Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Eli Zaretskii]

>  commit baaf97ce1a1f14e1b999ad17084022a725fbcccc
>  Author:     Sean Whitton <spwhitton@spwhitton.name>
>  AuthorDate: Fri Aug 30 11:43:35 2024 +0100
>  Commit:     Sean Whitton <spwhitton@spwhitton.name>
>  CommitDate: Fri Aug 30 11:47:35 2024 +0100
> 
>      ; Fix some ungrammatical uses of "allows to"
> 
>      * doc/emacs/anti.texi (Antinews):
>      * etc/NEWS:
>      * etc/NEWS.29:
>      * lib/cdefs.h:
>      * lisp/tab-bar.el (tab-bar-select-restore-windows):
>      * lisp/vc/vc-git.el (vc-git-print-log-follow):
>      * m4/gnulib-common.m4 (gl_CHECK_FUNCS_SET_RESULTS):
>      (gl_CHECK_FUNCS_ANDROID): Fix ungrammatical uses of "allows to".

It is wrong to fix problems in Gnulib files in our repository, because
those fixes will be overridden when the files are next time imported
from Gnulib.  Problems in Gnulib should be reported to the Gnulib list
instead.

More generally, "allows to" is AFAIK not necessarily a grammatical
mistake, it's a stylistic preference.  So this kind of changes should
not have been made automatically and without discussion.

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Alan Mackenzie]

Hello, Eli.

On Fri, Aug 30, 2024 at 20:40:22 +0300, Eli Zaretskii wrote:
> >  commit baaf97ce1a1f14e1b999ad17084022a725fbcccc
> >  Author:     Sean Whitton <spwhitton@spwhitton.name>
> >  AuthorDate: Fri Aug 30 11:43:35 2024 +0100
> >  Commit:     Sean Whitton <spwhitton@spwhitton.name>
> >  CommitDate: Fri Aug 30 11:47:35 2024 +0100
> > 
> >      ; Fix some ungrammatical uses of "allows to"
> > 
> >      * doc/emacs/anti.texi (Antinews):
> >      * etc/NEWS:
> >      * etc/NEWS.29:
> >      * lib/cdefs.h:
> >      * lisp/tab-bar.el (tab-bar-select-restore-windows):
> >      * lisp/vc/vc-git.el (vc-git-print-log-follow):
> >      * m4/gnulib-common.m4 (gl_CHECK_FUNCS_SET_RESULTS):
> >      (gl_CHECK_FUNCS_ANDROID): Fix ungrammatical uses of "allows to".

> It is wrong to fix problems in Gnulib files in our repository, because
> those fixes will be overridden when the files are next time imported
> from Gnulib.  Problems in Gnulib should be reported to the Gnulib list
> instead.

> More generally, "allows to" is AFAIK not necessarily a grammatical
> mistake, it's a stylistic preference.  So this kind of changes should
> not have been made automatically and without discussion.

No, Eli, as a native English speaker, I can definitively say "allows to"
is wrong.  It's a fairly subtle point, but instantly marks its writer as
a non-native speaker.

"Allow" takes a direct object (which may be a gerund), not an
infinitive.  Further complicating the structure, that direct object may
be the person or agent who is being permitted or enabled, not
necessarily the action that is so enabled.

For example:
Wrong: This allows to add text properties to the text.

Correct: (i) This allows the addition of text properties to the text.
(ii) This allows adding text properties to the text.
(iii) This allows text properties to be added to the text.
(iv) This allows the programmer to add text properties to the text.

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Eli Zaretskii]

> Date: Fri, 30 Aug 2024 18:14:59 +0000
> Cc: Sean Whitton <spwhitton@spwhitton.name>, emacs-devel@gnu.org
> From: Alan Mackenzie <acm@muc.de>
> 
> > It is wrong to fix problems in Gnulib files in our repository, because
> > those fixes will be overridden when the files are next time imported
> > from Gnulib.  Problems in Gnulib should be reported to the Gnulib list
> > instead.
> 
> > More generally, "allows to" is AFAIK not necessarily a grammatical
> > mistake, it's a stylistic preference.  So this kind of changes should
> > not have been made automatically and without discussion.
> 
> No, Eli, as a native English speaker, I can definitively say "allows to"
> is wrong.  It's a fairly subtle point, but instantly marks its writer as
> a non-native speaker.

Being a native English speaker doesn't mean that you are always right
about use of English.  Please look on the Internet and you will see
that the above is not always correct.

> "Allow" takes a direct object (which may be a gerund), not an
> infinitive.  Further complicating the structure, that direct object may
> be the person or agent who is being permitted or enabled, not
> necessarily the action that is so enabled.
> 
> For example:
> Wrong: This allows to add text properties to the text.
> 
> Correct: (i) This allows the addition of text properties to the text.
> (ii) This allows adding text properties to the text.
> (iii) This allows text properties to be added to the text.
> (iv) This allows the programmer to add text properties to the text.

This page disagrees:

  https://english.stackexchange.com/questions/85069/is-the-construction-it-allows-to-proper-english

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Sean Whitton]

Hello,

On Fri 30 Aug 2024 at 08:40pm +03, Eli Zaretskii wrote:

> It is wrong to fix problems in Gnulib files in our repository, because
> those fixes will be overridden when the files are next time imported
> from Gnulib.  Problems in Gnulib should be reported to the Gnulib list
> instead.

Oops.  Thanks.  I didn't edit ORG-NEWS because of this issue.

It's just m4/gnulib-common.m4 I should undo, right?  Or do we not need
to undo it now, and just wait for the next gnulib import?

> More generally, "allows to" is AFAIK not necessarily a grammatical
> mistake, it's a stylistic preference.  So this kind of changes should
> not have been made automatically and without discussion.

It's not just stylistic -- it's certainly incorrect in American English,
which is Emacs's dialect.

-- 
Sean Whitton

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Sean Whitton]

Hello,

On Fri 30 Aug 2024 at 06:14pm GMT, Alan Mackenzie wrote:

> No, Eli, as a native English speaker, I can definitively say "allows to"
> is wrong.  It's a fairly subtle point, but instantly marks its writer as
> a non-native speaker.

I think you could have written this message with a kinder tone, Alan,
given, as you acknowledge, it's a subtle point for people to have to
learn.

Anyway, it does indeed mark out writing as coming from a non-native user
of English.  You see it relatively commonly in software contexts,
probably because lots of programmers are not native English speakers.

-- 
Sean Whitton

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Eli Zaretskii]

> From: Sean Whitton <spwhitton@spwhitton.name>
> Cc: emacs-devel@gnu.org
> Date: Fri, 30 Aug 2024 20:00:53 +0100
> 
> It's just m4/gnulib-common.m4 I should undo, right?

No, not just that.  Quite a few m4/*.m4 files are from Gnulib.  You
need to look at the Git history to know which ones.

> Or do we not need to undo it now, and just wait for the next gnulib
> import?

We don't need to undo, but we definitely should tell Gnulib folks
about the issue.

> > More generally, "allows to" is AFAIK not necessarily a grammatical
> > mistake, it's a stylistic preference.  So this kind of changes should
> > not have been made automatically and without discussion.
> 
> It's not just stylistic -- it's certainly incorrect in American English,
> which is Emacs's dialect.

I disagree, but that is not the important aspect of this.  The
important aspect is that changing grammar should not make the text
harder to read and understand.  By mechanically changing those places
to be grammatically you definitely made at least some of the text more
awkward and hard to read.  E.g., look at this one:

  -The command @code{recover-file} no longer allows to display the diffs
  +The command @code{recover-file} no longer allows displaying the diffs

Does this really read well to you?  Or how about this one:

  -dnl The first variable allows to distinguish all three cases.
  +dnl The first variable allows distinguishing all three cases.

"Allows distinguishing"? really?

You should have instead reworded the text to avoid the use of these
constructs, making the text more clear and side-stepping the issue of
"allow to" entirely.

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Eli Zaretskii]

> From: Sean Whitton <spwhitton@spwhitton.name>
> Cc: Eli Zaretskii <eliz@gnu.org>,  emacs-devel@gnu.org
> Date: Fri, 30 Aug 2024 20:03:54 +0100
> 
> Anyway, it does indeed mark out writing as coming from a non-native user
> of English.  You see it relatively commonly in software contexts,
> probably because lots of programmers are not native English speakers.

In technical texts these uses of "allows" omit "one" or "you", so the
text really says "SOMETHING allows _you_ to do whatever", but omits
"you".  Changing this mechanically to "allows doing" in many cases
makes the text much harder to read, a mouthful, really.

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Sean Whitton]

Hello,

On Fri 30 Aug 2024 at 10:18pm +03, Eli Zaretskii wrote:

> I disagree, but that is not the important aspect of this.  The
> important aspect is that changing grammar should not make the text
> harder to read and understand.  By mechanically changing those places
> to be grammatically you definitely made at least some of the text more
> awkward and hard to read.  E.g., look at this one:

Let me just note that I didn't do it mechanically.  I always read the
surrounding context if I change something like this, else I can't know
for sure that I did it correctly.

>   -The command @code{recover-file} no longer allows to display the diffs
>   +The command @code{recover-file} no longer allows displaying the diffs
>
> Does this really read well to you?  Or how about this one:
>
>   -dnl The first variable allows to distinguish all three cases.
>   +dnl The first variable allows distinguishing all three cases.
>
> "Allows distinguishing"? really?

Yes -- I don't find these the slightest bit unclear, unnatural or
difficult to read.  "allows distinguishing" doesn't stand out to me.

> You should have instead reworded the text to avoid the use of these
> constructs, making the text more clear and side-stepping the issue of
> "allow to" entirely.

Well, in this, as you are a non-native speaker, your opinion on whether
it's clear counts for more than mine.  So I would be happy to put in the
effort to try to make it clearer.  However, I don't think I'm capable of
doing a good job, because they all read perfectly fine for me.

My only suggestion would be "allows you to make Emacs distinguish" ?

-- 
Sean Whitton

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Sean Whitton]

Hello,

On Fri 30 Aug 2024 at 10:22pm +03, Eli Zaretskii wrote:

>> From: Sean Whitton <spwhitton@spwhitton.name>
>> Cc: Eli Zaretskii <eliz@gnu.org>,  emacs-devel@gnu.org
>> Date: Fri, 30 Aug 2024 20:03:54 +0100
>>
>> Anyway, it does indeed mark out writing as coming from a non-native user
>> of English.  You see it relatively commonly in software contexts,
>> probably because lots of programmers are not native English speakers.
>
> In technical texts these uses of "allows" omit "one" or "you", so the
> text really says "SOMETHING allows _you_ to do whatever", but omits
> "you".  Changing this mechanically to "allows doing" in many cases
> makes the text much harder to read, a mouthful, really.

It's not always 'one' or 'you'.

    The first variable allows distinguishing all three cases.

Here it is the code, not any human, that is to do the distinguishing.

-- 
Sean Whitton

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Sean Whitton]

Hello,

On Fri 30 Aug 2024 at 10:00pm +03, Eli Zaretskii wrote:

> This page disagrees:
>
>   https://english.stackexchange.com/questions/85069/is-the-construction-it-allows-to-proper-english

No, people there conclude it's ungrammatical, and they explain why.

Both the 19th century examples are not the form we're objecting to; it's
not the mere adjacency of 'allows' and 'to'.

-- 
Sean Whitton

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Ship Mints]

[-- Attachment #1: Type: text/plain, Size: 2395 bytes --]

I find Eli's English is better than native proficiency, especially
considering the bell curve is alive and well. I currently live in the USA
where English proficiency seems to be on a long, slow, steady downward
trend. I assumed he is a native, or near native, speaker/writer. He writes
well, and all day long on this list, about very complex topics (at least
compared to the general public), and is patient and very clear. He even
uses correct punctuation where I observe many Americans don't, or don't
bother; if they are studious enough to know it.

On Fri, Aug 30, 2024 at 5:01 PM Sean Whitton <spwhitton@spwhitton.name>
wrote:

> Hello,
>
> On Fri 30 Aug 2024 at 10:18pm +03, Eli Zaretskii wrote:
>
> > I disagree, but that is not the important aspect of this.  The
> > important aspect is that changing grammar should not make the text
> > harder to read and understand.  By mechanically changing those places
> > to be grammatically you definitely made at least some of the text more
> > awkward and hard to read.  E.g., look at this one:
>
> Let me just note that I didn't do it mechanically.  I always read the
> surrounding context if I change something like this, else I can't know
> for sure that I did it correctly.
>
> >   -The command @code{recover-file} no longer allows to display the diffs
> >   +The command @code{recover-file} no longer allows displaying the diffs
> >
> > Does this really read well to you?  Or how about this one:
> >
> >   -dnl The first variable allows to distinguish all three cases.
> >   +dnl The first variable allows distinguishing all three cases.
> >
> > "Allows distinguishing"? really?
>
> Yes -- I don't find these the slightest bit unclear, unnatural or
> difficult to read.  "allows distinguishing" doesn't stand out to me.
>
> > You should have instead reworded the text to avoid the use of these
> > constructs, making the text more clear and side-stepping the issue of
> > "allow to" entirely.
>
> Well, in this, as you are a non-native speaker, your opinion on whether
> it's clear counts for more than mine.  So I would be happy to put in the
> effort to try to make it clearer.  However, I don't think I'm capable of
> doing a good job, because they all read perfectly fine for me.
>
> My only suggestion would be "allows you to make Emacs distinguish" ?
>
> --
> Sean Whitton
>
>

[-- Attachment #2: Type: text/html, Size: 2968 bytes --]

RE: [External] : Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Drew Adams]

> > Anyway, it does indeed mark out writing as coming from a non-native
> > user of English.  You see it relatively commonly in software contexts,
> > probably because lots of programmers are not native English speakers.
> 
> In technical texts these uses of "allows" omit "one" or "you", so the
> text really says "SOMETHING allows _you_ to do whatever", but omits
> "you".  Changing this mechanically to "allows doing" in many cases
> makes the text much harder to read, a mouthful, really.

Even better is to change it to "lets you do whatever".

Or "If SOMETHING then you can do whatever", which
makes the _user_ the actor of the active voice.

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Mike Kupfer]

Sean Whitton wrote:

> On Fri 30 Aug 2024 at 10:18pm +03, Eli Zaretskii wrote:
> > You should have instead reworded the text to avoid the use of these
> > constructs, making the text more clear and side-stepping the issue of
> > "allow to" entirely.
> 
> Well, in this, as you are a non-native speaker, your opinion on whether
> it's clear counts for more than mine.  So I would be happy to put in the
> effort to try to make it clearer.  However, I don't think I'm capable of
> doing a good job, because they all read perfectly fine for me.

I'll take a stab at it.

doc/emacs/anti.texi: 

  The command no longer allows the display of diffs between a file and
  its auto-save file.

etc/NEWS: 

  This lets the user customize different switches...

etc/NEWS.29:

  *** Users can commit a subset of changes in a diffs buffer.

  This function provides a way to modify some or all of the settings of an
  existing connection-local profile.

  Specifying a cons as the FROM argument causes the measurement of text
  to start from a specified amount of pixels above or below a position.

lib/cdefs.h:

  ISO C99 lets you declare arrays...

lisp/tab-bar.el:

  A button in the window allows for the restoration of the killed
  buffer.

lisp/vc/vc-git.el:

  When this variable is nil, and the log ends with a rename, a button
  below the rename allows for the display of the log for the file name
  before the rename."

Given Eli's comment about changes related to gnulib, I won't suggest
anything for gnulib-common.m4.

HTH,
mike

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Po Lu]

Sean Whitton <spwhitton@spwhitton.name> writes:

> Anyway, it does indeed mark out writing as coming from a non-native user
> of English.  You see it relatively commonly in software contexts,
> probably because lots of programmers are not native English speakers.

AFAIK this construct is taught in the German education system, which
probably gives rise to the bulk of its users.  Not to imply that it
isn't ungrammatical.

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Sean Whitton]

Hello,

On Fri 30 Aug 2024 at 04:39pm -07, Mike Kupfer wrote:

> I'll take a stab at it.

Thanks.

"allows for Xing" reads less well than "allows Xing" to me, but perhaps
Eli will disagree.

-- 
Sean Whitton

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Sean Whitton]

Hello,

On Sat 31 Aug 2024 at 08:23am +08, Po Lu wrote:

> AFAIK this construct is taught in the German education system, which
> probably gives rise to the bulk of its users.  Not to imply that it
> isn't ungrammatical.

I have a qualification in teaching English as a foreign language and did
so for two years.  It feels like every country has its own set of both
incorrect and highly unidiomatic bits of English that somehow survive on
its syllabi.  (And I assume the same is true for other languages taught
in English-speaking countries.)

I haven't taught Germans, so thank you for sharing this.

-- 
Sean Whitton

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Eli Zaretskii]

> From: Sean Whitton <spwhitton@spwhitton.name>
> Cc: emacs-devel@gnu.org
> Date: Fri, 30 Aug 2024 22:00:07 +0100
> 
> > You should have instead reworded the text to avoid the use of these
> > constructs, making the text more clear and side-stepping the issue of
> > "allow to" entirely.
> 
> Well, in this, as you are a non-native speaker, your opinion on whether
> it's clear counts for more than mine.  So I would be happy to put in the
> effort to try to make it clearer.  However, I don't think I'm capable of
> doing a good job, because they all read perfectly fine for me.
> 
> My only suggestion would be "allows you to make Emacs distinguish" ?

I would reword to say something like "The first variable can be used
to distinguish..."  And similarly in the other cases.  For example:

   ** 'window-text-pixel-size' understands a new meaning of FROM.
  -Specifying a cons as the FROM argument allows to start measuring text
  -from a specified amount of pixels above or below a position.
  +Specifying a cons as the FROM argument allows starting the measurement
  +of text from a specified amount of pixels above or below a position.

Here I would say

  The FROM argument can now be a cons, which means to start measuring
  text from...

To summarize: IME, when people use problematic grammar, it is always
better to completely reword the text, and in many cases such rewording
will make the text much more clear and easy to understand.

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Gerd Möllmann]

Sean Whitton <spwhitton@spwhitton.name> writes:

> Hello,
>
> On Sat 31 Aug 2024 at 08:23am +08, Po Lu wrote:
>
>> AFAIK this construct is taught in the German education system, which
>> probably gives rise to the bulk of its users.  Not to imply that it
>> isn't ungrammatical.
>
> I have a qualification in teaching English as a foreign language and did
> so for two years.  It feels like every country has its own set of both
> incorrect and highly unidiomatic bits of English that somehow survive on
> its syllabi.  (And I assume the same is true for other languages taught
> in English-speaking countries.)
>
> I haven't taught Germans, so thank you for sharing this.

I think it's a meme that allows to make fun of Germans :-).

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Eli Zaretskii]

> From: Sean Whitton <spwhitton@spwhitton.name>
> Cc: Eli Zaretskii <eliz@gnu.org>,  emacs-devel@gnu.org
> Date: Sat, 31 Aug 2024 07:15:27 +0100
> 
> On Fri 30 Aug 2024 at 04:39pm -07, Mike Kupfer wrote:
> 
> > I'll take a stab at it.
> 
> Thanks.
> 
> "allows for Xing" reads less well than "allows Xing" to me, but perhaps
> Eli will disagree.

See my other suggestions: I think the rewording should be more
radical.  E.g., in this case:

  lisp/tab-bar.el:

    A button in the window allows for the restoration of the killed
    buffer.

I'd use something like

   Pressing a button in the window restores the killed buffer.
or
   You can press a button in the window to restore the killed buffer.

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Mike Kupfer]

Eli Zaretskii wrote:

> See my other suggestions: I think the rewording should be more
> radical.  

Agreed, a more radical rewording would likely help.

mike

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Sean Whitton]

Hello,

Okay, thanks, I went through and edited them all again (not the gnulib).

-- 
Sean Whitton

Re: emacs-30 baaf97ce1a1: ; Fix some ungrammatical uses of "allows to"

[Eli Zaretskii]

> From: Sean Whitton <spwhitton@spwhitton.name>
> Cc: emacs-devel@gnu.org
> Date: Sun, 01 Sep 2024 11:20:51 +0100
> 
> Okay, thanks, I went through and edited them all again (not the gnulib).

Thank you.