Assuming human-style vocal cords, what phonemes would a snouted animal be able to produce? I'm looking at developing a conlang (or several) for a setting, and it has several humanoid species with protruding muzzles. I can think of a few phonemes that can be ruled out immediately on the basis of lip immobility (/f/, /v/, and /w/ would be right out), but beyond that I'm kind of stumped.

(For what it's worth, I looked at this question, but it didn't really give me any leads. There's also a SciAm article that talks about canine communication, but it doesn't discuss this specific topic.)

  • $\begingroup$ many animals with muzzles have highly mobile lips why would you preclude sounds that need lips? $\endgroup$
    – John
    Nov 2, 2023 at 20:49

4 Answers 4


Since it seems like you're looking for constraints to help construct your language, I thought one helpful approach would be to lay out exactly what sounds you have to work with. I've taken a chart of the International Phonetic Alphabet (IPA) from Unilang.org and annotated it to indicate the sounds which are impossible without lips - crossed out in red.

Consonants Vowels

There's another thing to consider, if we're talking about carnivores like cats and dogs - their teeth aren't like ours, so it's likely they'd have some degree of difficulty forming dental consonants, and maybe alveolar consonants (with the tongue just behind the teeth), in the way that we're used to forming/hearing them. As their teeth don't form as close a 'seal' as ours, with larger gaps between pointier teeth, it's likely there'd be a lisping or 'whistling' sort of inflection on these consonants - which would be perceived as an impediment if they were trying to speak one of our languages, but would be the normal way of producing such sounds in a language developed by/for pointy-toothed carnivores. The effect will be more pronounced for the dental (marked in orange) than for the alveolar.

All this is also dependant on the shape and size of their tongue; dogs especially have larger tongues proportionate to their mouth cavity compared to humans, so they'll probably have more issues with this than cats. In humans, an abnormally large tongue is known as macroglossia, and affects individuals with Down Syndrome - a dog's speech might share certain characteristics commonly associated with the speech of individuals with this condition.

Something I haven't quite been able to define in full is whether or not a cat or dog would be capable of producing sounds that aren't in the human IPA. One thing that immediately comes to mind is a cat's purr, and off the top of my head I do not actually know much about how this sound is produced. Whatever mechanism is involved is very likely something that a cat would incorporate into its language, given the rich variety of ways in which real cats can purr. If you wanted to work it into this chart (i.e. to make a chart of your own) I would probably put it as a trill or fricative beyond the glottal. A canine language would probably also incorporate growling, which (when imitated by a human) is probably somewhere between a voiced uvular and glottal trill.

Given that their hearing is more sensitive in higher frequencies, it might be easier for them to perceive breathy/whistling sounds that are very subtle to our ears - accordingly they might assign linguistic significance to these sounds (e.g. allowing some air to escape through the nostrils, or between the teeth). If this was the case, it might be a part of their language that humans would struggle to understand/learn. More sensitive hearing overall might also lead them to assign significance to sounds that are almost imperceptible to us, like the sound of the tongue adhering to/separating from the roof of the mouth, or the sound of the throat closing during swallowing.


YouTube Searching: Do a search on YouTube for "talking dogs" or "talking cats". With enough training, they can do pretty much every single phoneme out there, despite the immobile lips.

Allophones & Accents: That happens because of allophones, which are similar groups of sounds that form the phonemes used in a language. In fact, you see that even in humans, where people thousands of miles apart learn the same phonemes as different allophones, which makes them harder to understand, but are still possible. We usually call these "accents." There's more than one way to speak a particular phoneme. They often sound more like what a deaf person sounds like when trying to speak; they're clearly not using the allophone we're used to, but they can be understood to some degree.

  • 2
    $\begingroup$ While empirical sources can be helpful, YouTube is hardly a reliable source, particularly given that various tricks can be employed to mimic speaking ability. $\endgroup$
    – Aerdan
    Jun 25, 2016 at 21:36
  • $\begingroup$ How is it not a reliable source since the videos literally demonstrate the very thing that you're asking about being possible? Are you referring to tricks by people or tricks by the animal? $\endgroup$ Jan 22, 2022 at 20:07

I would strike a balance between understandability (for readers/listeners), effect, and realism.

It would be burdensome to read a long book with extreme accents for main characters or dominant communication forms.

An accent, or linguistic differences, can highlight differences of nationality, origin, and race, and give the readers mental cues.

Realism is helpful; it might make perceptive readers enjoy the book more. But too much realism can ruin it (IE if most words have to be deciphered, and the reader must slow down too much).

Realism: I like the other answer highlighting phoneme groups that would be difficult. I should note that difficult phonemes tend to not make it into a language, even if one can learn to say them. The USA/american "R" sound is unique and difficult to say. But most foreigners can learn to say it (though it tends to sound softer or different).

Specific examples and suggestions:

  • Introduce the reader/con-attendees to the linguistic limitations once:
  • Using a character that is still learning English/common/whatever.
  • With a character who is tired/emotional and lets their accent slip
  • Aim for lighter and heavier accent variants (different skill levels in English)
  • Main characters have a minimal accent; in writing, it's usually ignored or noted.
    • "(words here)", he said in a thick X accent"
    • Rishriku said something, but she had to ask him to repeat it twice before realizing he was asking for a nasal wipe.
  • Think of the target audience, and don't overburden them with parsing your words.
  • Use a thicker accent only at first, to set the tone.
    • IE, don't misspell words all through the book.
  • Once you've set a tone, use context, phrases, etc to highlight the difference, rather than an annoying accent:
    • One or two words, phrases, altered translations, or spacer words used throughout. "You humans always worry, ka?" / "How would you say, hrrum, I prefer to speak to you in galactic common, thmmmm."
    • Consistently forget a word, phrase, or translation. "How fast is this, eh, space-engine? (hyperdrive)" / "Hand me the turn-twist-tool" he said, waving a bag of screws (screwdriver)
    • Misused slang, "This should be a bite of cake, right?"
    • Mixed metaphors, "It's not rocket surgery"
    • Unique slang throughout (read Ender's game) - best when the meaning is easy to infer.
    • Describe it (with humor): "The Kzindog muttered a stream of consanants. She had to replay the sounds in her mind to realize he had just said "You humans all smell the same."
    • Mention it: "You should have seen it" The guard said through his thick Kzindog accent.
    • Have them describe humans: "I laughed at his pink hairless face, eyes like potatoes" / "shut up, Round-Eyes."
    • Mention or make nonhuman sounds: "What did you think I was?" he growled, "A common dog?" / "shut up and get back to your chores," the feline alien said--and Joe had to remind himself that the purring voice was not a pleased voice, and these were not earth-cats.
    • Use altered punctuation or extra consonants sparingly: "What did youu--think I wasss? An earth-thing sssnake?"
    • Use different punctuation for different languages:
      • <Do you think this human will mind if I eat his uneaten carrion?> Rishthrak growled softly to his companion, who simply gave a tail-shrug.
      • Rishthrak, can you hear me? Grekil sent his telepathic message as powerfully as he could manage while keeping focus--and his lasblaster--pointed at the criminal.
    • Mention or display cultural bias, racism/specism: "Oh, sorry. I thought all humans liked potatoes" / "The only good thing you humans brought to the galactic community was potatoes--and that's not even a human invention."

I gave one answer already, but after coming back to this a few months later, I have some other ideas that constitute a separate kind of answer.

The summary is this: keep it practical, memorable, and fun. Don't make it annoying, frustrating, confusing or difficult

Here are my thoughts on what that means and how that works, in depth...

Naming schemes = fun, practical, and accessible

  • Names are one place you can have a lot of consistency/creativity/fun:
    • They CAN consistently appear in a story, and sound unique to a culture/civilization.
    • Then naturally add cultural/species flavor/reference without having to be overt or go out of your way.
    • Live events (the CON part of the OP question): Name guidelines might be a good way to make it easy
      • easy for amateurs to participate without much effort
      • easy for RPG characters or shared storylines.
      • Easy to prepare just once
      • Requires minimal practice after coming up with a name.
      • Rule sets need not be memorized, only referenced.

Creating a "sound" for a language and its names. You probably want a practical, pronounceable, memorable and identifiable set of sounds, so people can say the equivalent of "that sounds like Klingon!" or "Is that a southern accent?"

Start with a subset of English vowels and consonants (there are about 40 phonemes if I recall, with regional variations)

  • WHY create a sub-set of English sounds:
    • Pronounceable and thus memorable to readers
    • Won't slow English readers/audience down or frustrate/annoy.
    • Lets readers/audience just enjoy it without trying to remember stuff
      • eg, that "#" means "purring chitter sound"
      • eg, prevents second-guessing or stopping to check the book reference (and ruining plot flow or event participation).

If you really want something unique/unpronounceable... go ahead, but follow these guidelines/ideas

  • If you're passionate about coming up with something unique that humans can't pronounce or remember... that's fine, but... * Come up with easy English equivalents for any repeated use/reference in written form. * Have some exposition or conversation about why-this-nickname (because the characters want to reduce pain for the same reasons as the readers).
    * Think about how names get slaughtered and translated to equivalents in reality. It's a practical necessity. I will learn to respond to a Spanish pronunciation of my name when I'm around spanish-native speakers.
  • To demonstrate by example... Try to read/memorize "Xt'grktm" versus a pronounceable version "Zet-griktam", or "Łßňĵ#" vs "Luhss-nuhjay-(chitter)".

Define some nonverbal language, and use non-English/nonhuman sounds like body language.

  • Restrict unique sounds to nonverbal cues and filler words --

  • Allows you to use these traits to add flavor/tone but not direct meaning

  • Nonverbal cues have several advantages:

    • Don't need to be remembered, consistent, looked up in a reference, or constantly explained.
    • Need not be pervasive. You don't have to over-describe or over-use it, because it's not intrinsic to the langauge.
  • Non-verbal and non-auditory cues.

  • For con participants, these are probably easier to adopt and sprinkle into interactions, and don't have to be as consistent.

  • For readers, this can be a good way to sprinkle in some sense of alien/difference without overdoing it

  • Examples:

  • Simple sounds: purr, growl, chitter, whistle, clicking, rasp, rattle

  • Variations add flavor: soft purr, low growl, high-pitched chitter

  • Remember to add variations/combinations to the unique sounds.

  • Combinations add flavor/complexity later on: growling chitter, singsong purr, rasping whistle, rattle-growl, etc.

  • Invert/change the meaning: Just like the real world, nonverbal cues can be misjudged or misunderstood, like how the thumbs up is like a middle finger in certain countries, or pointing can be done with chin/lips, index finger, or the whole hand (and each can be offensive/quirky in the wrong countries)

    • e.g. a purr actually indicates anger/negativity in a feline species, and it makes frustrating situations worse between them and humans.
    • Only use this in long-form fiction where you can establish the context, and maybe even use it as a plot device for/against protagonist/antagonist/others.

Language and culture are linked, and are never homogenous

  • Remember: Language, culture, and nonverbal cues are all linked.
  • Avoid "planet of the hats" or having a single rule set
  • Make sure to identify one or two variants of language/pronunciation
  • These variations come from sub-cultures, continents, dialects, sub-species, or alternate belief systems for a race/world.
  • Don't overdo it--it should give variety but still be memorable and fun, allowing audience to easily remember the implications rather than giving them a burden of remembering/reminding/frustrating details.
  • If names are going to be part of the dialogue, or at all important, simply having those names appear may be enough reminder of language differences.

HOW to select a subset of English sounds - some ideas

  • You can be arbitrary--just pick a consonant or sound and strike it out. Not everything needs a reason.

  • Also be logical. Think about biology, morphology, nutrition, and cultural influences on language and tone

    • IE, decide that the vowel sounds "O/oo/ou/w" are difficult for feline inexperienced speakers (requires lip shapes that muzzles don't have).
  • Don't be absolutist - think about variation & difficulty

    • Decide which English sounds are hard
    • Do not exclude these sounds entirely--but soften the sounds and reduce their frequency based on how difficult they are.
    • e.g. muzzle mouthed: any labial and dental sounds are hard, but not impossible
    • Remember that most sounds can be imitated very closely in alternate ways. The dental sound "t" is hard for canine/muzzles, but easy enough to move to the tongue & palette -- sounding closer to a "d" but not alike. This is easy enough, in fact, I wouldn't even exclude it from the language.
  • Be logical about skill variations (fictional character skill and participant skills)

    • Characters/actors with less actual skill/practice can be the "more experienced English speakers" without worrying about messing up too much.
  • Come up with some common names or naming patterns from sub-cultures/tribes/backgrounds. In English, you have O'hare/O'donnell, Robinson/Anderson, Swensen/Jensen. In Hawaiian, they have the glottal stop (heard in uh-uh and noted by apostrophe in Hawai'i); many pacific island languages have far fewer consonants because of dietary/malnourishment over generations - reducing dental sounds mostly.

  • Decide how names are constructed and used. Do people go mostly by surname, tribe? Are there common endings for male/female or other identifying morphological/familial traits? Avoid having a single answer for any of these.

  • Muzzles, specific examples/ideas:

    • If the muzzle-mouthed crowd are mostly from normal human cultures or heavily influenced by them, this should give less weight to any pronunciation differences--focus on softening or slightly altering some vowels and consonants. This is also easier on reader/actor/audience, and gives an out/explanation for those whose pronunciation is more English.
    • For casual actors/participants, find simple rules. IE, say everything while making a 😬 grimacing/smiling/toothy mouth shape with cheeks stretched & frozen. It's much easier than learning any complex rules.
    • Experiment with talking without moving your lips much. Identify which vowels and consonants come out differently or not at all. Try to imagine how they might compensate if required to do so (you can make most sounds using other means, or close enough). Note which ones require more effort and practice. For example, if you have sharp teeth, "th" might have to be produced using the soft palette--it can be done, but it sounds slightly off. Think about how ventriloquists sound (especially amateurs) -- their characters often have a certain accent because their lips aren't moving and they are maintaining a smile
    • Have mercy on amateur/community actors' mouth muscles. Think about this carefully.

I realize I am answering partly from a written context, but remember that a con-lang will be written as well, many rules are the same, and many wb.se users are into writing--so answering for them too.

Every rule probably has its exceptions, of course. You have to know your audience, subject, tone, and intent. If you intend for the audience to be confused during a short skit, and to listen only to conversational tones and impressions, then by all means, do what you need to do. Long interactions rules will be different than short ones.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .