compounds, created when Latin apophony stopped being productive.
Blatantly contradicts this?
if we were asked today to create a compound of dare we would create verbs exactly like addere and condere
I could do, but it'd be gratuitous.
All this in addition to your lying twice about de Vaan and also lying by citing a paper that directly contradicts your claim that addere and subdere are compounds of dare.
Don't bother replying at this point. I'm done and have come to realise that PUC's five-word reply was precisely what nonsense like this deserves. Nicodene (talk) 08:19, 27 January 2024 (UTC)Reply
dāre
There is no dāre, only dăre (as page 121 correctly mentions – see also here)
as a first-conjugation irregular
The same authors do not mention that dare belongs to the first conjugation elsewhere. As I said earlier, it depends on the context: dictionaries usually don't mention athematic verbs, but only more or less regular verbs, trying to force them into one of the four conjugations. I am fine calling these verbs “first-conjunction irregular” and “third-conjunction irregular” for heuristic reasons, but ultimately their etymology tells a different story, which definitely needs to be presented as such in a paragraph that discusses their etymology.
yet you could not apply same reasoning to redditis?
I can, but it is probably a futile exercise: If you had given *reddătis to a proto-Latin speaker, they would have immediately changed it to reddĭtis, due to Latin apophony (the same phenomenon tha transform facio into inficio – note the difference between the later circumdătis and the regular 1st-conj. amātis). But attributing the -i- of redditis to a “normalization” of the verb within the third conjugation and saying that the vowel is a thematic vowel typical of the latter would explain it too. Which one is it then? As I said, it is a futile excercise. Furthermore, I would like to remind you once again that de Vaan himself considers reddere as a compound of dare (p. 174).
Blatantly contradicts this?
Here I don't get what you are asking.
All this in addition to your lying twice about de Vaan and also lying by citing a paper that directly contradicts your claim that addere and subdere are compounds of dare.
That is a serious accusation. If you have the strength to go through I would like to ask you to quote my words in green and say “here you lied”.
--Grufo (talk) 09:13, 27 January 2024 (UTC)Reply