Language desk
< June 29	<< May \| June \| Jul >>	July 1 >

Welcome to the Wikipedia Language Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

June 30

Irregular words — what are they made of?

Some month ago, it was mentioned that the computer mouse was given the plural form "mouses", not "mice". That is not without logic; if "mouses" is an acceptable plural, it carries additional information, i.e. "plural of mouse" + "not the animals".

However, that was dropped; I've read a lot of "mice" in a computing context: driver software, BIOS, etc. "Mouses" do creep (no pun intended) into conversation with the digital veterans but hardly anywhere else today.

Now, my "real" question: Do words tend to get irregular when used frequently? Is there any chance that web pages will some day be "lunk", not "linked"? Will they cease to be "pages", and be "pige" from then on, perhaps? (The latter are mere examples, not asked for answers. However, "VAXen" and "boxen"...) - ¡Ouch! (^{hurt me} / _{more pain}) 10:16, 30 June 2014 (UTC)[reply]

I don't see "pige" anyway, because its sound can be indistinguishable from "page", depending on dialect.~

It's not so much that words "become" irregular when they are used frequently, in the sense that they suddenly and arbitarily assume some new irregular feature, but historically they do tend to often preserve old patterns while the bulk of other words moves on through regular language change; this then will result in irregularity in some frequent words. Take the irregular verbs in English: most of them are a remnant of a once-regular, more frequent pattern (the "strong verbs"). While most other verbs that once belonged to this paradigm moved on and assimilated to the larger group of "weak verbs" (our modern "regular verbs"), the high-frequency members of the class preserved the old patterns. Across languages that have inflectional patterns, it is often found that the highest-frequency items are the ones that are most likely to have irregular forms (like, in English, the most irregular verbs are be and have). It's because of the way such words are stored and represented in our brains. Simplifying somewhat: only if a word is reasonably frequent will the brain go to the trouble of storing each of its inflection forms separately, while for less frequent words it is more efficient for the brain to just store the single form of the root and derive the inflected forms on-the-fly when they're needed. This way, frequent words can "afford" to have irregular forms, while for infrequent ones the brain will prefer applying regular rules.

In the specific instance of English computer jargon, there is probably a different factor at play though: it's just a matter of humorous linguistic creativity characteristic of computer nerds that they invent these pseudo-archaic irregular forms, like "boxen" or "Unixen". Fut.Perf. ☼ 10:56, 30 June 2014 (UTC)[reply]

So everything, with few exceptions, shifted towards s-plural and ed-passive?

And why, I wonder? Because the speakers were used to the most common irregular words, and making them regular just sounded "more wrong" than regular uncommon words?

I did a bit "research" of my own, and these are the patterns I found:

the en-plural of ox (VAX, box, UNIX, regex seem to be in here just because they are -x words)
- children might be an irregular en-plural.
the ee-plural of some words containing -oo- (wikt:goose and wikt:tooth are fairly common)
- The ee-plural of loot is definitely one that runs on "Rule Of Funny"; sometimes used in in-game chat, meaning a lot of loot.
unchanged plural. Possibly the most common non-s-plural.
the i-plural of some words containing -ou- (like mouse).
All of the above are mainly found on short words; not sure if that's a general rule or just because long words tend to be uncommon.

On verbs, I found the i/a/u pattern, like sink/sank/sunk. The most common verb pattern without -ed.

In spoken English, the i/a/u pattern often avoids the -ked ending.
wikt:prunt doesn't show up as passive of print; that would be another candidate like link because it's a very common word (depending on profession, though) and it does save a syllable. That would be my other candidate for new irregularity, but when it comes up, it's between coworkers and hardly, if ever written.
Words that do have the -ked ending, somehow sound wrong. Example: asked, linked, thanked.
- Justified with traffic. :)

An i/o/o pattern is quite noticeable, too. "Rise and shine" is a twofer here.

think avoids the -ked ending, but uses a different pattern: i/ou/ou, like find and fight.

About the very common verbs be, do, have, the defining character is that more often than not, they are verbs in syntax only; usually, they don't carry a meaning. ("Don't carry" is just the opposite of "carry"; one could say that "don't" is to verbs what "not" is to adjectives. The German language uses nicht for both. The "do" part doesn't add anything.)

To sum it up, the "irregular irregulars" I found shared the following criteria:

They followed a certain pattern; verbs seem to prefer i/a/u.
The spots must be vacant! (I.e. "think" can't go i/a/u, because "thank" and "thunk" are words in their own right.)
Easier to pronounce, or one syllable shorter than the regular form.
Long words just don't seem to do it.

Adjectives tend to get more comparable (that is, "comparabler"?), not less, when used a lot. The Rule of Short seems to be more important than everything else. "Awesomer" has replaced "more awesome" in spoken English. Definitely one of the disgustingest developments. ;) - ¡Ouch! (^{hurt me} / _{more pain}) 08:41, 1 July 2014 (UTC)[reply]

One example I've noticed - at least in spoken London English - is that the past tense of the verb 'to text' is pronounced almost indistinguishably from its present tense - the verb in 'I text him every day' sounds the same as the verb in 'I text(ed) him yesterday'; I conjecture we may have a new strong verb (using the let/let/let pattern) in the wild. AlexTiefling (talk) 12:06, 1 July 2014 (UTC)[reply]

The patterns you mention, Ouch, are not productive, but are remnants of patterns that were productive more than 2,000 years ago but have since undergone enough sound change to be only partly recognizable. For example, the vowel changes that remain for a few noun plurals, such as mice, are remnants of the Germanic umlaut, while the vowel changes in verb tenses are remnants of the more than 4,000-year old Indo-European ablaut. (See Germanic strong verb.) Some other verb irregularities, such as think / thought are slightly more recent (like 2-3,000 years old) and are the result of sound changes operating differently on different forms of frequently used but formerly regular verbs. (See Germanic weak verb.) To answer your original question, the words that are most commonly used are the ones for which different forms are most likely to survive when sound changes make them irregular. It is difficult to think of an example of this happening in the present, since these changes happen so slowly that we often aren't aware of them, though AlexTiefling's example might count (though I don't think that particular form is common in the United States). Marco polo (talk) 15:21, 1 July 2014 (UTC)[reply]

@Alex, I've heard the same formation for to text here in SW Ontario. "I will text him today; he texed me yesterday." The people I've heard it from generally have a problem with consonant clusters (the plural of breast is rendered as breast or breast-ts if they try to enunciate), so that may be at play. Matt Deres (talk) 16:29, 2 July 2014 (UTC)[reply]

Most irregular plurals in English were actually root nouns and other consonant stems in Old English; oxen is a noun of the weak declension, i. e., a noun whose stem used to end in n (Middle English still had lots of weak nouns; eye and shoe have preserved n-plurals in Scots, too, although shoe was originally a strong noun, i. e., a noun whose stem used to end in a vowel). Also, cow, which is originally a root noun too, still has an irregular plural ky(e) in Scots, which is archaic or dialectal in English, where a (now equally archaic or dialectal) n-plural kine has also arosen (double-marked for plural with the different vowel and n-ending just like brethren). The vowel alternations in these nouns are a relatively recent development, as they were not yet there in Proto-Germanic and arose in the immediate prehistory of Old English (as a trace of the former nominative plural ending *-iz).

However, the vowel alternation in the strong verbs goes back all the way to Proto-Indo-European and the origin of the Proto-Indo-European alternation, called Ablaut or apophony, is not clear.

Fun fact: The stem of the Proto-Indo-European ancestor of ox ended in n too, but the nominative singular may have lacked it, so ox – oxen could reflect a really, really old alternation. (Also, the American English plural was borrowed into some Native American languages, but in the meaning of the singular.) --Florian Blaschke (talk) 00:50, 4 July 2014 (UTC)[reply]

Thanks Marco Polo et al, I wouldn't have guessed that the verbs were that old. I thought they were about 948 years old, or maybe 12 to 13 centuries. Bummer.

Interesting read, that Germanic strong verb article. So, i/o/whatever is class 1, and i/a/u and i/ou/ou are class 3. No wonder I saw those patterns; the article confirms they are the two fairly common classes.

Some others, like the rare irregular passive of wikt:pwn, bwn, definitely runs on what Tropers call Nonsensotium; OTOH, there's not much ablauting (you know that every English noun can be verbed?) you can do without any vowels. And it's rare anyway; very few communities even recognize it as a word.

I'd suggest tuxt for both past and passive of text — that is, if I used it a lot, which I do not (I use the noun more frequently). And I don't want to attract the ire of the Linux community either. - ¡Ouch! (^{hurt me} / _{more pain}) 12:56, 4 July 2014 (UTC)[reply]

the meaning of Hollywoodism

Please teach me the meaning of "Hollywoodism" in the following passage.114.169.204.22 (talk) 11:51, 30 June 2014 (UTC)dengen[reply]

   As Barney first began to peruse Lance's initial contribution,
   he smiled at the Hollywoodisms. But by the time he had finished, 
   he felt deeply unsettled. Obviously, he had opened up Pandora's 
   box.---Erick Segal, Doctors, p.492.

It's hard to know for sure without more context, but Hollywoodisms here probably means language or narrative structures (such as plots) typically used in Hollywood films or in the U.S. corporate film industry (known as "Hollywood"). Marco polo (talk) 14:05, 30 June 2014 (UTC)[reply]

We have an article on an Iranian conference, The International Conference on Hollywoodism, which appears to be a forum for criticism of Hollywood's approach to history and politics in general, and there's a 1998 movie Hollywoodism: Jews, Movies and the American Dream, the title of which speaks for itself. "People who liked this also liked A Bit of Fry and Laurie"? _Not_ a show which I would have thought appealed to the anti-Semitic component of the movie-going public, but there's no accounting for taste. Tevildo (talk) 16:14, 30 June 2014 (UTC)[reply]

Just from that snippet, I'd go with "Hollywood cliches". Without the plural, I may have leaned toward the other sorts of -ism. But with the S, they seem like truisms instead. What was Lance's intital contribution? InedibleHulk (talk) 17:05, 30 June 2014 (UTC)[reply]

From looking a bit deeper into the context, I suspect that it's not so much cliches/truisms, as Hollywood film industry jargon. --Itinerant1 (talk) 18:27, 30 June 2014 (UTC)[reply]

Yeah, you've convinced me to suspect this, too. Terms from this lexicon or something. InedibleHulk (talk) 18:37, 30 June 2014 (UTC)[reply]

There's the infamous jargon of "Variety" magazine ("Sticks Nix Hick Pix", "legitimate BO boffo", "eye net"=CBS, "alphabet net"=ABC, "peacock net"=NBC etc. etc.)... AnonMoos (talk) 23:53, 30 June 2014 (UTC)[reply]