Here's a world much like Earth, except it's inhabited by a race of shapeshifters. On one side of the appearance divide there is an appearance much like a large Earth wolf; on the other is an appearance much like an Earth human.

The actual mechanics of the shapeshifting process are undefined, and I'm willing to do a little fudging there (but not really to the point of just saying "use magic"), but both forms should keep an exterior appearance actually resembling their respective species. So no humanoids with a large muzzle, nor brachycephalic wolves.

It's not too hard to find references to IPA charts, but I'm having trouble applying those to my particular question.

So basically, if we're allowed to make the internal anatomy as that of either a wolf, a human, something in between or something completely different, but are constrained by the external apperance of a human and a wolf and to the extent possible would like for them to also be able to make sounds resembling spoken English at least in human form (though with a strong accent is fine), then what sounds would be common to both forms of such a creature? For bonus points, rather than just making a list, also explain why those and not some other ones.

To clarify, these creatures (in either form) don't need to reliably fool someone with a training in medicine or anatomy (think doctor or veterinarian), especially upon close medical examination, into thinking that they are what they appear to be. However, they should maintain the respective outward appearance as much as is reasonably possible (small adjustments are allowed, but not major deviations, as illustrated e.g. by the human-with-a-muzzle example above).

Also to clarify, you are allowed to adjust the internal anatomy! So there is no need to be constrained by a wolf's (or a human's) voicebox, for example. What I want to avoid changing much is the exterior appearance. It is however best if the exterior appearance is as close to the only thing as possible that changes when they shapeshift.

  • 2
    $\begingroup$ Would hiding a human mouth in the back of a wolf's mouth be allowed? $\endgroup$
    – user25818
    Commented May 1, 2018 at 22:15
  • 1
    $\begingroup$ @notstoreboughtdirt I think I'd rather not go that route, to be honest. It'd definitely add a lot of complexity. But remember that you're allowed to tweak the internal anatomy, as long as the exterior appearance meets what I'm after. $\endgroup$
    – user
    Commented May 2, 2018 at 5:31
  • 1
    $\begingroup$ Look up Kevin Grevioux on Youtube. He played the werewolf Raze in Underworld. When I saw the movie, I assumed they had digitally altered his voice for the role. Then I heard him speak. Holy crap! That's what I would expect a lycanthrope passing as a human to sound like. $\endgroup$ Commented Jun 2, 2018 at 5:13

3 Answers 3


Let's start from an easier position. Let's start with what a wolf can't do.

  • Wolves form very few front-of-the-mouth sounds. A whistle, for example, is beyond them. As would be a hiss or an "oooh" sound (indeed, O and U vowels might be beyond them). Wolves don't have significant lip control (at least I've never seen a dog smirk, though that's likely not conclusive evidence....) so they have a limited control of pitch.

  • Wolves also can't form tongue-sounds such as clicks and the "th" sound. They can't roll their R's or make a buzzing sound.

Or can they?

However, your wolf isn't just a wolf, it's an intelligent brain inside a wolfey-looking body. It remembers speech, does it not? Two interesting tidbits from here:

  • The shape of an individual's vocal tract is partly genetic, partly learned.

Which suggests that your intelligent wolf can increase the sounds it could reproduce. A good example are those tongue sounds. Yes, that lengthy thing might be a challenge to click off the soft pallet or push up against teeth, but it could be learned nonetheless.

  • More than 100 muscles work together when you sing or form a single phrase.

Here's where some of your world rules can work in your favor. I frankly don't know enough about canine physiology to know if that block of 100 muscles exists in the species — but I'm willing to bet it doesn't. In other words, all the jaw, neck, shoulder, and chest muscles that we use to produce a very complex set of sounds simply may not exist in wolves.

At least not normal wolves. Perhaps your shapeshifters retain this particular trait? That would mean a slightly increased chest area (might impact the use of front legs) and a thicker throat/neck area. It might also mean a thicker snout to permit greater lip and fore-palate control. In other words, if you know what you're looking for, you can see the difference between a wolf and a lycanthrope.

And that's what makes creature design interesting, isn't it?

  • 1
    $\begingroup$ "It remembers speech, does it not?" It definitely should. They do maintain memories and mental faculties (though limitations in their body may mean they're less able to make use of them in one form or the other), so I don't think it's much of a stretch that this would extend through speech. "if you know what you're looking for, you can see the difference between a wolf and a lycanthrope" That's acceptable; they don't need to fool a doctor (or a veterinarian), as long as they can pass for both forms to a more casual observer. $\endgroup$
    – user
    Commented May 2, 2018 at 5:38
  • $\begingroup$ So you're basically saying there's little reason I can't have what I want, with human-like and wolf-like creatures alike being able to make human-esque sounds, with only minor tweaking of the outer appearance of either form? $\endgroup$
    – user
    Commented May 2, 2018 at 5:39
  • $\begingroup$ That's what I'm thinking. You'll get a shift in pitch or timber due to the long snout and palate, but other than that I think that the combination of brain + extra muscle set makes the creature completely believable. $\endgroup$
    – JBH
    Commented May 2, 2018 at 13:34
  • $\begingroup$ @JBH: Spot on, but don't forget the other direction, Wolves are already capable of pitch control to some extent, but are capable of more guttural sounds then most humans, so therefore the human form would most likely be able to perform these easier as well. $\endgroup$ Commented May 29, 2018 at 11:17
  • 1
    $\begingroup$ Also worth considering the amount of communication that is non verbal, body language pays a huge part in wildlife and still a huge part in humans as well, might be interesting to see the human forms acting somewhat wolf like in body language while talking because they've forgot what form they are in, i know this wasn't part of the question, just worth noting $\endgroup$ Commented May 29, 2018 at 11:20

Well, seeing as the more limited of the two forms (vocally, I mean) is the wolf, I find it logical that most such shared sounds would originate from the wolf side. Wolves have several distinctive sounds that can be transferred through to the human:

Whining. Whining is quite a common canine sound which often signifies sadness, discomfort and several other emotional states. I would assume it would carry well to the human side and would enrich the social communitation between speciments of that species. Since wolves have a pack mentality, whining would often signify a lower social standing in relation to another specimen.

Growling. This would probably be less impressive and more subdued in human form, but it communicates threat quite clearly so I think it would transfer.

Sighing. A sigh is quite universal. Granted, its not exactly a full sound, but it does transfer well.

Laughing. While it would be a bit distorted in wolf form, its still transferable.

'Huh?' An inquiring sound would probably transfer well as well and its useful.

Groans. When aiming to voice displeasure, a groan is a basic sound and it would transfer well.

As a side note, I would advise you to think of other things that would be shared, like body language. wolves use body language, the position of the body, ears and tail for social ranking. This is bound to transfer to the human side in some way, you just need to determine how.

  • $\begingroup$ Don't forget about yawning.... $\endgroup$
    – Shalvenay
    Commented May 2, 2018 at 3:00
  • $\begingroup$ I feel that you're missing out on the part where the internal anatomy is allowed to be adjusted, and therefore miss out on a lot of possibilities. Could you Edit to elaborate on why that wouldn't allow human-like sounds, if you feel that remains the case? $\endgroup$
    – user
    Commented May 2, 2018 at 5:32

Let's start by looking at the possible human sounds in the IPA's pulmonic consonants, it's points and manner of articulation, and figure out what in this human-based inventory a wolf may be able to produce consistently. Pulmonic consonant chart from ipachart.com

Right off the bat we should note that animals like wolves have relatively little control over their front lips (only really moving the front of their mouth for a submissive grin) making the articulation of the lips for the bi-labial and labio-dental columns infeasible in my opinion.

a dog wearing a submissive grin

Dental sounds, remembered from a human form, may be possible since they do not require difficult lip movements. I can't see any reason why alveolar/post alveolar sounds would be infeasible for a long-muzzled creature, but the tip of a canid's tongue has less strength and articulation than a human's, so that will potentially be an issue that could prevent these phonemes from being reliably produced. Same with Palatal, Velar, etc. Actually those consonants with further back point of articulation would probably be the most natural for a dog or wolf to make; think growls, murrs and other far-back fricatives that dogs are fond of.

Retroflex may be tough, as it'd require the creature to stretch their mouth open wider to accommodate the curve of such a long tongue.

I'm very dubious about how a creature with a long tongue would produce trills, especially since dogs have thinner tongues with less articulation at the tip.

Plosives all seem out of the comfort zone of most dogs, so keep that in mind.

Approximants are your friends because they do not require total closure/contact in order to be produced, and therefore will be easier to make in general. Lateral Approximants and Fricatives are also good, as a lateral movement of a dog/wolf's lips is completely natural to them.

It'd be good, when considering canine conlangs, to watch videos of 'talking dogs' and listen to what phonemes the dog produces most naturally. Dogs tend to have a preference, to my ear, for: velar, uvular, pharyngeal and glottal fricatives; what sounds like a lot of uvular trills; aspiration; length, tone and stress distinction on vowels.

Dogs also don't have a uvula so they wouldn't be able to produce uvular sounds.

Look into those light green and gray squares in the image; those are points + manners of articulation deemed rare or impossible for a human. The whine a dog makes isn't anywhere on here of course, but it'd be a natural sound for the wolf form, and replicable for the human form. (What any whine would be called on this chart is a question that I'm not sure I'm equipped to tackle.)

I made a rough chart of consonants to seek or avoid: those best left alone are blacked out, and the most important ones boxed in red. The most of a dog/wolf's articulation is in the back of the mouth, so that's where the phonemes are.

Retroflex approximants are often made when dogs yawn, so I left them in this modified chart. That said they are unlikely to be any more phonemic than the clicks an English speaker makes when tsking at someone. The problem with labial consonants is not having the articulation to make a perfect mouth seal, so a labial approximant is perfectly fine.

![modified ipa chart

And what about vowels? Vowels are easy as pie for a dog to make, and it's a good place to put tone, length, nasalization, aspiration, rhotacizing, vocal fry and stress for variation, so you don't need to have many vowels at all to be able to thicken your phonemic inventory.

IPA vowel chart

I wouldn't discount wolves or dogs as incapable of rounded vowels, as they can round their lips just fine to howl. I've heard many a dog pronounce a phoneme equivalent to /ɔ/, though in vowels especially I think you can hear the difference in a dog/wolf's skull and mouth shape.

That said the IPA vowel chart still applies with it's handy mapping of place/manner of articulation. A dog can pronounce a rounded back vowel with an open mouth, and even if it has different auditory qualities than a human /ɒ/ it's still a /ɒ/ to me.

However I'm sure a shifter would be able to mimic human vowels just fine with practice, but it wouldn't be natural for every shifter to force their mouths to work like that. These vowels are analogous enough that it seems likely that shifters would just have a 'human dialect' and a 'wolf dialect' that are mutually intelligible.

Front vowels could be tricky though, especially the front closed set /i/ and /y/ which Ive never heard a dog make. It's hard for me to make conclusive statements on this however, as most vowel sounds I hear out a dogs mouth have significant vocal fry or heightened pitch, making them very different to my ears.

Do what feels right, listen to the sounds dogs make (they're much closer to wolves than most people think)


You must log in to answer this question.

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