In English, for example, the word "don't" is made up of 4 letters ("d", "o", "n" and "t"), and one punctuation mark ("'").

However, there seems to me to be no reason for this distinction. Without any of the letters, the word is incorrect, and without the punctuation mark the word is incorrect.

The same applies to "light-hearted", which, without the "-" is not a correctly written word.

I gather that in some languages, for example Spanish, the "n" is a different letter to "ñ", whereas in English that would likely be considered different punctuation (e.g. "naïve" and "naive" are generally considered the same word, and are both correct).

The field is even more muddied by symbols such as "œ", which are not in the alphabet, but also aren't really punctuation!

What causes the distinction between letters (part of the alphabet) and punctuation (other symbols, required as parts of words but not in the alphabet?) - and why aren't "'" and "-" part of the alphabet? Are there written languages which don't make this distinction at all?

  • 1
    This question is based on a false premise. Apostrophes and hyphens in written words are not punctuation marks. Punctuation is separate from written words.
    – Rosie F
    Commented May 3, 2020 at 6:27
  • 3
    Letters correspond in some way to phones. Punctuation doesn't. That's the basic difference, though I'm sure you could find some exception.
    – curiousdannii
    Commented May 3, 2020 at 8:14
  • 2
    @curiousdannii were and we’re sound different due to the apostrophe, for example.
    – Tim
    Commented May 3, 2020 at 12:17
  • 1
    Spelling conventions re: hyphenation are rather arbitrary, cf. light hearted (Defoe) or lighthearted (Collins).
    – Alex B.
    Commented May 3, 2020 at 12:41
  • 2
    @Tim No, they sound different because they're different words. Them having different spellings isn't the cause of that. Otherwise homographs wouldn't exist!
    – curiousdannii
    Commented May 3, 2020 at 12:48

3 Answers 3


Good question!

The distinctions made between all these different components of a writing system are observed by linguists and then given names.

The defining feature of a punctuation mark is that it's purpose is to aid the reader, traditionally when reading aloud. For instance, the spaces we use between words in English is a punctuation mark since it acts as a delimiter or a tokenizer for each word. Without it, reading English would be significantly more difficult.

The "correctness" of a written word is a convention that is collectively decided by our language community through usage. For instance, the hyphenation of "light-hearted" is intended to help aid the reader, but if we all decided to write it without the hyphen, that would become the convention. Conventions can change, but they change much slower for writing systems than they do for spoken word!

The symbol "œ" is a letter that is used in languages other than English. The best way to describe it is a grapheme, which is basically a building block of a writing system, or orthography. Everything we have discussed so far is a grapheme. You can think of graphemes as the electrons, neutrons and protons of writing systems!

This analogy works well because there are different kinds of graphemes, too, such as punctuation and alphabetic letters.

The tilde (that's the "~" above the "n") in "ñ" is a diacritic, which is just a punctuation mark that is added to an already existing character to give it additional meaning, usually appearing above or below a letter. They are not limited to Latin, either.

Diacritics in Abjad writing systems

  • Are there any languages that dont use spacing? It seems so useful!

    • Japanese orthography includes no spacing and no capitalization, but readability is achieved through contextual usage of their different writing systems: hiragana, katakana and kanji.
  • What about numbers? Like "1" and "6"?

    • Those are logograms. It's worth noting that there are other numeral systems, like in Chinese where one can use "一", "二" and "三" for "1", "2" and "3" respectively. Do you see how the number of lines in the corresponds to the quantity it describes? Those characters in particular could be called ideograms.
  • Each word of a sentence in English, can be classified into different PoS. Is there any word, symbol or so, that cannot be classified so classified? When we say parts of speech, does it refer only to speech, and not writing? I have always thought which part of speech a punctuation mark is.
    – Ram Pillai
    Commented May 11, 2020 at 2:55
  • 1
    @RamPillai It's important to remember that all such categorisations come after the fact - we don't learn them when learning to speak or read, only when trying to describe the system. There's also no universal and comprehensive list of such categories, and the exact definitions depend on what aspect of language you're trying to discuss. As such, it doesn't really mean anything to say something "cannot be classified" - a linguist might decide that a word doesn't fit any of their current categories, but they can simply invent a new category to put it in.
    – IMSoP
    Commented Jul 10, 2020 at 11:21

It is kind of a historical artefact.

Letters (or signs for syllables, consonants, or words) came first historically, everything else, including word separators or spaces between words, came much later by centuries or even milennia.

So your last question can be answered first: There are languages written completely without punctuation and they include Classical Greek and Classical Latin as written in antiquity (modern editions add word spaces and punctuation to help the modern reader, but the original manuscripts from antiquity don't have them).

Historically, punctuation was first added to religious texts to preserve their exact pronunciation when the original language changed and to avoid ambiguities. We can often find several layers of punctuation in religious texts: Vowel marks or diacritics for pronunciation, and a second layer of cantillation marks to guide reciting the texts with the correct intonation.

Another thing are abbreviatures: To save time (of scribes) and space (on expansive parchment) a lot of abbreviatures and shorthands are introduced in writing. They were mostly given up after the introduction of typesetting and printing but some artefacts of that are left in our present day writing system, e.g., the ampersand sign &.

  • 2
    I'd distinguish between diacritics which mark phonetic information and punctuation which at most marks pauses - I don't think they're really very similar.
    – curiousdannii
    Commented May 6, 2020 at 8:44
  • 1
    @curiousdannii While that's a useful distinction, I don't think it's quite as clean as you imply. We wouldn't normally call "-" a diacritic, but in "co-operate" it marks a phonetic distinction (which might otherwise be written "coöperate"). Meanwhile, "!" and "?" may well influence the intonation of the word or whole phrase preceding them; something which in other contexts might be indicated by diacritics on particular syllables.
    – IMSoP
    Commented Jul 10, 2020 at 11:29

From a modern European point of view, it's tempting to see the concept of an "alphabet" as a universal aspect of writing, and that writing is inevitably made up of "letters" and "punctuation".

This is because we're familiar with the Latin, Cyrillic, and Greek alphabets, which are actually closely related, and have been adapted for different languages and contexts over thousands of years. They thus all share a basic structure:

  • A fixed list of core symbols for representing consonant and vowel sounds, which we call "letters".
  • In some languages, a set of reusable markers for creating additional letters by modifying this core set, which we call "diacritics".
  • A separate set of conventional marks added to those core symbols to represent other aspects such as pauses, emphasis, or intonation, which we call "punctuation".

If you look further afield, you will find many writing systems which can't be described in this way:

  • In an "abjad", the core set contains only consonants, with vowel sounds optionally represented by modifying marks (e.g. Hebrew, Arabic)
  • In an "abugida", the symbols in the core set each represent a consonant plus a vowel (e.g. Indic scripts)
  • In some systems, characters for syllables are systematically built from a set of smaller components (e.g. Korean Hangul)
  • Different languages need to encode different dimensions of speech, so some use markers for stress (e.g. Spanish), aspiration (e.g. polytonal Greek), or tone (e.g. Vietnamese). Sometimes, these are optional, but sometimes omitting them is as incorrect as omitting the dot on a lower-case "i" in an English text.
  • In an ideographic writing system, each symbol represents an entire word or concept (e.g. Chinese)
  • On the other hand, the separation between words may not be indicated by the script at all (e.g. Thai)
  • In many languages the system theoretically represents sounds, but doesn't do so systematically (e.g. English!)

These aren't neat categories, either; many writing systems have evolved over time, and under the influence of other systems, to have combinations of these concepts. Terminology has then been invented to describe how they work, both for native users learning to read and write them, and for linguists to compare them.

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.