Talk:-icity

Definition from Wiktionary, the free dictionary
Jump to: navigation, search

RFC[edit]

TK archive icon.svg

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

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


-icity[edit]

A lot of the derived terms listed here don't look to me to have been formed with the suffix -icity, rather, -ic and -ity. — lexicógrafo | háblame — 00:32, 5 October 2010 (UTC)

Is there a difference? I mean, if some of them ended in /ˈɪ.kɪ.ti/ I would see the problem, but as it is, it seems that -icity is the suffix you get when you add -ity to a word ending in -ic, no? —RuakhTALK 00:54, 5 October 2010 (UTC)
It would help if we had some semantic content to the "definition". Several of our affix entries have no semantic content whatsoever, [[-ity]] being an example. I don't think the facts usually justify that, certainly not "-ity", which most define as forming nouns of "nature or condition or state, quality or degree" from adjectives. I would be fascinated to hear the semantics of -icity.
This does illustrate a problem with our poor conceptualization of what we mean by derivation in general and affixation in particular.
  1. Is derivation supposed to be a historic process or a morphological one?
  2. For words formed in the same language do we have different rules than for cross-language derivation?
  3. To understand historical derivation of words formed by affixation we would have to do a great deal of research using OED and our own corpus research as most other dictionaries don't bother.
If we take a historical approach to this we would need to investigate each word to determine whether the word ending in -ic preceded the word ending in -icity. If we found some attested ones or numerous nonce coinages illustrating productivity we could justify an entry.
If we take a morphological approach maybe we could direct user to the components for the "normal formation". DCDuring TALK 03:06, 5 October 2010 (UTC)
I guess the thing is … the suffix -ity, when appended to a pre-existing word, does three things:
  • it moves the stress to the last syllable, then appends /ɪ.ti/.
  • together with the stress shift, it causes a whole bunch of random vowel changes that only make sense in terms of historical forms and etyma. (Think of "real" vs. "reality", for example, where the stress gets shifted into the second half of an normally monosyllabic word; or "urbane" vs. "urbanity", "humane" vs. "humanity", "-ose" vs. "-osity".)
  • it combines with various other suffixes in specific ways. ("-ose" + "-ity" = "-osity", "-ous" + "-ity" = "-ity" or sometimes "-osity", "-ic" + "-ity" = "-icity".)
In a sense, even if a word in "-ic" predates a word in "-icity", the derivation is nonetheless kind of [word minus -ic] + -icity; for example, "toxicity" is formed from "toxic" by dropping /ɪk/ and adding /ˈɪsɪti/. Because it's not possible to add -ity to a word in -ic; when you try, you get -icity.
There are surely words in -icity that don't come from words in -ic (for example, multiplicity comes directly from French, multiplic being quite rare), but I'm not sure if there are words that come from -icity directly without coming from -ic. If so, then they should probably have their own special table, but that doesn't mean the rest of those words should be removed from the entry.
RuakhTALK 11:36, 5 October 2010 (UTC)
You could certainly look at it that way. I almost always ignore pronunciation considerations, probably wrongly. Partridge's Origins, which has a good section on affixes, has an entry for it, too, but not Quinlon's "Affixes" web resource.
OTOH, affixes also can change the pronunciation and even the spelling of bases. Origins also has -acity.
I would venture that all or almost all the derived terms are historically either derived from English (some vintage) words ending in -ic or (-ice ?) or borrowed and respelled, especially from French (some vintage). In the source language, or farther back in derivation, such endings are always sequential, AFAICT (eg, [la] teneo > tenax (tenac-) > tenacitas > [fro] tenacite > tenacity <??? ten + -acity). To me taking a synchronic view of such probably-never-productive-in-English units seems highly questionable.
But, maybe we should encourage the creation and use of such compound suffix entries to shorten the largely unusable lists produced by {{suffixsee}} when applied at [[-ity]]. DCDuring TALK 12:49, 5 October 2010 (UTC)
Move to rfv or rfd. Mglovesfun (talk) 22:58, 7 October 2010 (UTC)

moved to rfv -- Prince Kassad 15:18, 13 February 2011 (UTC)

RFV[edit]

TK archive icon.svg

The following discussion has been moved from Wiktionary:Requests for verification.

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


There seems to be a dispute on whether this really is a formative English suffix. Most terms seem to be either -ic + -ity, or directly borrowed from a foreign language. Cites would be helpful in determining whether this really exists. -- Prince Kassad 15:18, 13 February 2011 (UTC)

Keep. Don't let the spelling deceive you: the fact that /-ɪk/ + /-ɪti/ = /-ɪsɪti/ should be documented at [[-icity]], no matter how sum-of-orthographic-parts it might be. (It should also be documented at [[-ity]], and maybe also at [[-ic]], but the ones don't preclude the other.) —RuakhTALK 18:02, 13 February 2011 (UTC)
That's dangerous ground isn't it? We deleted -cede a few months ago on the grounds it didn't exist, but you can still pronounce it. Mglovesfun (talk) 11:56, 14 February 2011 (UTC)
With -cede the complaint seems to have been that it doesn't form words in English, and doesn't even really have an expressible meaning in English. With -icity, however, nom says that it does form words in English, by adding -ity to words in -ic. My point is that that's misleading: -ic + -ity = -icity looks reasonable, until you try pronouncing the three forms and realize that that's not how it works. —RuakhTALK 12:26, 14 February 2011 (UTC)
I tend to think this might be attestable anyway, but I don't think we should keep it whether citable or not. Mglovesfun (talk) 12:45, 14 February 2011 (UTC)
How is the stem pronunciation alteration evidence in favor of keeping this? It seems to me that many suffixations may be accompanied by pronunciation changes to the stem, ranging from altered stress, through vowel changes and orthographic changes (eg, consonant doubling), to consonant pronunciation changes. It seems reasonably clear that the previous existence of an adjective ending in "ic" is required for the vast majority of the formations ending in "icity".
As it stands the entry is deceptive implying productivity that may not be warranted. It is relatively easy to populate a "related terms" or, worse, "derived terms" section by copying from OneLook all words ending in "icity". The value that Wiktionary should offer is making the distinction between actual derivation and coincidental spelling similarity. I would suggest a comparison of the current entry for -arian and earlier versions that did not exclude Tocharian and centenarian from the small number of legitimate formations.
It seems to me that this a clear case of the desirability [licensed by "desirable"] of a hard redirect, this time to the legitimate suffix -ity, unless there is actual evidence of derivativations directly from -icity. DCDuring TALK 16:09, 14 February 2011 (UTC)
It's true that many suffixes entail pronunciation changes in the stem; in fact, -ity itself induces a stress shift, together with associated vowel changes. But there's no general rule that it changes a preceding /k/ to a /s/; that's a specific property of -ic + -ity-icity. —RuakhTALK 01:23, 16 February 2011 (UTC)
I have tried to find examples of words that end in Citations:-icity that do not end in -ic. - -sche 23:41, 15 February 2011 (UTC)
I read some months ago a book on morphology that made the point that the best evidence that an affix was productive in a period was the existence of a number of unattested (possibly unattestable), but intelligible words in that period. It seems dumb that the criteria we seem to apply (three attested derivatives) would lead us to exclude currently productive affixes. What number of recent formations others would find convincing? Perhaps 6-10? Nine total cites for unattestable formations would arguably be the citation equivalent of three attestable formations.
-icity might well meet that standard. COCA and OneLook have searches that support wildcards. OneLook search captures words that don't seem to be included in normal search. DCDuring TALK 00:15, 16 February 2011 (UTC)
For the record, I think there are a lot of problems with the "it's not an affix unless there are citeable words that can only be explained by its having been added directly, in Modern English, to an pre-existing stem" approach. I guess the idea is that affix entries aren't actually worthwhile in and of themselves, but rather are only useful insofar as other entries' etymology sections can link to them? But personally I think that affix entries are worthwhile in and of themselves, and am getting rather annoyed with the frequent attempts to enforce the contrary view by making demands at WT:RFV that aren't actually supported by any policies. —RuakhTALK 01:23, 16 February 2011 (UTC)
I was arguing in favor of those without attestable derivatives, but with evidence of a derivation process. Without some kind of evidence how would one know that an affix is not a figment of one's imagination? DCDuring TALK 02:15, 16 February 2011 (UTC)
You were describing one problem with it. My complaint is that there are a whole raft of problems with it, not just one. As for evidence — I think evidence is great. I just think there's more than one kind of evidence. You'll notice that my "keep" vote was based on evidence: -icity has a specific pronunciation that's not explainable as just -ic + -ity. —RuakhTALK 03:03, 16 February 2011 (UTC)
How does the fact that adding "-ity" to a word ending in "ic" convert the hard "c" to soft "c" by itself make "-icity" a morphological unit? Does the fact that "frolic" becomes "frolicking" make "-king" either a suffix or an inflectional ending? DCDuring TALK 05:00, 16 February 2011 (UTC)
That principle is what my objection is based on. Mglovesfun (talk) 11:21, 17 February 2011 (UTC)
Re: "word ending in 'ic'": You mean "word ending in -ic". Re: "morphological unit": Did I use that term? Re: "-king": It means we should have an entry for the sum of the two parts, i.e. [[frolicking]]. Which we do have, so I'm not sure why you're objecting. —RuakhTALK 12:21, 17 February 2011 (UTC)
An argument analogous to yours would justify an entry for -king. DCDuring TALK 14:09, 17 February 2011 (UTC)
Sorry, I don't see it. Please make the argument. —RuakhTALK 14:14, 17 February 2011 (UTC)
Another note ci- is almost always pronounced /si/ in English. Mglovesfun (talk) 14:57, 17 February 2011 (UTC)
My complaint with -icity is the same as the one for -cede. The whole RFV is asking does this form words in English. Your argument seems to be we should keep it regardless of attestation because the pronunciation is interesting - which I strongly dispute. I don't like the idea of keeping things whether they are attested or not - it rather defeats the object of having a page dedicated to verification. Mglovesfun (talk) 14:59, 17 February 2011 (UTC)
Citations:-icity would seem to justify a pass anyway. Almost any non-Romance word that can be attested suffixed with -icity would probably justify a keep. Mglovesfun (talk) 15:03, 17 February 2011 (UTC)
Re: "Your argument seems to be we should keep it regardless of attestation because the pronunciation is interesting": Not at all. My argument has two parts: (1) it's obviously incredibly well attested, as everyone knows who speaks English, and as the nominator acknowledged; and (2) we should keep it because the pronunciation shows that it's not SOP. Because of quirks of this spelling system, it looks SOP, but looks, as we know, can be deceiving. English does not have any sort of regular alternation between /s/ and /k/. It happens (due to the history of the Romance languages) that in this spelling system, both sounds are sometimes represented <c>, with rules for when that's possible and when it's not (e.g., <ci> is (almost?) never pronounced /kɪ/), but that shouldn't trick us into thinking there's some underlying archiphoneme /C/ that's sometimes realized as /s/ and sometimes as /k/. —RuakhTALK 15:14, 17 February 2011 (UTC)
The nominator says nothing of the kind here, by my reading.
AFAICT, arguments for inclusion could be:
  1. that -icity is or has been productive (diachrionic) or
  2. that we keep any combination of morphemes that people might imagine as being productive because they are "really" reinventing the terms that end in "icity" rather than pulling them whole from their mental lexicon (synchronic).
I advocate the diachronic line and would prefer that we only accept such evidence. I am not sure what would constitute evidence along the synchronic line that would be sufficient and effectively available to us. I suppose that the mere fact that someone creates an entry for -icity constitutes evidence that at least one person views it as meaningful morphological unit in their idiolect. Perhaps such evidence is sufficient. Or we could have native speakers, or experts, or anyone so willing express themselves on the topic, followed by a vote. DCDuring TALK 16:07, 17 February 2011 (UTC)
The nominator says that "Most terms seem to be either -ic + -ity, or [] ". This statement clearly presupposes that there are words in -icity.
Your two arguments for inclusion form a false dilemma. They both assume that a suffix should only be included if it's productive, and differ only in the evidence required to demonstrate productivity. I reject the assumption, so lie outside of your evidence spectrum.
At heart, I think y'all are making a sum-of-parts argument ("this is just -ic + -ity") couched as an attestation argument ("we need to attest this as something other than just -ic + -ity"). I, however, am not bothered by the fact that this is formed from -ic + -ity, any more than I'm bothered by the fact that hot dog is formed from hot + dog.
If there is community consensus for your view, and we can objectively codify it, then we can start using WT:RFV to enforce it. But so far these WT:RFV listings make it seem like y'all are trying to bluff your way past a lack of demonstrated consensus.
RuakhTALK 16:25, 17 February 2011 (UTC)
OK. PK acknowledged a possibility, which may account for why he brought the matter for verification. AFAICT the nom statement does not acknowledge a fact.
What is an affix if it is not or has not been productive?
I acknowledge that I didn't make explicit allowance for the completely evidence-free approach.
I understand that there is a Mexican proverb to the effect that one "makes a path by walking it". I think that's what we're doing. Are we ready for the dynamite and earth-movers? DCDuring TALK 17:45, 17 February 2011 (UTC)
Re: "What is an affix if it is not or has not been productive?": An affix.
Re: "I acknowledge that I didn't make explicit allowance for the completely evidence-free approach": You're indenting it as though it's in reply to my comment, but it actually seems to be studiously avoiding replying to my comment. ;-)
RuakhTALK 18:11, 17 February 2011 (UTC)
An affix is a morpheme that is involved in affixation. Affixation would seem to be a process that is either part of the history of a language/dialect/etc as a whole (diachronic) or of an individual's learning and use of a language (sychronic). I don't see other real-world possibilities. Do you?
You haven't advanced a rationale for why the phonetic phenomenon should be counted as evidence. Is there one?
I have tried to answer your points as they apply to my arguments as best I can. What have I missed? DCDuring TALK 18:23, 17 February 2011 (UTC)
On an unrelated note, I think that "Spanishicity", "Frenchicity", "Dublinicity" and "Parisicity" support "-icity" as different from "-ic"+"-ity", because there is no word "Frenchic" (though misscans of "Frenchie"), "Spanishic", "Dublinic". - -sche 19:18, 17 February 2011 (UTC)
I'm impressed; but, playing Devil's advocate for a moment, I'm not sure how you would show that. Even if Dublinic weren't otherwise attested (it actually is, but apparently not durably archived), that wouldn't necessarily mean this wasn't {Dublinic}ity. If we prefer to see the /k/-to-/s/ transformation as belonging to -ity, as DCDuring apparently does, or to the following <i>, as Mglovesfun apparently does, then a nonce occurrence of -icity could just as well be seen as a nonce occurrence of -ic together with a nonce occurrence of -ity. Analogously: if a word that ends in -hood exists only in plural, should we conclude that -hoods is not just -hood + -s? This is why I think quote-unquote "attestation" is the wrong approach. Certainly we shouldn't include affixes that literally aren't attested, but nearly every affix that's ever been listed on this page is actually in clearly widespread use. —RuakhTALK 19:42, 17 February 2011 (UTC)
Cool. That is the kind of evidence that would probably satisfy Ingo Plag, the author of Morphological Productivity: Structural Constraints in English Derivation (Topics in English Linguistics, 28) (1999). I read his Word Formation in English. I have no idea what he would actually say about our discussion or about "-icity", but all the novel formations fit into his suggestion that forms not found in any dictionaries are the best evidence for productivity of affixes. DCDuring TALK 20:19, 17 February 2011 (UTC)
My two cents: I agree with Ruakh that "-icity" should be an entry (or, if it were unattested, at least a redirect to -ity or -ic), so that in that entry (respectively in -ity) the pronunciation can be explained. I also agree with DCDuring that "whoever says A must say B" and if "-icity" should be an entry, then "-king" should also be an entry — or at least a redirect to -ing or -ic, so that in the entry the phenomenon "traffic" → "trafficking" can be explained. I understand Ruakh's point that nonce "Spanishicity" could be only nonce "Spanishic+ity" instead of "Spanish-icity", but I think the 1985 example is then good, because its "Germanicity" is distinctly difficult to interpret as "Germanic-ity". - -sche 22:18, 17 February 2011 (UTC)
To clarify my opinions in a few respects: (1) I'd be pretty O.K. with "-icity" being a redirect to "-ity". I think I'd prefer a full entry, but I'm not sure; and anyway I'd accept a redirect if that's more palatable to the editors who don't want an entry. (2) I'm actually not strongly opposed to having an entry for "-king", but I nonetheless see it as completely different and unrelated to the question of "-icity". I actually don't even begin to see the connection. "-king" isn't part of English per se, only an artifact of English orthography; what's happening is that the sound /k/ is being spelled <c> in some forms and <ck> in others. This gives the illusion that "-king" is an inflectional ending, but it's just an illusion. By contrast, "-icity" is a well-attested combination of two affixes, clearly an actual part of the language (whether or not you're willing to describe it as an "affix"), and the question is whether this part of the language warrants its own entry. My reason for saying it does is that it has a unique and unpredictable pronunciation. (As Mglovesfun says, the pronunciation happens to be apparent from the spelling; but it's not a general property of English that affixes are immutable strings of letters whose pronunciations are determined by spelling. If it were, the gerund-participles of "panic" and "hop" and "hope" would be "panicing" with an /s/, "hoping" with an /oʊ/, and "hopeing" with your-guess-is-as-good-as-mine.) —RuakhTALK 02:36, 18 February 2011 (UTC)
I have added a few references to the entry -icity. Also, I suggest, that whether we have -icity or -king or neither, we should have notes about -icity (pronunciation) and -king (added k) at -ic (note about pronunciation and k) and at -ity (note about pronunciation) and at -ing (note about k). - -sche 20:58, 18 February 2011 (UTC)
Passed as clearly in widespread used; feel free to submit a RFD, but I suspect it will be kept for no consensus. - -sche (discuss) 11:29, 2 August 2011 (UTC)