🏆 I provide private lessons on Emacs, Linux, and Life in general: https://protesilaos.com/coach/. Lessons continue throughout the year.

On nuance, grammar, and improving one's style

The following is an excerpt of an exchange that initially was about the capitalisation of proper nouns but ultimately developed into a commentary on matters of style and grammar. I have omitted the personally identifying information and am sharing this with permission in the hope that others will find it useful.


[Prot edit: Here the person is quoting from a file I maintain (Emacs-related, but don’t worry about it).]

While on it, you might want to also guard against writing sentence fragments instead of sentences. I often see it in your texts, including in Emacs, but nowhere else. For instance, take the following two “sentences” from C-h v modus-themes-completions RET:

While Helm, Ivy et al. appear slightly different than their original looks, as they are toned down a bit.

Whereas the other group of packages will revert to an even more nuanced aesthetic with some additional changes to the choice of hues.

With regard to sentence fragments: that is usually intentional because I do not agree with the standard guidelines for grammar. It is not how we speak and I find them forced in some cases. I treat language as a medium of communication first and foremost, where what matters is the message being conveyed: for as long as you understand what I am stating, then I see no major problem. Not to give the wrong impression though, as I do follow rules of grammar albeit not as faithfully as one would expect.

[Prot edit: I provide an example.] Socrates is human. Athenian in origin. A philosopher. Father of seven children.

If you submit this in class you will fail. I disagree because I do not treat each sentence atomically but only in its context. Yes, “Athenian in origin” does not stand on its own: it does not make sense once you remove it from its context. Why would you remove it though? It was meant to be read as part of a gestalt form. The context cannot be ignored.

The above example is a perfectly intelligible sequence of statements as we combine them in our mind and thus understand that all relate to Socrates. Not that I write as in the example, as I find it a bit too terse. Just noting that I take a macro view of language instead of scrutinising it clause-by-clause, sentence-by-sentence. There are cases where the micro approach is necessary and I do use it when I must.


Interesting! Thank you for taking the time to explain your view of the English grammar. I see your point. You say “I see no major problem,” and I agree. Still, I see some minor problems. For example, if I wanted to quote you, I have to grab multiple sentences. For another example, a blind person will hear a weird intonation when their screen reader tries to read the fragments. For yet another example, some people will pay with their time to back-trace and re-parse on the first read, thinking they misread the sentence somehow. (My case. Every single time.) Nothing major, but I also fail to see any significant advantages of the practice.

Speaking of the English language, how about a fun challenge for you? When doing a live video, do not say the word “be” in any form. Doing so forces one to speak in active voice. In turn, one often must know more about the topic at hand. I have no idea if you like these sort of challenges, but if you do, I would love to see it!


[Prot edit: my reply.]

For example, if I wanted to quote you, I have to grab multiple sentences.

I think that is an unintended net positive because it make misquotes more difficult. Though this is not a strong point I am making and it should not be read as an excuse.

For another example, a blind person will hear a weird intonation when their screen reader tries to read the fragments.

I was not aware of this. It seems crucial then! Is there some setup I can try to test how a screen reader works? Documentation on how to do it should suffice. I want to improve things for that use-case and am willing to change my ways for it.

For yet another example, some people will pay with their time to back-trace and re-parse on the first read, thinking they misread the sentence somehow. (My case. Every single time.) Nothing major, but I also fail to see any significant advantages of the practice.

This is an issue as well. I will try to improve things.

Speaking of the English language, how about a fun challenge for you? When doing a live video, do not say the word “be” in any form. Doing so forces one to speak in active voice. In turn, one often must know more about the topic at hand. I have no idea if you like these sort of challenges, but if you do, I would love to see it!

I had never thought about that before. I know that the use of passive voice is discouraged, but I am not conscious of how often I express myself in those terms. As already implied, I don’t treat it as wrong in my writings, but that is due to how I approach grammar, so maybe I am not aware of the scale of the issue and thus cannot anticipate how such a challenge would go. I will have to try it during some live stream, but then we need some means of measuring my performance: I usually am reluctant to rewatch the video…

By the way, I have tried to incorporate artbollocks-mode or writegood-mode in my workflow, while I have flymake-proselint set up, though I seldom follow their suggestions [Prot edit: those are programs that help you with prose]. To rekindle my memory, I just activated writegood-mode. It should help surface issues in my current message:

  • In the phrase “should not be read as an excuse” the “be read” comes with a warning. Sure, I could state it as “Though this is not a strong point I am making and it is not an excuse”, but such a change alters the meaning because in the first case I am advising you not to interpret it as an excuse whereas in the second I am telling you that it is not. I wanted to convey the former meaning.

  • The “is discouraged” in the phrase “I know that the use of passive voice is discouraged”, is another warning. Again, I can reformulate the statement such as “I know that we should avoid the passive voice” but that rewording changes the exact meaning (as with the previous example) and feels awkward in the sense that we are following the letter of the law without understanding the spirit of the law.

To be clear: I am not arguing against your points. I am genuinely curious to learn more about this and am willing to improve my style, though the packages/programs that claim to assist one’s writing tend to enforce a less expressive style which is not a clear improvement in a multitude of cases (I wanted to write “in many cases” but apparently “writegood-mode” says it is wrong, as if you will definitely not understand what I am saying, so I used the more awkward “in a multitude of cases”).


[Prot edit: the reply I got.]

Great points and discussion!

In the phrase “should not be read as an excuse” the “be read” comes with a warning. Sure, I could state it … “it is not an excuse”.

You replaced one “to be” with another “to be”.

For fun, ask Emacs to fontify all occurrences of “to be”, namely “am”, “is”, “are”, “was”, “were”, “been”, and “being” [Prot edit: “fontify” means to have the program highlight those occurrences]. Then try to write some thoughts. You will struggle to no end, but it will sharpen your writing and thus your thinking.

Important: Instead working around the passive voice and “to be” syntactically, focus on semantics. Identify appropriate actions, actors, objects, quantities, and the like.

Consider your case:

| passive    | "it should not **be** read as an excuse" |
| active     | "it is not an excuse"                    |
| no "to be" | "it does not excuse me"                  |

As for the word “many”, it has low precision.

Consider your case:

| imprecise, **100 or 11,000 out of 12,000** | "in **many** cases"   |
| more precise, meaning 50+ precent          | "in most cases"       |
| even more precise, about 90+ percent       | "in almost all cases" |
| precise and also readable                  | "almost always"       |

The writegood-mode rejects the first one and rightly so.


[Prot edit: my reply.]

For fun, ask Emacs to fontify all occurrences of “to be”, namely “am”, “is”, “are”, “was”, “were”, “been”, and “being”. Then try to write some thoughts. You will struggle to no end, but it will sharpen your writing and thus your thinking.

I will give it a try. I already added writegood-mode to my workflow.

Consider your case:

| passive    | "it should not **be** read as an excuse" |
| active     | "it is not an excuse"                    |
| no "to be" | "it does not excuse me"                  |

Thanks for making this crystal clear! The problem, then, is with how I think (or over-think) because I still believe that the passive voice is closer to my intent in that specific context. I meant “it is not an excuse, though you can read it that way (and that ultimately does not matter)”. To my mind, the active voice here means “trust me and take it for what it is (because it does matter)”. The third variant has a different meaning which is technically correct though not exactly what I wanted to say.

Yes, there are contexts where the passive voice obscures the meaning and I admit that it is easy to unintentionally cause confusion by the wanton use of the passive voice or some other form of expression. My point is that clarity/obscurity is not an innate property of the passive voice, but of the wider context: clarity/obscurity is an emergent property. The macro view of what is being communicated cannot be ignored (writegood-mode has two warning for this sentence, by the way, even though this is an abstraction and the passive voice does not detract from its meaning).

By enforcing the rule without accounting for the context, we are risking the loss of nuance. I understand that “nuance” may be what those rules want to diminish or eliminate so as to improve the clarity of the given statement, though one must consider the hidden cost of such an endeavour. I am not prepared to provide assent to the notion that nuance is necessarily undesirable, at least not without considering the context. Thus, I ultimately am wary of the decontextualisation of linguistic patterns as it alters the constitution of the case under consideration by essentially changing the meaning.

As for the word “many”, it has low precision.

Consider your case:

| imprecise, **100 or 11,000 out of 12,000** | "in **many** cases"   |
| more precise, meaning 50+ precent          | "in most cases"       |
| even more precise, about 90+ percent       | "in almost all cases" |
| precise and also readable                  | "almost always"       |

The write-good-mode rejects the first one and rightly so.

I am aware of this and feel that we are victims of following the rule to the letter while losing sight of its spirit. Imprecision is exactly what I wanted to communicate. Let’s think of it this way: if I had opted for any of the other three styles I would be implying that I am confident/certain in what I am stating and that, presumably, I can back that up with hard numbers. Whereas I was being dubitative: I was expressing myself on a topic I have not researched in detail and for which I only have anecdotal evidence (i.e. my own experience), to the effect that I was forwarding an opinion that came with the tacit proviso of “maybe I am wrong because I do not know the extent of the issue”.

In other words, my nuanced formulation contained meaning (my dubitativeness) that we would lose if we were to follow the rule to the letter.

[We should not forget that I am not a native speaker and may still make mistakes; mistakes which I hope to learn from. Though note that I have similar opinions on grammar/style for my mother tongue.]


The macro view of what is being communicated cannot be ignored (writegood-mode has two warning for this sentence, by the way, even though this is an abstraction and the passive voice does not detract from its meaning).

I like it. You say “what is being communicated cannot be ignored”, and I ask: Says who? So what? For me, writing or speaking in E-Prime, which goes beyond the passive voice with no “to be”, often reveals weak ideas, weak or missing sources, and more. Of course, I find it hard. Then, I also find it interesting and fun, hence the challenge!

Imprecision is exactly what I wanted to communicate.

Then, “some” would do. The word “many”, like the word “very”, often means nothing. These words add a false sense of precision, and weak sources use them for that reason. The less research, the more “very” and “many”.

P.S. The two sentences below, for example, show clear writing and thinking. They contain no weasel words, no passive voice, and they also constitute E-Prime. To me, they read like butter (sans the unnecessary “then”). Well done!

If you agree, then I can prepare a draft and show it to you for a final confirmation. I think we can omit the details about the themes and just cover the rest of the discussion.


[Prot edit: my final reply.]

The macro view of what is being communicated cannot be ignored (writegood-mode has two warning for this sentence, by the way, even though this is an abstraction and the passive voice does not detract from its meaning).

I like it. You say “what is being communicated cannot be ignored”, and I ask: Says who? So what? For me, writing or speaking in E-Prime, which goes beyond the passive voice, often reveals weak ideas, weak (or missing) sources, and the like. Of course, I find it hard. Then, I also find it interesting and fun, hence the challenge!

Here I should take a step back to note that I am making an argument for an edge case. I do not advocate abandoning grammar or disregarding rules of style/semantics altogether. It just so happened that some statements made for good examples in favour of an eclectic approach that allows for case-by-case variations.

Imprecision is exactly what I wanted to communicate.

Then, “some” would do. The word “many”, like the word “very”, often means nothing. These words add a false sense of precision, and weak sources use them for that reason. The less research, the more “very” and “many”.

Understood.

As for the screen readers, I have used proprietary ones before my re-conversion to open-source.

Fine. I want to study this further.

P.S. For a good example, the two sentences below showcase clear writing (and thus clear thinking). They contain no weasel words, no passive voice, and also constitute E-Prime. To me, they read like butter (sans the unnecessary “then”).

If you agree, then I can prepare a draft and show it to you for a final confirmation. I think we can omit the details about the themes and just cover the rest of the discussion.

I agree and have already committed to improving my style. The “then” should not be there.