Talk:-otomy

-otomy
This should IMHO be deleted as redundant to -tomy. All the items can be analyzed as Xo- + -tomy or X + -o- + -tomy if need be. For instance, can be analyzed as  +  or aden- + -o- + -tomy if need be. To place e.g. metrotomy to multiple suffix categories, one for -tomy and one for -otomy, seems to create avoidable redundancy. See deleted Talk:-oscopy for a similar treatment. We have no -oplasty. A check in Category:English suffixes shows we do not have this all that often, relative to the total number of suffixes we could treat like that. -ocracy, and -ology are some examples of what we do have. Here, again, czarocracy can be analyzed as czar + -o- + -cracy, with no need of -ocracy suffix.

An alternative to this RFD would be to make it a matter of policy, but we have some precedent so let us see whether there is support for deletion here. --Dan Polansky (talk) 07:57, 2 October 2022 (UTC)


 * Delete. More examples of suffixes -oX that can be handled as just -X or if need be -o- + -X: ; ; ; ; ; ; ; ; ; ; ; ; . --Lambiam 12:55, 2 October 2022 (UTC)
 * The proposed treatment is in keeping with WT:MWO: It has no -ometry, -ostomy, -otomy, -onomy, and -ology. But it has -ocracy. It has -onymy, traced by Wiktionary to ὄνομα, so the -o- is not the interfix but rather part of the etymon. It has -osis. By contrast WT:OED has no -otomy, but it has -ocracy, -ology, -ologist, -ological, -olol, -ometer, and -ometry, all as combining forms. It has -osis. I propose to follow MWO for minimalist treatment. -onym is a special case, etymologically. Why MWO has -ocracy I don't know. We can investigate whether -osis is a special case as well. We can expand our suffix entries with notes that some derivations are sometimes analyzed as containing an -oX suffix but that we chose to analyze it as -o- + -X. That should do and help up avoid double suffix entries and double categories. --Dan Polansky (talk) 11:37, 16 October 2022 (UTC)
 * I added sources to our entry. MW has speedometer as an example using -o-; AHD has acidophilic. We analyze both entries using -o-, not using -ometer or -ophilic. If we keep -otomy, some editors will invariably keep using it in etymologies, resulting in inconsistent treatment. As for whether -o- is a morpheme, that does not seem decisive: it is a "linking element". Perhaps it is "speedo-" and "acido-" that are the morphemes; it is not clear why "-ometer" and "-ophilic" should be more of candidates for morphemehood than "speedo-" and "acido-". Category:English terms interfixed with -o- has over 1,400 entries; many cases of similar analysis/etymology can be found there. However, some should perhaps be analyzed using Xo- combining forms: archaeography could be analyzed using archaeo-. --Dan Polansky (talk) 06:25, 20 October 2022 (UTC)
 * The entry for even admits that it is not a morpheme, and you are still basing your argument on how our categorisation structure works, which is wrong. As for why it doesn't attach to  or, that's because they aren't prefixes and don't behave like them. Basic stuff. Theknightwho (talk) 06:34, 20 October 2022 (UTC)
 * Our entry on -o- is no authority and does not even say "not a morpheme". Furthermore, so what if -o- is not a morpheme but perhaps a "morph" or "linking vowel"? It is an element of analysis. Suffix -y attaches to "skin" to create "skinny", and similarly, -o- could attach to "speed" to create "speedo-". Affixes do attach to free morphemes. Linking elements -n- and -o- are accepted in German, Czech and other Slavic compounding etymologies, avoiding the need to create combining form entries such as Wolken- for Wolkenkratzer, Bundes- for Bundestag or modro- for modrooký. Thus, modrooký = modrý + -o- + oko + -ý. What is to be avoided is duplication, not only of categories but also of suffix entries. The minimalist approach is well enough sourced to be a linguistically acceptable option. Is the proposal here that -o- should not be used in our English etymologies? --Dan Polansky (talk) 10:17, 20 October 2022 (UTC)
 * It’s not relevant that it doesn’t use the exact words “not a morpheme”, as it says it’s inserted between morphemes - a completely unnecessary statement for anything that is itself a morpheme. You’ve also provided nothing to suggest it is a morpheme, so we cannot assume that it is one, either. This is relevant, because it determines whether the suffixes that include it are alternative forms; evidently, they are.
 * The idea of avoiding duplication in alternative forms also goes against our approach everywhere else on the site, as you very well know. Given your heavy focus on categorisation, I can only conclude that this is yet another misguided attempt to sweep anything awkward under the rug when it goes against your own over-hasty analysis and attempts to pigeonhole everything based on whatever your latest category obsession is. That is a terrible approach to building a dictionary. Theknightwho (talk) 14:04, 20 October 2022 (UTC)
 * (outdent) 1) Search for finds linguistic sources using the term to refer to linking elements. They do think linking elements are morphemes. 2) The claim that it is not a morpheme is not backed up by any linguistic source so far; it is based on an uncertain inference from wording. To describe an interfix as connecting two morphemes is natural (it follows from its definition) and does not really imply it is not a morpheme. 3) If it is not a morpheme but rather a glue-like element used to link morphemes, then like glue it does not attach with priority to one of the connected surfaces. 4) If we accept "inserted interconsonantally between two morphemes", this does not imply -o- creates -oX but rather that it is inserted between X and Y to create words. This description does not imply -ometer as an intermediate product. 5) Spellings like cool-o-meter are a hint at this kind of analysis. There is no need to create -o-meter alongside -ometer to account for cool-o-meter and clap-o-meter. 6) I propose to keep analyzing speedometer as speed + -o- + -meter rather than changing it to speed + -ometer. To make sure this analysis is consistently applied, having no entry for -ometer is the most practical option: when there is no -ometer, it is not available in etymologies. I propose to follow a) Merriam-Webster, b) Wiktionary current predominant practice, c) the result of Talk:-oscopy, d) even OED to a large extent (see next point). 7) A minimalist approach is practical and would serve the readers well. Looking just at the neo-classic compounds starting with "acido-" that we have and that are in OED, we would need -ocyte, -ogenesis, -ogenic, -oleous, -olysis, -ophil, -ophile, -ophilia, -ophilic, -ophilous, -ophyte and -opore (-ophilic is the only bluelink, created in 2018, used in zero etymologies). Not even WT:OED has any of the forms; it has e.g. -genic. OED on "acidogenic": "acid n. + -o- connective + -genic comb. form." So even OED does not consistently and fully play the -oX game (it has -ology). My proposal: let's take the practice that we follow in all but a few cases and apply it consistently. Let's go by the Occam's razor heuristic. Let's not invent (a + (b + c)) when (a + b + c) is fine and does not prefer the former over ((a + b) + c). Let's not create a plethora of -oX forms that we never had and most of which are absent from most dictionaries, e.g. per  vs. . --Dan Polansky (talk) 07:11, 21 October 2022 (UTC)
 * Calling it "uncertain inference from wording" only makes sense if you're intentionally refusing to understand why I don't think it's a morpheme, as you've failed to engage with my explanation. 3 is completely false, because it ignores the option to simply model the two variants as alternative forms that depend on whether the stem ends in a consonant or vowel; you cannot argue that there are two variants of the stem, however (and examples like are not relevant, as they are prefixes where my argument also applies). The existence of terms like  is also completely irrelevant, because (a) that only exists with the suffix, and (b) is riffing off the English word . Just because you don't like the fact that  doesn't neatly fit into your model doesn't mean that you have to model the existence of a semantically irrelevant morhpeme in order to explain it. Much simpler to take the usual approach of noting that the suffixes often have phonetic variants that depend on the final consonant of the stem. An approach that is, in fact, a lot more common in linguistics than yours.
 * As for "taking a minimalist approach", you've failed to explain why we should only take that approach here, while we don't anywhere else on the dictionary. As someone who is usually highly inclusionist, it is one of the more glaring examples of the way you will argue totally contradictionary positions depending on what you want in any given moment (such as trying to brush any awkwardness under the rug to make categorisation easier while using the WT:LEMMING argument to argue for exclusion; the latter of which is something that you explicitly said you should not do on more than one occasion). Calling my approach (a + (b + c)) is also a pretty obvious misrepresentation of what I'm saying, too. Theknightwho (talk) 09:27, 21 October 2022 (UTC)
 * Maybe you could provide some external sources to support the notion that -o- is not a morpheme and that the -oX approach is more common or "usual" in linguistics. I provided sources for existence of -o- as an element of analysis (multiple dictionaries including MW and OED), sources lacking most -oX forms while having -X forms (multiple dictionaries including MW and OED), and anyone can search for "linking morpheme" to easily find academic sources online. I find my analysis compelling and well backed by sources. --Dan Polansky (talk) 10:03, 21 October 2022 (UTC)
 * Perhaps you could engage with the main substance of my point instead of fixating on the word "morpheme"? It just comes off as though you've only read the first sentence of my reply, because you haven't addressed my main argument(s) at all.
 * Continuing to use WT:LEMMING to argue for exclusion, despite explicitly saying that you would not and should not do that in the past, is not a good argument by the way. Theknightwho (talk) 10:15, 21 October 2022 (UTC)
 * Again, as for "An approach that is, in fact, a lot more common in linguistics than yours": if that is true, it should be easy to provide at least one external linguistic source. I am eager to learn more from external sources, to broaden the perspective. --Dan Polansky (talk) 13:12, 21 October 2022 (UTC)
 * Again, you're avoiding the substance of my argument. It's intellectually dishonest. I also said that modelling suffixes as having variants is more common than inferring the existence of a link - it was a general point you've not only hyper-fixated on for disingenuous reasons, but obviously misinterpreted in order to feel like you've "won". Egotistical nonsense. Theknightwho (talk) 14:56, 21 October 2022 (UTC)
 * Why not share with us your sources for common enlightenment? --Dan Polansky (talk) 15:34, 21 October 2022 (UTC)
 * You seriously want me to source the use of with suffixes, and how it doesn't involve the creation of a morpheme? That Wikipedia article should help.
 * Go on then - please explain what the morpheme means in . If it's a morpheme, it must have some kind of semantic value. By the way: the fact that linking morphemes exist does not inherently mean that  is acting as one. We even define it as an interfix, too, and interfixes are not morphemes (unlike infixes). Theknightwho (talk) 16:07, 21 October 2022 (UTC)
 * I don't trust Wikipedia. When a Wikipedia article is well sourced, it is possible to trace a statement to the sources it traces to. What needs external sourcing, not Wikipedia, are the claims that are subject to disagreement, e.g. "interfix is not a morpheme" or "the -otomy analysis is more common in linguistic sources". As for semantics of morphemes, cranberry morphemes have no known meaning. I played the sourcing game by tracing -o- existence to multiple external sources, by tracing interfix to multiple external sources, by creating empty morph and tracing it to multiple external sources, by pointing out to "linking morpheme", where the tracing to external sources is available in Google search. There is empty morpheme (semantics-free morpheme), but I found only two sources to trace it to. I will quote one of my sources, boldface mine: "Interfixes (also  called  linking  elements  in  English  and Fugenmorpheme  in  German)  refer  to  the  phonetic  material  some-times  found  in  compound  words  at  the  constituent  boundary.". The source may be wrong; I don't know. If you give us an external source claiming that an interfix is not a morpheme, we will have a more complete picture of what sources are saying. Without it, we have your claim against the claim of multiple sources found by searching for "linking morpheme", and as far as I am concerned, the multitude of external sources win. I may well be wrong, the sources may be wrong, you may be right, but that's the sourcing game. --Dan Polansky (talk) 10:51, 29 October 2022 (UTC)
 * You have seriously (and I think disingenuously) misrepresented what a is, which is a morpheme that has an opaque meaning to speakers because it exists in fossilized constructions. You also don't seem to understand that  and  are not synonyms.
 * The concept of an empty morpheme is also controversial, and it also doesn't appear to be relevant here given that we can explain the presence of an here as being a purely phonetic element.
 * Given that Wikipedia has plenty of sources, as you can already see, and the fact that you are calling a single paper that you've pulled from Google a "multitude of external sources", I don't think there's much point in continuing this discussion, as you are clearly not engaging in good faith. Theknightwho (talk) 11:39, 6 November 2022 (UTC)
 * Sure, let's talk about me. Good idea. Now, let's try something different. Should Wiktionary be allowed to mark up Czech word as  +  +  in its etymology? If so, does it mean that Czech etymologies are allowed to use, although it has no semantics and therefore is, allegedly, not a morpheme? --Dan Polansky (talk) 10:57, 4 January 2023 (UTC)
 * Keep. It is spurious to analyse the intermediary "o" as a morpheme in its own right. This is clearly an epenthetic alternative form. To delete because of categorisation issues is also totally wrongheaded - that problem is obviously possible to solve in other ways, and is an issue that exists in many languages. Plus, the idea of regulating content based on how well it conforms to our current categorisation system is the opposite of how we should be approaching things, and therefore something I will never support. Theknightwho (talk) 22:26, 19 October 2022 (UTC)
 * Keep per the above. AG202 (talk) 13:23, 21 October 2022 (UTC)
 * Keep per above. MedK1 (talk) 21:54, 28 August 2023 (UTC)
 * Delete per proponent. PUC – 11:09, 23 October 2022 (UTC)
 * Keep per the above. AG202 (talk) 13:23, 21 October 2022 (UTC)
 * Keep per above. MedK1 (talk) 21:54, 28 August 2023 (UTC)
 * Delete per proponent. PUC – 11:09, 23 October 2022 (UTC)


 * A question for the RFD participants on the fence: should English be forbidden from English etymologies/morphologies of neo-classical compounds going forward? --Dan Polansky (talk) 19:53, 30 January 2023 (UTC)

Kept as no consensus, 3 keeps vs. 3 deletes. (I would abstain.) However I note that Dan basically executed the deletion anyway by emptying Category:English terms suffixed with -otomy (the one entry there now was added by Kiwima months afterwards), you may want to review entries in Category:English terms suffixed with -tomy to see which should be where. —Al-Muqanna المقنع (talk) 00:00, 10 September 2023 (UTC)


 * @Dan Polansky Colour me shocked that Dan Polansky was only a stickler for rules when convenient. We should mass revert this. Theknightwho (talk) 00:22, 10 September 2023 (UTC)


 * Compare Grease_pit/2023/July and the various kinds of "semi-distinct" suffixes discussed there. I hope we eventually decide on and implement a general approach (rather than discussing isolated affixes, e.g. here we're ruling on -(o)tomy but not the other terms Lambiam mentions like -(o)nomy), regarding what form of the affix the templates should display (presumably whatever the 'surface' form is) vs link to (maybe the lemma form instead), and which form(s) we should categorize it by. - -sche (discuss) 18:52, 10 September 2023 (UTC)