De Vulgari Regularitate (earlier: substratums)
- One clarification of my point before I proceed, in case some people missed it: the whole point is, why Latin plurals (and cases, and conjugations, and gender and deponents and what not) in English as they were used in Latin? What we emprunt are the words (roots, pre- and suffixes.) Everything else done on them is using English (or modern) devices.
I didn't mean to discuss only some stray cases such as virus and substratum.
> "Don't we?" Do you mean ordinary speakers or linguists pressed to give gramatical advice?It's hard to say what makes the former prefer this or that form. If functional simplicity were always given precedence, we'd have given up "sheep" and "geese" in favour of "sheeps" and "gooses" a long time ago. Aren't the old forms stupid? If we managed to get rid of them, they'd soon be forgotten like "kye", "beech" and "een" for "cows", "books" and "eyes".
Yes, one may equally stand against the Old English plural relics, as in sheep and geese.
Whether OE or Latin, or some other language, these irregularities can be done away with.
> My theory is that speakers are normally conservative and try to conform to the norm (if they know how) rather than try to "repair" their language so that it will be simpler and more logical. If children say "sheeps", we correct them rather than applaud their rationality. As a result, the norm evolves rather slowly, as if _against_ our best efforts to perpetuate it. You may say "substratums" as a matter of principle, but you wouldn't do so at a time when such forms still attracted social stigma. Who wants to sound ignorant and uneducated?[Ash]
But since linguists are (hopefully) not fraught with the risk of attracting censure, or at least, being labeled ignorant, can we hope there would be some push to natural usage from their quarters?
English, of course, is not academie-regulated (and I am not suggesting it,) It functions rather democratically, if I can say so. Demoglottally?? If everyone finds it odd to change, who will bell the cat? Why not people of this category, the linguists et al. (Only don't do it while writing an English exam, but I hope most folks here are past that stage ;-)
Is it expecting too much? Or being too impractical by hoping or urging for something like this?
Of course only an ignoramus (and a snob at that) would attempt to use unetymological plurals like "ignorami" or "octopi". I prefer "cactuses" and "formulas" to "cacti" and "formulae", but I won't say or write "phenomenons". Perhaps I'm inconsistent, but so is everybody else in one respect or another :).
Even 'formulas' from 'formulae' and the kind were a break from existing norm. Now my question is, from a linguistic standpoint, leaving political and other forces aside, what can possibly be a positive prod to encourage this regularizaiton? I have special hopes on English: since it has given up such a huge bulk of its older devices, I am inclined to place a little more hope on it for all its potential for tradition-ridding.
> I thought I had the answer to the question as to why we don'tsee the Latin plural of "virus" used -- that it is fourth declension
so the nom. pl. is spelled like the nom. sing. But wrong again!
Checking my Allen & Greenough
<snip snip snip>
> To sum up, since <vi:ri:> is not even an authentic Latin plural, everyone should feel free to regularise it in English.[Ash]
This is exactly what I expected. Why take recourse to Latin or other language originals, when there is a simpler way, just -s every word to plural it! Isn't the usage of Latin plurals the result of a prudish attitude?
One more thing on "substratums." I personally find the "mz" ending a
little difficult to say. Is this an idiosyncracy, or could it be general
among English speakers and influence the decision to retain the Latin plural?
Aw c'mon! When you can say bums, customs, poppadums and conundrums, why do you recourse to Latin for some words? The difficulty or the sense of it, I suspect, stems not from some assurance somewhere that there is a Latin plural out there for the taking, but perhaps from the need to explain it now. But objectively, aren't all these just as pleasing or awkward, just as easy or difficult? How about the tons of -isms in the language, aren't they worse by this count?
I would endorse such "urplurals" on just one ground: that form itself confers a special sense to it irrespective of the presence of the singular form either in the "empruntee" (original language) or in the "emprunter" (borrowing language.)
Such words are fine, because it can be supposed that they are in fact borrowed in their plural form directly, such as data and literati (even when the singular exists with a slightly or largely different sense.) This is in principle the same as emprunting a verb form and using it as a noun (ignoramus, exit) or any such cross-functional usage.
PS: In the lighter vein...
A friend of mine (who's from India) was telling me about this incident that took place in a chemistry class:
The teacher enters the class, walks up to the blackboard, and writes "Formulae." He turns to the students to start his lecture, but in a moment turns back. Shaking his head, he erases the word, and writes "Formulas." Again after a pause, not satisfied with himself, he effaces that word too, and finally confidently writes.. "Formulaes."
To digress, there is some interesting discussion taking place regarding IE origins a Germanic-L@yahoogroups. You might want to take a peek at it.
- --- In cybalist@y..., "Ash" <equinus100@h...> wrote:
>missed it: the whole point is, why Latin plurals (and cases, and
> One clarification of my point before I proceed, in case some people
conjugations, and gender and deponents and what not) in English as
they were used in Latin? What we emprunt are the words (roots, pre-
and suffixes.) Everything else done on them is using English (or
This question probably needs a thesis to answer it. I strongly
suspect it depends on the two languages involved. I will just rush
through the cases where I know something:
Greek to Latin:
The nouns were inflected using the Greek endings where the two
systems were directly comparable. This is more noticeable in poetry
than prose. This was overwhelmingly a late, learnèd phenomenon.
is complicated enough that Latin text books teach the versions of the
declension for the Greek nouns. The more oblique cases have the Lain
Latin had a more extensive case system than Greek. The nominative
singular of most Greek nouns contains a flexional ending, but can be
identified with a Latin one. (A striking case is Latin Plato: for
Greek Plato:n - this is a regular correspondence between the two.)
The gender systems correspond.
Although the nominative and accusative inflections can be readily
mapped between the two systems, they are rarely identical even when
converted to Latin sounds. The common, well matching systems are:
1st decl: Latinised Greek a:, a:n; ae, a:s v. Latin a, am; ae, a:s
e:, e:n; ae, a:s v. Latin a, am; ae, a:s
a:s, a:n; ae, a:s v. Latin a, am; ae, a:s
e:s, e:n; ae, a:s v. Latin a, am; ae, a:s
2nd decl: Latinised Greek os, on; oe, u:s v. Latin us, um; i:, o:s
ros, ron; roe, ru:s v. Latin er, (e)rum; (e)
on, on; a, a v. Latin um, um; a, a
3rd decl Latinised Greek is, in; i:s, i:s v. Latin is, em/im;
x, ca; ces, cas v. Latin x, cem;
o:n, o:na; o:nes, o:nas v. Latin o:, o:nem;
o:n, ona; o:nes, o:nas v. Latin o:, inem;
There are other patterns that match well, but some that don't match
My feeling is that if you have been schooled in the Greek
declensions, and often criticised for mangling the endings, you will
think twice before completely Latinising the Greek inflexions on the
Latin to German:
Nouns are inflected using the full Latin system. I would presume
that this is a learned phenomenon. The nominative singular of a
Latin noun already contains a flexional ending, sometimes subtractive!
The German cases are a subset of the Latin system. German plurals
are fairly irregular.
The gender systems correspond.
Why the German system should be so complicated, I do not know.
Perhaps it is just a carry-over from when German replaced Latin. Do
any other modern languages follow this pattern?
Arabic to Persian:
Only the pausal forms are used - case is not an issue.
Modern Arabic and Persian do not have case inflections. (You might
want to count the definite object suffix, but that is used
indifferently (almost) with Persian nouns, Arabic nouns, and
pronouns.) Persian plural formation is (now) very regular.
Persian does not have gender any more than English does - in fact, if
anything, less so.
It is not at all obvious to me why one should feel compelled to use
the Arabic plural. (There is supposed not to be any compulsion
nowadays.) It is just conceivable that the Persian animate plural
looked like the Arabic dual, and was thus avoided to avoid looking as
though one had got it wrong.
English to Thai:
Borrowed nouns are not inflected. But then, as a first
approximation, Thai has no inflections. Thai has a classificatory
system, but it seems to be semantically based, and it has no
interaction with any English features. Borrowed nouns can readily
acquire what you might call a class prefix; this is not the same as
the classifier. (Examples are <rot ben> = 'merc', <rot ciip> = 'land
rover' for various classes of motor vehicle :))
Aramaic to Modern Hebrew:
Some borrowed nouns are inflected in Aramaic. Spoken modern Hebrew
plurals are irregular. The gender systems are the same. I cannot
see any necessity for nouns to be inflected in Aramaic. Perhaps it
simply reflects a period of bilingual scholarship.
Latin to English:
English nouns do not display case, and Latin nouns are borrowed in
the nominative and only inflected for number, not for case.
For second declension Latin nouns in -us, there are two modes of
borrowing - with and without the ending. When nouns are borrowed
without the ending, they undergo normal English inflections.
Dinosaur names in -saurus are a productive example - brontosaur v.
brontosaurus (or _Brontosaurus_, for these are generic names).
Perhaps this holds the key. If the singular is 'brontosaur', then
there is no question that it will be pluralised as 'brontosaurs'.
The English plural of 'brontosaurus' is 'brontosauruses', and English
has a traditional objection to ending a word in multiple sibilants -
cf. the endingless possessives, such as "Jesus'". Moreover, if one
has gone to the trouble of adding the syllable 'us' to keep a Latin
appearance, one might as well go the whole hog and replace it by 'i'.
Keeping to the field of palaeontology, it occurs to me that in fact
there are several other schemes for naturalising generic names. The
subfamily name yields a noun/adjective in -ine, and it would appear
that 'australopithecines' is used to avoid saying 'australopitheci'
or 'australopithecuses', and likewise I suspect the family name is
similarly used to naturalise names in '-suchus' as '-suchid'.
None of this explains what is happening with Latin nouns in -um and -
The treatment of the Latin 3rd declension is more interesting. Only
three patterns are common: -ex, -ices (e.g. index); -ix, -ices (e.g.
matrix); -sis, -ses (e.g. oasis). The last one may be maintained by
euphony; in speech, I even use the pattern to make the plural
of 'diocese'. It is possible that some of the above forms are
maintained by a combination of the plural being relative common and
mellifluousness. I remember having to learn that the plural
of 'suffix' was not 'suffices'.
I will cut short this exposition on the noun before it does turn into
I suspect the genesis of the use of the foreign inflections is that
the user of these words and his audience had both been formally
taught the foreign language. Adding the wrong endings to often
already inflected words would smack of ignorance, reducing the
credibility of the user. (I'm not sure how far this argument is
applicable to Arabic and Persian.)
Verbs are a very different matter. It can be very difficult for an
inflecting language to borrow a verb from another, dissimilar
inflecting language. Middle English managed to do it from French,
but there may have been enough similarity. Even then, some verbs,
e.g. 'render', were taken over in their infinitive form. There is a
curious pattern - 3rd conjugation verbs could be taken over from
Latin in their present stem, but otherwise the supine stem was used
as the stem. This is the only manner in which English can be said to
follow the Latin conjugations. German solved the problem of
borrowing French verbs by inflecting the infinitive, whence the -
ieren suffix that has spread through the continental Germanic
In none of these cases have foreign inflections been borrowed
directly. English has derived two inflections from Latin - the agent
noun (in -er) and a passive verbal adjective in (-able) - that it can
apply to any semantically suitable verb, and one further, redundant
inflection, the verbal noun, which does not exist for all verbs.
This is perhaps more a matter of word derivation than inflection. In
all cases, these inflections have been formed by analysing words
borrowed (or coined) separately. In fact, the -able inflection could
not be described as such in Latin!
> That's the true revolutionary spirit! They MUST be dood away with (see? "dood" is the regular past participle of "doo", which is the regular spelling of "do"). But if we stand against them, will other people follow us -- all the English-speaking mans, womans and childs?Thanks Piotr.
Well, I was not at all talking about a complete grammatical regularization, just one minor modification in something that is tantalizingly close to being regular. But I see the point, it won't work this way.
> Only artificial languages can remain fully logical for ever, ...so if you want regularity, learn Esperanto.
... and hence artificial and dead. I was talking about (one of the) living, growing, changing languages thow.
>>> This question probably needs a thesis to answer it.know that already?
>> Or even several thesises.
> Interestingly, I just can't accept 'thesises' as English! Did you
The -sis ending, I'd agree, is the only real awkward situation that English possesses. But that's again how one subjectively assesses.
PS: Thanks Richard for that interesting analysis. Wonder if you have ever worked on the question of regularity of languages.