I've had an idea recently to use color in order to disambiguate several usages of the same word in a text which refer to two different instances of the same concept. Since StackExchange does not allow use of colors in formatting, in the reproduction of text below I use bold and italic fonts instead:

When you start to critically think about thinking, it's easy to start doubting your thinking the same way that you doubt your thinking.

This doubt makes critical thinking about thinking an antimeme (an idea with self-censoring properties): it makes it harder to develop the thought, and it makes it harder to communicate it.

I assume that I'm not the first person in the world to come up with that idea: is there a prior art of using formatting for disambiguation? I am looking for particular examples as well as analysis of this technique.

  • 6
    You would have to be very careful with colour selection to avoid making the text inaccessible to the colour -blind.
    – Chenmunka
    Commented Oct 15, 2022 at 20:55
  • 1
    I haven't seen any with colors, but definitely have seen some with different fonts. Most famous one is the summoning chant for Cthulhu. Probably you can play around that? In some games it's also sometimes used to denote "this is some otherworldly words that mere mortals can't comprehend or even see"
    – justhalf
    Commented Oct 16, 2022 at 6:48
  • 1
    @justhalf, do you have some particular examples of that "otherworldly words" case I can take a look at? Commented Oct 16, 2022 at 9:42
  • I don't have specific examples right now, but will let you know if I can find one.
    – justhalf
    Commented Oct 16, 2022 at 12:05
  • 1
    my copy of "The Never-ending Story" by Michael Ende has two colors: red and green... one to indicate things are happening in the "real world" and other to indicate things are happening on Fantasia... just a swift FYI as it does not answer your question but came to mind when reading this post.
    – DarkCygnus
    Commented Oct 17, 2022 at 20:16

8 Answers 8


That sounds confusing, no matter the formatting. It might be clearer to stick to longer descriptions (e.g. "critical thinking" as opposed to "the act of thinking") or to pick a synonym (e.g. "reasoning", "contemplation"), define it once and then consistently use that.

  • I agree that it might be possible to disambiguate this text somehow differently, but I do not agree that writing out longer descriptions is necessary a better way to do that. When reading more verbose descriptions reader has to maintain longer context and might be lost more easily. Commented Oct 15, 2022 at 21:23

Never use color alone to indicate anything that matters.

Some people are colorblind, and will be unable to tell apart colors that may look entirely different to you. The most well-known pair of such colors is red and green, but other forms of colorblindness may affect other combinations of colors, and so it is safest to assume that any two colors might be confused. While color can sometimes be used to help indicate things, it should never be used in isolation.

In philosophy, I have occasionally seen multiple meanings indicated with numerical subscripts (so you might write think1 and think2, which can be done with <sub></sub> tags in Stack Exchange). However, this convention is rare in other fields, and would likely need to be explained. Also, it's used for differentiating meanings, rather than instances of the same meaning, so I would avoid this usage for your precise use case. Instead, it's often best to add words or recast the sentence so as to remove the ambiguity. This will usually require you to write more verbosely, but clarity should usually be prioritized over brevity. For example:

When you start to critically think about thinking, it's easy to start doubting your thinking-about-thinking the same way that you doubt your thinking.

This doubt makes critical thinking about thinking an antimeme (an idea with self-censoring properties): it makes it harder to develop the thought, and it makes it harder to communicate it.

I have boldfaced the only change to make it obvious, not to suggest that you would actually set it in bold in your manuscript. Note that the second sentence did not need to be modified at all, because it is already obvious which sense is meant.

  • Do you have concrete examples of philosophical texts using subscripts? I would be interested to take a look at those. Commented Oct 16, 2022 at 9:34
  • Instead of a sub element, you could also use the underscript numerals such as and (not to mention ₀₃₄₅₆₇₈₉ - and , , & ). The real characters work better when text is copy-pasted, for example. Commented Oct 16, 2022 at 17:35
  • 2
    @AndrewShulaev: This appears to be less common than I remembered. I did find some use of subscripts in this article, but it was in relation to pronouns and their antecedents, not to meanings and senses.
    – Kevin
    Commented Oct 16, 2022 at 18:17
  • @TobySpeight: The position of the Unicode Consortium is that those characters basically only exist as a compatibility hack, so any software that does "compatibility" normalization may silently convert them into regular numerals and break your formatting. If you are writing in a rich text environment, you should use the formatting features of that environment, and not try to coerce everything into Unicode characters. Copy-pasting should preserve rich text if the destination supports it.
    – Kevin
    Commented Oct 16, 2022 at 18:26
  • That sounds lovely in theory. Sadly, in practice, most transfers lose the markup (e.g. when copying into a new Stack Exchange answer or into email). The only target I've seen that works from rendered HTML is Libreoffice. I've certainly not found anything that converts subscript markup into subscript characters for plain-text targets. Whereas using subscript characters from the start works pretty much everywhere. I'm looking forward to better support for formatted text transfers! Commented Oct 17, 2022 at 7:14

Software documentation makes heavy use of alternative fonts to distinguish words used in a normal English sense from words used with a programming language meaning. I haven't seen colour used for that, though I have seen background shading.

On web pages, of course, colour is often used to signal clickable hyperlinks.

Missals (used to contain the text of a liturgical service such as the Catholic Mass) and Orders of Service often use text in two colours (black and red) to distinguish the actual text of the service from commentary or instructions (c.f. stage directions), or to distinguish text spoken by different participants. Red letter bibles use red text for the words of Jesus. But this is expensive, and font variations are a more common device.

  • 1
    Colour is of course quite common when the documentation includes pieces of source code, in the form of syntax highlighting Commented Oct 16, 2022 at 20:55
  • 2
    @SE-stopfiringthegoodguys: But the color is descriptive not prescriptive -- it doesn't cause the words to take on a different meaning, it merely reflects the meaning they have. OTOH a change in font, or shaded background, does designate a different meaning.
    – Ben Voigt
    Commented Oct 17, 2022 at 20:01
  • 1
    The difference there is that the documentation you're referring to is distinguishing things of very different kinds — source code from narrative. The question seems to be asking about different instances of the same kind of thing, so it's perhaps not comparable.
    – gidds
    Commented Oct 17, 2022 at 21:14

If you're tempted to do this, one thing to consider is whether this is even feasible to do with the way you intend to publish your text (cost of colour print, style requirements of a periodic, and so on). This won't be an issue if you're self-publishing on your own website, but most other ways, it's likely it will - and even on your own website, if there's anyone who'd like to print your work out, then you're setting them up for running into an issue.

Thus, colour-coding is quite an impractical method at least, and may turn out straightout impossible.

I'm afraid I don't recall any instance of colour-coding text in non-fiction. In fiction, the one I know of is the Neverending Story, where it isn't meaning of the words but whole storylines that is colour-coded. Consequently, the manufacturing cost of the book is fairly high, and so is its price in the bookshop.

Programming editors often colour-code parts of the code for viewer's convenience, but the colour isn't part of the actual code, it's automatically added according to what each part of the code does, and bears no additional information at all.

For other formatting options, bold and italics are used in sciences to denote physical quantities, sets, vectors and so on, both in equations and when the symbols are used in a sentence - where it quite helps to tell the difference between, say, "a" (indefinite article) and "a" (acceleration).

I don't recall any usage for altering the meaning of terms. Admittedly, there would probably be more of a demand for such a thing in humanities (such as your example), which I'm simply not familiar with.

What actually comes to mind as most similar to what you're trying to do is (fiction, again) the distinction between animals and Animals in the Chronicles of Narnia, where lowercase ("a horse") denotes ordinary animals we know, and capitalised ("a Horse") denotes sentient beings populating the fantasy country. It took me three books to even notice there was a difference other than context.

And yet another option I've seen is a different font type, such as Terry Pratchett used in his Discwold books to denote that a character was speaking a different language (curly font for Klatchian). He was reportedly hated by printers for it. And in translation into my language, the Klatchian lines in the curly font substituted all letters with diacritics (about a third of all letters) with non-letter characters.

One thing you may notice about the examples I've listed is that formatting helps the reader's orientation, but it usually isn't the only way to distinguish between two different things - the reader would already have enough information to figure the distinction out without any difference in formatting. And I think there's a good reason.

Whichever formatting method you might choose to distinguish between two terms where the words are identical and the terms only differ by formatting, this will make it possible to make a distinction in written text, but not in speech. What you write will be impossible to discuss in a spoken conversation. That seems like a huge setback, so I'd strongly suggest not using formatting on its own for distinction at all, but defining differing terms in words instead, such as, I don't know, "common sense thinking" versus "self-reflective thinking".

  • Thanks for all the similar examples in fiction! As for publishing in print, I don't think it's that much of concern: it might be possible to use alternative encoding (e.g. subscripts, as suggested in another answer) for print version while using color encoding on the web. Commented Oct 16, 2022 at 9:41
  • FWIW, some editions of The Neverending Story dropped the colour-coding in favour of italicisation, much like @AndrewShulaev in his question here. A few years back, I asked about this on Literature SE (but no answers yet). Commented Oct 17, 2022 at 17:53

As mentioned by Divizna it is very atypical to color certain words in a normal book, with the Neverending Story being the outlier here (and I'm quite sure many editions would have dropped it if it wasn't referenced by the story itself). A typeface change is generally used instead where that would have been an option.

I have mainly found this in the context of different languages rather than word meanings, though. For example, a novel is written in English, where a character says some words in Spanish. Then the book is translated into Spanish. The original text in Spanish is rendered in italic to show it's in a different language than the rest of the text (a footnote explaining it the first time).

This is not only done as a result of translations, but it is also done in originals. For instance, in Asterix and the Goths (Astérix et les Goths) the speech balloons are written in Gothic font when the characters speak in Gothic, while the normal font is used when they are speaking in French Gaulois:

[Image from https://majorspoilers.com/2013/11/28/retro-review-asterix-goths-1963/

This makes both languages intelligible for the readers, while most characters actually speak a single language, with not understanding each other playing an important role.

Conversational guides also tend to make heavy use of font and colors to differentiate between the multiple languages in their texts (source language, target language, pronunciation…)

  • Actually, some editions of The Neverending Story (including the first one I read) did drop the colour-coding in favour of italicisation. A few years back, I asked about this on Literature SE but no answers yet. Commented Oct 17, 2022 at 17:53

Another strong argument against this (besides others mentioned in answers) is that our mental representation of text is based on spoken language, not written. Most well-established inline text-formatting clearly corresponds to something in spoken language: punctuation mostly signals timing, bold and italic correspond to marking emphasis with stress and tone, and so on. So following such formatting comes very naturally to a reader. Colour doesn’t correspond to anything spoken. Reading your text aloud, how would you distinguish the red word from the green? A reader has no idea, and hence no intuitive way to imagine it when reading the text; so it’s much harder for them to process the colour as part of the text while reading.

Formatting with no spoken counterpart can be useful for other things: for instance, signalling the structure of a text, or directing the reader’s attention. Colour can be good for such purposes. But for information that’s meant to be part of the textual content itself, colour just won’t work well for most readers.


It doesn't use color, but the closest example I can remember of using formatting to distinguish context is Terry Pratchett's Discworld series. The character of DEATH (as opposed to the phenomenon of death) is in all caps, as well as things directly associated with DEATH, and all of his dialogue (only DEATH's lines are in all caps, while people respond in mixed case). Part of this imagery stems from an introductory description of how DEATH is more REAL than regular reality, as the one character (or phenomenon) which is unavoidable under any circumstances.

(The DEATH OF RATS was a ancillary character, apparently because rats die at such a rate as to justify their own specialist, but its only line was SQUEAK.)


Italics is prescribed by AP Stylebook and Chicago Manual of style to use for "words as words" -- that is, when you refer to a word itself (not using it functionally), you put it in italics, or quotes if font changes aren't possible. For example:

Manners and Manors are homonyms.

This would seem to apply to the question of distinguishing words being defined.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.