First post. Wanted to title it "Speaking in tones," but that's not very informative.

Long ago, I learned a little about talking drums and whistle speech as long-range communication tools based not on direct semantic coding (Morse, e.g.) but on tonal and rhythmic paraverbals of natural language. (There's probably a better way to express that.) Now there are videos demonstrating such wonderful languages in use.

More recently, I got to wondering how this might play out when linguistically sophisticated speakers of Asian and other tonal languages converse with non-tonal language speakers in the latter's tongue - say, an educated, multilingual native Chinese speaking English with a monoglot American. For example, might the Chinese speaker

  • Flatten tonality generally

  • Follow the tonal syntax (and semiotics, if any) of native speaker

  • Adopt a group defined/defining tonal system

  • Make up (consciously or otherwise) an idiosyncratic tonal system

  • Use tonality in ways that the English speaker may not even sense, but that helps the Chinese speaker with his/her own feelings/expression

  • Use Chinese tonal cues as a covert way to express (for speaker's own benefit) insult, praise, feelings

  • do., in ways that carry meaning only to other tonal-language speakers who may be present

Et cetera. A unifying question might be: When speakers of tonal languages speak non-tonal languages to native speakers of the latter, do they make any special use of tonal inflection, and/or make any special effort to suppress tonality? Or do fluent multi-linguals tend to compartmentalize the "tonality" function as needed?

Not looking for help on a term paper or any real-life situation; just curious. Thanks for any insights!

  • The Q seems to assume that the dimension of pitch is unused in non-tonal languages, whereas in tonal languages it serves to provide 'paraverbal' information. I don't think it works like that. First of all, while lexical tone obviously doesn't influence the final pitch contour in a non-tonal language, that doesn't mean that the pitch channel is clear - it's needed for stress and intonation. Secondly, the function of lexical tone is not to finesse meaning - tones are not a more sophisticated form of intonation but are as integral a part of the syllable as consonants or vowels.
    – rchivers
    Commented Sep 7, 2020 at 20:23
  • From that POV it's not surprising that, IME, speakers of SE Asian tonal languages have a very hard time learning words without attaching tones to them. This means that, when they are speaking a non-tonal language, they constantly have a spurious pitch input that can only mess up their stress and intonation. I feel that some people make an effort to suppress tonality but this is virtually always a failure, so would say that the tonal patterns of their L1 tend to come through and clash with the stress and intonational patterns of the non-tonal language.
    – rchivers
    Commented Sep 7, 2020 at 20:29
  • Thanks for these comments, rchivers. I did not mean to imply that pitch is unused in non-tonal languages or is separable from other aspects of vocalization in tonal ones. My question is whether the tonal speakers make some conscious or unconscious semantic use of tone whether or not the listener can understand this. Your second comment suggests that tonal speakers may find it as hard to avoid this "noise" (as perceived by non-tonal listeners) as non-tonal speakers find learning tones in, e.g., Chinese. Very interesting.
    – cTen
    Commented Sep 7, 2020 at 21:31
  • I'd still like to know if some tonal-language speakers use un-needed tones to communicate in code when the overt message is non-tonal. Seems like a nice gimmick for a spy thriller. Or maybe actual spies use this already?
    – cTen
    Commented Sep 7, 2020 at 21:36
  • Well let's see how the thread develops, but I can't make sense of the idea of a semantic use of tone, or of tones being a covert channel for a message that is superimposed on the literal meaning of the words. To me the tones are part of the make-up of the words and don't have any more meaning in themselves than the vowels or consonants. I suppose you could contrive a system where conspirators mispronounced the words of another language in pre-defined ways to embed hidden information, but although speaking a tonal language might make that a bit easier, it's basically possible for any language.
    – rchivers
    Commented Sep 7, 2020 at 23:11

2 Answers 2


There's an interesting answer when it comes to English as spoken by Cantonese speakers from Hong Kong, where tone can even distinguish words that would be homophones in English.

Generally, stressed syllables in English will come out with tone 1 (contour 55), which is high level tone. Unstressed syllables are of two types: final unstressed syllables end up with tone 4 (contour 21), and other unstressed syllables end up as tone 3 (contour 33) or tone 6 (contour 22).

As an example, the English word reminder will end up as [wi˧ majn˥ dɑ˨˩]. A longer word like qualification would be [kwɔ˨ li˨ fi˨ kej˥ sɵn˨˩].

That said, certain English words which would be homophones in other pronunciations can be distinguished in HK English solely by tone, for example for [fɔ˨] versus four [fɔ˥]. The reason for this particular pair should be clear, given that one is a (typically unstressed) preposition. The special thing about HK English would be that the distinction between the two would be maintained in HK English even if for occurred in a stressed context, for example "I am for and not against."


IMO, we do not know a general answer, but anecdotally I believe that the intonational system of English used by second language speakers whose first language is tonal is non-randomly related to the tonal phonetic of their first language, thus the English of a Logoori speaker and that of a Kamba speaker will be intonationally different. However, this influence is diminished as a function of increasing fluency in English. In other words, they suppress influences from their native language, because that is not how English speakers talk.

I should also point out that speakers of non-tonal languages can also have what looks like a tonal overlay in their English, as in the case of Korean. Standard Korean is non-tonal, but Koreans speaking English exhibit (at least initially) a strong influence of Korean intonation, which makes their English "sound" more tonal that e.g. speakers of German. Having interacted with speaker of standard and Kyeongsang Korean (the latter is a tonal dialect), I can't say that there is a noteworthy difference in their respective dialects of English. It is therefore not easy to say to what extent an observed second language effect is because of the tonal system of the first language (if it has tone), or the intonational system (all languages have one). The niche that would have to be studied to pin this down is tonal vs. non-tonal dialects of a single relatively uniform language.

  • 1
    Much obliged for these informative insights. I didn't include in the OP one of the motivations for this question: Many years ago, I attended a workshop by a linguist who'd worked in intelligence. That person demonstrated a simple yet subtle way to encoding information and recognition signals in normal syntax and semantics. I wondered how many other channels there might be for conveying signals (even at low bandwidth) that might lie outside the notice of most people, and verbal steganography via situationally irrelevant but not clashing use of tonality seemd like a good way. Thanks!
    – cTen
    Commented Sep 12, 2020 at 23:12

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.