Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

[Respite] pronunciation (last “e” is silent -- unlike [Despite] ) -- Another example?

87 views
Skip to first unread message

henh...@gmail.com

unread,
Dec 26, 2022, 12:38:45 PM12/26/22
to

> Respite is pronounced: [Res-pit]. The last “e” is silent


(even though it's in the dictionary...) i've never heard the [Respite] pronunciation that rhymes with [Despite]



What's another word with a similar (final) [Silent E] at the end ?




----------- other than proper names, like Charlotte ------- i guess... in German, the last E is always pronounced...

Dingbat

unread,
Dec 26, 2022, 6:46:41 PM12/26/22
to
On Monday, December 26, 2022 at 9:38:45 AM UTC-8, henh...@gmail.com wrote:
> > Respite is pronounced: [Res-pit]. The last “e” is silent
>
In EnUS as in the original Old French RESPIT.
https://www.pronounceitright.com/pronunciation/respite-uk-1915
>
> (even though it's in the dictionary...) i've never heard the [Respite] pronunciation that rhymes with [Despite]
>
It ends with ITE in EnUK but doesn't rhyme with DESPITE;
it's RESS-PITE, not RE-SPITE.
>
> What's another word with a similar (final) [Silent E] at the end ?
>
I've heard GRANITE as GRAN-IT in EnUS; it's GRAN-ITE in EnUK.
Spenser reveled in silent Es; I presume YGOE, aka YGO, had a silent E.

henh...@gmail.com

unread,
Dec 26, 2022, 7:15:39 PM12/26/22
to
On Monday, December 26, 2022 at 3:46:41 PM UTC-8, Dingbat wrote:
> On Monday, December 26, 2022 at 9:38:45 AM UTC-8, henh...@gmail.com wrote:
> > > Respite is pronounced: [Res-pit]. The last “e” is silent
> >
> In EnUS as in the original Old French RESPIT.
> https://www.pronounceitright.com/pronunciation/respite-uk-1915
> >


how (why) would the French pron. be preserved over centuries ???


> >
> > What's another word with a similar (final) [Silent E] at the end ?
> >
> I've heard GRANITE as GRAN-IT in EnUS; it's GRAN-ITE in EnUK.


Pomegranate, Candidate, (pron. as -Dit ?)

Definite

Precipice

Ruud Harmsen

unread,
Dec 27, 2022, 7:24:03 AM12/27/22
to
Mon, 26 Dec 2022 09:38:44 -0800 (PST): "henh...@gmail.com"
<henh...@gmail.com> scribeva:
>> Respite is pronounced: [Res-pit]. The last “e” is silent

Completely new to me.

>(even though it's in the dictionary...) i've never heard the [Respite] pronunciation that rhymes with [Despite]

Both possible, says
https://www.collinsdictionary.com/dictionary/english/respite Collins.

>What's another word with a similar (final) [Silent E] at the end ?
>
>----------- other than proper names, like Charlotte ------- i guess... in German, the last E is always pronounced...

--
Ruud Harmsen, http://rudhar.com

wugi

unread,
Jan 1, 2023, 12:16:23 PM1/1/23
to
Op 26/12/2022 om 18:38 schreef henh...@gmail.com:
>
>> Respite is pronounced: [Res-pit]. The last “e” is silent
>
>
> (even though it's in the dictionary...) i've never heard the [Respite] pronunciation that rhymes with [Despite]
>
>
>
> What's another word with a similar (final) [Silent E] at the end ?

Looks like the office-suffice mess.

--
guido wugi

henh...@gmail.com

unread,
Jan 1, 2023, 11:19:14 PM1/1/23
to
Precipice



Date, Pomegranate, Candidate, (pron. as -Dit)

Definite


Age, Sage, -- Adage, Advantage, Message

Peter T. Daniels

unread,
Jan 2, 2023, 10:26:22 AM1/2/23
to
Incidentally, the "e" is silent in both those words.

wugi

unread,
Jan 2, 2023, 3:04:29 PM1/2/23
to
Op 2/01/2023 om 16:26 schreef Peter T. Daniels:
As in both respite and despite?

--
guido wugi

Peter T. Daniels

unread,
Jan 3, 2023, 10:36:51 AM1/3/23
to
Yup. (The final e isn't silent in e.g. anemone.)

The hen person is using "silent e" to mean what little children
learn to call a "helping vowel," the one that marks the vowel before
it as "long": create, complete, despite, promote, dispute. (Compare
bat bate, pet Pete, bit bite, cot cote, cut cute.) Richard Venezky,
who did the best analysis of English orthography, calls this "e" a
"marker." He identifies several "markers" in English orthography.

(It doesn't work for the other "long/short" pair as in boot/book.)

wugi

unread,
Jan 3, 2023, 11:15:01 AM1/3/23
to
Op 3/01/2023 om 16:36 schreef Peter T. Daniels:
And for our example "respite", where final -e is not altogether a
'helping' marker.

--
guido wugi

Peter T. Daniels

unread,
Jan 3, 2023, 12:36:06 PM1/3/23
to
(Not at all.)

I think that's a rare exception. (There are resources to check, but
it would be time-consuming.)

Oh -- did an idiom trip you up? "Not altogether" means 'to some
(small) degree'. Maybe you did intend 'not at all'.

Christian Weisgerber

unread,
Jan 3, 2023, 2:30:07 PM1/3/23
to
On 2023-01-03, Peter T. Daniels <gram...@verizon.net> wrote:

> The hen person is using "silent e" to mean what little children
> learn to call a "helping vowel," the one that marks the vowel before
> it as "long": create, complete, despite, promote, dispute. (Compare
> bat bate, pet Pete, bit bite, cot cote, cut cute.) Richard Venezky,
> who did the best analysis of English orthography, calls this "e" a
> "marker." He identifies several "markers" in English orthography.

Just as for many spelling conventions, there has to be an etymological
basis to this. I assume there was a class of words with the shape
'CVCə. Maybe the vowel lengthened in a stressed open syllable? The
schwa became silent, but -e was reinterpreted as a marker for a
long vowel. Eventually a silent -e was added to words where it was
not etymological, e.g. cnif > knife. Later, the Great Vowel Shift
diphthongized the long vowels.

> (It doesn't work for the other "long/short" pair as in boot/book.)

Those aren't a historical long/short pair, that would be bout/but,
where the "ou" spelling already marks the long vowel.

Pre-GVS, "boot" had /oː/, which didn't have a short counterpart.


... Well, something along those lines, anyway.

--
Christian "naddy" Weisgerber na...@mips.inka.de

Peter T. Daniels

unread,
Jan 3, 2023, 3:48:54 PM1/3/23
to
The dictionary transcription uses double-o with a macron or a
breve centered above them, and the Unicode folk have never
seen fit to provide for them.

Ruud Harmsen via Google Groups <google@rudhar.com>

unread,
Jan 4, 2023, 6:27:26 AM1/4/23
to
Actually, this is possible in Unicode. Via what John Wells calls tie
bars, but are inverted breves in Unicode, I found this block:
https://unicode.org/charts/PDF/U0300.pdf, using the codes
035D COMBINING DOUBLE BREVE
035E COMBINING DOUBLE MACRON

To try it out, I created a mini html file containing this:
<html>
<p>oo with breve: o&#x35d;o
<p>oo with macron: o&#x35e;o
</html>

After opening it in a browser, like with:
firefox nameoffile.htm &
it results in this:
oo with breve: o͝o
oo with macron: o͞o

Ruud Harmsen via Google Groups <google@rudhar.com>

unread,
Jan 4, 2023, 6:44:30 AM1/4/23
to
How well this looks, depends on fonts. My Linux system uses DejaVu Serif
as the default for displaying the HTML in Firefox. In it, the breve and macron
are well centered over the two o’s. But the vertical distance between letters
and diacritics are too small for my liking. (or is it "to my liking"?)

My now site-wide standard font Gentium Book Plus (by SIL International)
also centers well, but the vertical distance is too large. What I'd like is a
compromise between the two.
Strangely, the situation is the reverse with the single breve and macron (304
and U+306): Gentium places them rather close, and DejaVu too wide apart.

In Google Groups, I get them displayed in Roboto or Liberation Sans (Firefox
doesn't clearly indicate what is used for what), which has a small vertical
distance, and the centering is wrong: the breve is over the rightmost o, even
more to the right, and the macron is over the rightmost o too, but slightly more
to the left.

Ruud Harmsen

unread,
Jan 4, 2023, 6:48:16 AM1/4/23
to
Wed, 4 Jan 2023 03:27:25 -0800 (PST): "Ruud Harmsen via Google Groups
<goo...@rudhar.com>" <goo...@rudhar.com> scribeva:

>> The dictionary transcription uses double-o with a macron or a
>> breve centered above them, and the Unicode folk have never
>> seen fit to provide for them.
>
>Actually, this is possible in Unicode. Via what John Wells calls tie
>bars,

Here:
https://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm
found via my link list in:
https://rudhar.com/foneport/en/lingglos.htm

>but are inverted breves in Unicode, I found this block:
>https://unicode.org/charts/PDF/U0300.pdf, using the codes
>035D COMBINING DOUBLE BREVE
>035E COMBINING DOUBLE MACRON

Peter T. Daniels

unread,
Jan 4, 2023, 9:44:03 AM1/4/23
to
I copy-pasted it into Word and changed the font from "Helvetica" (which
has nothing to do with actual Helvetica) to TNR (they soar above the letters,
far too high) and Brill -- where it works, but Brill is a proprietary font that can
only be used in Brill's publications. (It's a lot easier to type diacritics in Word
than what you went through -- either the Unicode followed by Alt-x, or choose
from the Symbols panel, or create a keyboard shortcut

It's thus a problem if font designers haven't provided for their proper
positioning. Scrolling through my installed fonts, more of them do it
right than wrong. The font they used in my book, Gentium, does it
wrong; so does the other SIL font Doulos. (I don't have Charis installed.)

I never noticed that group of seven double marks. Thanks. (In Word's
Symbols pane they have no labels. They're just before the group
of Combining Superscript Roman Letters (why?) and after some
items that don't seem terribly useful -- "zigzag above," "double ring
below," etc.)

> My now site-wide standard font Gentium Book Plus (by SIL International)
> also centers well, but the vertical distance is too large. What I'd like is a
> compromise between the two.
> Strangely, the situation is the reverse with the single breve and macron (304
> and U+306): Gentium places them rather close, and DejaVu too wide apart.
>
> In Google Groups, I get them displayed in Roboto or Liberation Sans (Firefox
> doesn't clearly indicate what is used for what), which has a small vertical
> distance, and the centering is wrong: the breve is over the rightmost o, even
> more to the right, and the macron is over the rightmost o too, but slightly more
> to the left.

I can't select the font in GG.

wugi

unread,
Jan 4, 2023, 12:07:49 PM1/4/23
to
Op 3/01/2023 om 18:36 schreef Peter T. Daniels:
> On Tuesday, January 3, 2023 at 11:15:01 AM UTC-5, wugi wrote:
>> Op 3/01/2023 om 16:36 schreef Peter T. Daniels:

>>> The hen person is using "silent e" to mean what little children
>>> learn to call a "helping vowel," the one that marks the vowel before
>>> it as "long": create, complete, despite, promote, dispute. (Compare
>>> bat bate, pet Pete, bit bite, cot cote, cut cute.) Richard Venezky,
>>> who did the best analysis of English orthography, calls this "e" a
>>> "marker." He identifies several "markers" in English orthography.
>>> (It doesn't work for the other "long/short" pair as in boot/book.)
>>
>> And for our example "respite", where final -e is not altogether a
>> 'helping' marker.
>
> (Not at all.)
>
> I think that's a rare exception. (There are resources to check, but
> it would be time-consuming.)
>
> Oh -- did an idiom trip you up? "Not altogether" means 'to some
> (small) degree'. Maybe you did intend 'not at all'.

Did your didactic mode interfere with a sense of irony?

--
guido wugi

Peter T. Daniels

unread,
Jan 4, 2023, 3:22:11 PM1/4/23
to
With an alloglot, the difference between an intentional and an
unintentional anomaly is a fine one,

Ruud Harmsen

unread,
Jan 4, 2023, 5:30:42 PM1/4/23
to
Wed, 4 Jan 2023 06:44:02 -0800 (PST): "Peter T. Daniels"
<gram...@verizon.net> scribeva:
>It's thus a problem if font designers haven't provided for their proper
>positioning. Scrolling through my installed fonts, more of them do it
>right than wrong. The font they used in my book, Gentium, does it
>wrong; so does the other SIL font Doulos. (I don't have Charis installed.)

Is that "Gentium Plus" or "Gentium Book Plus"? Not the same thing. I
use the latter everywhere on my website now. When experimenting a few
days ago, I loaded Gentium Book Plus from Google Fonts, which seems a
confortable method. But then I found those are the "Basic" variants of
the two Gentium fonts, which (for efficiency for those who do not need
them) lack many special characters, including the IPA characters for
English sh and ch in ship and chip. The tricky part is that browsers
(and probably also Word processors) then insert a character from some
other system font that does have the character or the combination, but
which doesn't always look good. Fonts often differ in height, weight.
etc. as you know.

Example: I want all Greek from Gentium Book Plus now. But on one oage
with a lot of Hebrew too, I chose the preference order "Ezra SIL,
Gentium Book Plus, serif". Then somehow Firefox rendered a Greek mu as
the micro sign µ from Ezra, which has Latin and Latin Extended, but no
Greek. I think that is incorrect. But it happened. I solved it by
marking that column as Gentium only, no Ezra, which I could do because
that table column does not have any Hebrew.

>I never noticed that group of seven double marks. Thanks. (In Word's
>Symbols pane they have no labels. They're just before the group
>of Combining Superscript Roman Letters (why?)

That happens to be the order in Unicode. Quote:
==
Medieval superscript letter diacritics
These are letter diacritics written directly above other letters.
They appear primarily in medieval Germanic manuscripts,
but saw some usage as late as the 19th century in some
languages.
0363 COMBINING LATIN SMALL LETTER A
==

>and after some
>items that don't seem terribly useful -- "zigzag above," "double ring
>below," etc.)

Yes:
==
Miscellaneous additions
0358 COMBINING DOT ABOVE RIGHT
• Latin transliterations of the Southern Min
dialects of Chinese
0359 COMBINING ASTERISK BELOW
035A COMBINING DOUBLE RING BELOW
035B $? COMBINING ZIGZAG ABOVE
• Latin abbreviation, Lithuanian phonetics and
medievalist transcriptions
==

From:
https://unicode.org/charts/PDF/U0300.pdf
In the block "0300-036F Combining Diacritical Marks",
https://rudhar.com/lingtics/uniclnkl.htm .

>> My now site-wide standard font Gentium Book Plus (by SIL International)
>> also centers well, but the vertical distance is too large. What I'd like is a
>> compromise between the two.
>> Strangely, the situation is the reverse with the single breve and macron (304
>> and U+306): Gentium places them rather close, and DejaVu too wide apart.
>>
>> In Google Groups, I get them displayed in Roboto or Liberation Sans (Firefox
>> doesn't clearly indicate what is used for what), which has a small vertical
>> distance, and the centering is wrong: the breve is over the rightmost o, even
>> more to the right, and the macron is over the rightmost o too, but slightly more
>> to the left.
>
>I can't select the font in GG.

Neither can I. I can only have Firefox report what is used, by
clicking "Inspect".

Peter T. Daniels

unread,
Jan 5, 2023, 9:19:16 AM1/5/23
to
On Wednesday, January 4, 2023 at 5:30:42 PM UTC-5, Ruud Harmsen wrote:
> Wed, 4 Jan 2023 06:44:02 -0800 (PST): "Peter T. Daniels"
> <gram...@verizon.net> scribeva:

> >It's thus a problem if font designers haven't provided for their proper
> >positioning. Scrolling through my installed fonts, more of them do it
> >right than wrong. The font they used in my book, Gentium, does it
> >wrong; so does the other SIL font Doulos. (I don't have Charis installed.)
>
> Is that "Gentium Plus" or "Gentium Book Plus"? Not the same thing. I

I have Gentium and Gentium Plus. I don't know what variant their typesetter
used. Fortunately she's in England not India, but I know she uses InDesign
because some of the table cells got initial capitals. (WWS, pre-Unicode,
was done in FrameMaker, which is greatly superior, but Adobe bought it
to eliminate the competition and did not (has not?) continue to improve it.)

It's not a great font for book publication, but it's apparently the only one
they have that can handle all of phonetics.

> use the latter everywhere on my website now. When experimenting a few
> days ago, I loaded Gentium Book Plus from Google Fonts, which seems a
> confortable method. But then I found those are the "Basic" variants of

Google's "Noto" family of exotics is for the most part decent.

> the two Gentium fonts, which (for efficiency for those who do not need
> them) lack many special characters, including the IPA characters for
> English sh and ch in ship and chip. The tricky part is that browsers
> (and probably also Word processors) then insert a character from some
> other system font that does have the character or the combination, but
> which doesn't always look good. Fonts often differ in height, weight.
> etc. as you know.

> >Combining Superscript Roman Letters (why?)
>
> Quote:
> ==
> Medieval superscript letter diacritics
> These are letter diacritics written directly above other letters.
> They appear primarily in medieval Germanic manuscripts,
> but saw some usage as late as the 19th century in some
> languages.

And that was a reason for including them at such an early stage of
creating the inventory?

Ruud Harmsen

unread,
Jan 5, 2023, 11:30:47 AM1/5/23
to
Thu, 5 Jan 2023 06:19:15 -0800 (PST): "Peter T. Daniels"
<gram...@verizon.net> scribeva:

>On Wednesday, January 4, 2023 at 5:30:42 PM UTC-5, Ruud Harmsen wrote:
>> Wed, 4 Jan 2023 06:44:02 -0800 (PST): "Peter T. Daniels"
>> <gram...@verizon.net> scribeva:
>
>> >It's thus a problem if font designers haven't provided for their proper
>> >positioning. Scrolling through my installed fonts, more of them do it
>> >right than wrong. The font they used in my book, Gentium, does it
>> >wrong; so does the other SIL font Doulos. (I don't have Charis installed.)
>>
>> Is that "Gentium Plus" or "Gentium Book Plus"? Not the same thing. I
>
>I have Gentium and Gentium Plus. I don't know what variant their typesetter
>used. Fortunately she's in England not India, but I know she uses InDesign
>because some of the table cells got initial capitals. (WWS, pre-Unicode,
>was done in FrameMaker, which is greatly superior, but Adobe bought it
>to eliminate the competition and did not (has not?) continue to improve it.)
>
>It's not a great font for book publication, but it's apparently the only one
>they have that can handle all of phonetics.
>
>> use the latter everywhere on my website now. When experimenting a few
>> days ago, I loaded Gentium Book Plus from Google Fonts, which seems a
>> confortable method. But then I found those are the "Basic" variants of
>
>Google's "Noto" family of exotics is for the most part decent.

Yes, that too.

But Google also offers technical facilities for many other fonts, also
many that were made by others, provided they are made available under
a free licence, like SIL’s OFL.

https://fonts.google.com/
https://developers.google.com/fonts/faq
https://developers.google.com/fonts/docs/css2

With a single line of CSS code you can import the font from their
servers, and refer to it everywhere in a site. In their demo, phonetic
symbols worked, but not when I used their method on my site. I found
that that is because what they offer is a limited Gentium font. The
docu was unclear about that.

So now I have the full font on my own site, as downloaded from SIL.

Browsers will cache everything, so there isn't much of a performance
penalty.

>> the two Gentium fonts, which (for efficiency for those who do not need
>> them) lack many special characters, including the IPA characters for
>> English sh and ch in ship and chip. The tricky part is that browsers
>> (and probably also Word processors) then insert a character from some
>> other system font that does have the character or the combination, but
>> which doesn't always look good. Fonts often differ in height, weight.
>> etc. as you know.
>
>> >Combining Superscript Roman Letters (why?)
>>
>> Quote:
>> ==
>> Medieval superscript letter diacritics
>> These are letter diacritics written directly above other letters.
>> They appear primarily in medieval Germanic manuscripts,
>> but saw some usage as late as the 19th century in some
>> languages.
>
>And that was a reason for including them at such an early stage of
>creating the inventory?

I don't know. I also don't know which Unicode version first had them.
They sometimes add subblocks in existing larger blocks.

Christian Weisgerber

unread,
Jan 5, 2023, 1:30:06 PM1/5/23
to
On 2022-12-27, henh...@gmail.com <henh...@gmail.com> wrote:

>> > > Respite is pronounced: [Res-pit].
>
> Pomegranate, Candidate, (pron. as -Dit ?)
>
> Definite
>
> Precipice

quarantine

Peter T. Daniels

unread,
Jan 5, 2023, 2:01:55 PM1/5/23
to
On Thursday, January 5, 2023 at 1:30:06 PM UTC-5, Christian Weisgerber wrote:
> On 2022-12-27, henh...@gmail.com <henh...@gmail.com> wrote:

> >> > > Respite is pronounced: [Res-pit].
> > Pomegranate, Candidate, (pron. as -Dit ?)

(lots of unstressed -ate nouns)

> > Definite
> > Precipice
>
> quarantine

Not unstressed/reduced. AmE /iyn/ [ijn], conceivably BrE /ayn/ [ajn],
so not in this set of exceptions.

Ross Clark

unread,
Jan 6, 2023, 10:21:25 PM1/6/23
to
Actually we little children called it "silent e" if we called it anything.

I spent a little time looking for more examples of this particular case
which we might call "useless e" -- unmotivated for the speller and
misleading for the reader. Here's a longer list (mainly of two-syllable
words):

respite
granite
definite, infinite etc.
pomegranate
candidate
private
opiate
curate
pirate
senate
palate
agate
minute

olive
octave
massive, captive etc

promise, premise, mortise, anise
purpose

welcome
awesome, handsome etc.

medicine
feminine
heroine
engine
famine
ermine
urine
doctrine
fortune
missile
adjs in -ile docile, virile, agile, etc
gunwale
capsule

office
palace
precipice
college
pillage
adage
advantage
message

Some of these probably have alternative pronunciations (as with
"respite") which would make the -e unproblematic for those who so
pronounce them.
The majority would be better spelled without the -e. But we can't do
that with the -ce and -ge items, in which the -e is doing its other job
of indicating a "soft" value for the preceding consonant.

Tim Lang

unread,
Jan 7, 2023, 5:40:40 AM1/7/23
to
On 07.01.2023 04:21, Ross Clark wrote:

>Actually we little children called it "silent e" if we called it anything.

Also in numerous surnames & 1st names, e.g. Clarke, Greene, O'Toole,
Brooke, Drake, Boone, Boyle, Doyle, Jake, Lorne, Luke, Caine, Laine,
Chalke, Combe, *gue, *hue, *hoe, Hoare, Hoyte, Home/Hume (whereas
-ham: Nottingham, Tottenham etc; the rest of them -hams in Bavaria
and Austria :-)), Joule (but this is French), Keefe, Lane, Lowe, Meade,
Moore, Wolfe, Payne, Peake, Wayne, Scarfe, Scaife, Sharpe, Steele, Tyne,
Veale, Yoe (this one, I assume, variant of Yeo(man)). (...)

Presumably, most of these -e's are French and other (Romance or Celtic)
"etymological" remnants. Or having only "ornamental" reasons (in other
epochs). Irish O'Toole could have been spelled O'Tool, since in the
original there is no final -e: Tuathal. And -oo- represents a "con-
tracted" -uatha-. (BTW: it seems that Celtic tuath stays in the cognate
relationship with "teuta, diut, teut..." then "Deutsch" and "Dutch".
All of them meaning "people" in the sense of "populus, population,
a mass," which in German has been replaced by "Volk", since "Deutsch"
and "Teutsch" only have been used as ethnonyms for over 800 years now.
Celtic god's name Teutates/Toutatis also contains this word.)

Tim

henh...@gmail.com

unread,
Jan 8, 2023, 1:25:46 AM1/8/23
to
thanks for the list.... i couldn't post for almost 2 days and so i sent email to Dr.BenLizro but it was bounced back. ( benlizro@ihug......nz <------ what does BenLizro mean ? )


in case of
> olive
> octave (etc.)

(it seems that) The E is there to make it wf (well-formed).

oliv
octav


a word ends in V only if it's of foregin-origin, as in Slav, Yugoslav, (or Russian) ... Kiev, Molotov

(or a contraction)

Peter T. Daniels

unread,
Jan 8, 2023, 9:58:15 AM1/8/23
to
Yes, that's Venezky's third "marker" use of e. English doesn't like
words to end with v (shiv, rev are usually given as the only ones
that aren't [foreign] names).

Ross Clark

unread,
Jan 8, 2023, 2:18:54 PM1/8/23
to
It's an old email address that no longer exists. Still serves as a kind
of identifier on newsgroups. The "-ro" is me.

>
> in case of
>> olive
>> octave (etc.)
>
> (it seems that) The E is there to make it wf (well-formed).
>
> oliv
> octav
>
>
> a word ends in V only if it's of foregin-origin, as in Slav, Yugoslav, (or Russian) ... Kiev, Molotov
>
> (or a contraction)
>

That may be a correct observation, but I don't believe there is actually
any such "constraint" on English spelling. (Actually spellings of
"oliv(e)" with and without the -e are attested for as long as the word
has been in the language.) Just why this restriction on final <v> came
about might make an interesting historical study, but it is not a good
reason for preserving the present spelling.

Christian Weisgerber

unread,
Jan 8, 2023, 6:30:06 PM1/8/23
to
On 2023-01-08, henh...@gmail.com <henh...@gmail.com> wrote:

>> olive
>> octave (etc.)
>
> (it seems that) The E is there to make it wf (well-formed).
>
> oliv
> octav
>
> a word ends in V only if it's of foregin-origin, as in Slav, Yugoslav, (or Russian) ... Kiev, Molotov

Or relatively recent coinages like "shiv".

I think you will find more spelling oddities involving v, such as
the o in "love", because the origins of English orthography predate
the typographical distinction between u and v, and various devices
were invented to keep the two readings of the same letter apart and
avoid further confusion between uu/vv and w.

(French also has a few such oddities, e.g. the nonetymological h
in huile < Latin oleum to avoid reading it as "vile".)

wugi

unread,
Jan 9, 2023, 6:45:21 AM1/9/23
to
Op 8/01/2023 om 23:33 schreef Christian Weisgerber:
And Spanish: hueso (L. os), huelo (oler, olor), huevo (L. ovum)...

--
guido wugi

Peter T. Daniels

unread,
Jan 9, 2023, 7:56:20 AM1/9/23
to
On Sunday, January 8, 2023 at 6:30:06 PM UTC-5, Christian Weisgerber wrote:
> On 2023-01-08, henh...@gmail.com <henh...@gmail.com> wrote:
>
> >> olive
> >> octave (etc.)
> >
> > (it seems that) The E is there to make it wf (well-formed).
> >
> > oliv
> > octav
> >
> > a word ends in V only if it's of foregin-origin, as in Slav, Yugoslav, (or Russian) ... Kiev, Molotov
> Or relatively recent coinages like "shiv".
>
> I think you will find more spelling oddities involving v, such as
> the o in "love", because the origins of English orthography predate
> the typographical distinction between u and v, and various devices

That one, and "come" and "mother," and some others, seems to be
because a long string of minims (short vertical strokes) was deemed
to be confusing or even unreadable, so <u> became <o> (the top of
the letter was closed).

> were invented to keep the two readings of the same letter apart and
> avoid further confusion between uu/vv and w.

<w> hadn't been invented yet.

Christian Weisgerber

unread,
Jan 9, 2023, 9:30:06 AM1/9/23
to
On 2023-01-09, Peter T. Daniels <gram...@verizon.net> wrote:

>> the o in "love", because the origins of English orthography predate
>> the typographical distinction between u and v, and various devices
>> were invented to keep the two readings of the same letter apart and
>> avoid further confusion between uu/vv and w.
>
><w> hadn't been invented yet.

Exactly. Instead, <uu> was written. So today's <uv> and <w> were
indistinguishable.

Ruud Harmsen

unread,
Jan 9, 2023, 10:07:09 AM1/9/23
to
Mon, 9 Jan 2023 12:45:17 +0100: wugi <wu...@scrlt.com> scribeva:
Yes. It's a spelling issue: no Spanish word can begin with ue, ua,
etc. Hence also Chihuahua for is said as Chiwawa.

Ruud Harmsen

unread,
Jan 9, 2023, 10:39:57 AM1/9/23
to
Mon, 09 Jan 2023 16:07:06 +0100: Ruud Harmsen <r...@rudhar.com>
scribeva:
Ougadougou: Wagadugu, French has a simular issue.

Christian Weisgerber

unread,
Jan 9, 2023, 11:30:08 AM1/9/23
to
On 2023-01-09, Ruud Harmsen <r...@rudhar.com> wrote:

>>> (French also has a few such oddities, e.g. the nonetymological h
>>> in huile < Latin oleum to avoid reading it as "vile".)
>>
>>And Spanish: hueso (L. os), huelo (oler, olor), huevo (L. ovum)...
>
> Yes. It's a spelling issue: no Spanish word can begin with ue, ua,
> etc. Hence also Chihuahua for is said as Chiwawa.

But "Chihuahua" is of non-Spanish origin, even if the etymology is
uncertain. In Nahuatl names at least, <h> corresponds to underlying
/h/ ~ glottal stop.

Dingbat

unread,
Mar 22, 2023, 1:21:42 AM3/22/23
to
Luv is an alternate spelling of Love.
Chevrolet named a pickup truck LUV.
'Lurve is in the air' is one expression
with the other alternate spelling.
Lurve has a terminal e even tho dropping
it wouldn't change its pronunciation.

Peter T. Daniels

unread,
Mar 22, 2023, 8:37:31 AM3/22/23
to
On Wednesday, March 22, 2023 at 1:21:42 AM UTC-4, Dingbat wrote:

> Luv is an alternate spelling of Love.
> Chevrolet named a pickup truck LUV.
> 'Lurve is in the air' is one expression
> with the other alternate spelling.
> Lurve has a terminal e even tho dropping
> it wouldn't change its pronunciation.

A minor orthographic rule of English is that words don't end with v.
The exceptions are usually abbreviations This one would be eye-dialect..

Christian Weisgerber

unread,
Mar 22, 2023, 11:30:06 AM3/22/23
to
On 2023-01-08, Ross Clark <benl...@ihug.co.nz> wrote:

>> a word ends in V only if it's of foregin-origin, as in Slav, Yugoslav, (or Russian) ... Kiev, Molotov
>>
>> (or a contraction)
>
> That may be a correct observation, but I don't believe there is actually
> any such "constraint" on English spelling. (Actually spellings of
> "oliv(e)" with and without the -e are attested for as long as the word
> has been in the language.) Just why this restriction on final <v> came
> about might make an interesting historical study, but it is not a good
> reason for preserving the present spelling.

Old English didn't have final [v], it was devoiced to [f]. For this
reason, modern English words that end in /v/ are the result of
* loss of a final vowel in Middle English, still reflected in spelling
as -e, in particular in verbs: "have", "leave", etc.;
* borrowing;
* late formations, e.g. "shiv".

Old French also didn't have final /v/, devoicing it to /f/ instead,
e.g. "neuf, neuve". Final /v/ appeared in French only when the
final schwas disappeared, which are still retained in spelling as
-e. So French as the predominant source of borrowings only provided
"-ve".

For orthographic choices involving u and v, it's also worth keeping
in mind that English orthography was standardized at a time when
those letters were not yet distinguished. Although that may not
be relevant here, given the lack of final -u in English.
0 new messages