Posted by & filed under Podcast.

Today we talk to you all about vowel harmony, taking Turkish, Finnish, Moro, and Mongolian as case studies to help you figure out the intricacies of what choices you need to make in vowel harmony systems and how you can introduce interesting

Top of Show Greeting: Maksinaunminverbe

Links and Resources:

Special mention (from the outtakes): A giant freaking basket


χαιρετε, ὠ φιλε (“Greetings, friends”),

Over at the Conlangers group on deviantART, we’re hosting this little thing called NaCoWriMo. Basically, participants have until [Nov. 31] to write a 200-word passage in one of their conlangs. It’s not a translation challenge, but rather a bona fide composition challenge. I thought some of the Conlangery listeners would be interested in participating. It’s never too late to sign up, and anybody who wants to can do so at


13 Responses to “Conlangery #74: Vowel Harmony”

  1. Pete Bleackley

    After David said that he wasn’t sure whethe r its noun affixes were cases or not, I now want to know more about Moro. Unfortunately, Wikipedia says practically nothing about it. But it does say it in a choice of 6 languages, including Breton.

  2. AlucardNoir

    Off topic question: If a pidgin is a simple form of communication and a creole is the natural development born of a pidgin to facilitate communication, and if most languages have a tendency to get simpler and simpler over time, how do complex proto languages arise in the first place? I mean take PIE for example, it has a grammar more complex then anything in Europe at this time, hell even ancient Greek and classical Latin were simpler then PIE(grammatically speaking) and the’re several times more complex then some languages like English for example.

    On topic: is vowel/consonant harmony an acceptable way of circumventing highly regular conlangs and making them look(sound/feel) more natural?

    • admin

      On the first: I’m not sure how strong the tendency toward simplification is. Languages certainly can gain complexity in certain ways — by grammaticalizing new things, for instance. Another issue to consider is how you measure linguistic complexity. Objectively determining how “complex” a language is is problematic enough that many linguists simply avoid the problem altogether and assume that (after taking all factors into account) all languages are equally complex (excluding pidgins, which are not considered full languages).

      To the second: Vowel harmony is a great way of achieving what I call “regular irregularity”, where a perfectly regular process can create something that seems irregular to someone unfamiliar with the system or who is linguistically unsophisticated — especially if your vowel harmony system includes multiple features harmonizing in different ways (ie Turkish having all vowels harmonize on [back] but only high vowels harmonizing on [round]). It’s still not a substitute for genuine irregularities, though — since once someone figures out a highly regular vowel harmony system, it’s going to become very transparent.

      • AlucardNoir

        1. Tx for answering, I knew that beforehand, I just don’t agree with this relativism that seams to permeate modern western culture, the theory of gravity and the theory of relativity are not equal, why should it be any different in any other field, religion, philosophy… linguistics.

        2. Regular irregulars, would you mind expanding on that? it sounds very interesting. (if that sounded disingenuous, it was; I just want to pick your brain for ideas, since I personally dislike irregulars and see this as a great opportunity to advance my engelang)

        • David J. Peterson

          I personally dislike irregulars and see this as a great opportunity to advance my engelang

          Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!

          So, let’s go back to square one. Different conlangs are different. A conlang is defined by its goals. Naturalistic conlangs attempt to appear like natural languages. No natural language is perfectly regular. Ergo…

        • admin

          First, I will say that I will admit that Conlangery is very much focused on naturalistic conlangs. The hosts have always been naturalistic artlangers, and thus, that is our perspective. Perhaps there is room out there for an engelanger podcast?

          On relativism: the problem with linguistics (and anthropology in general) is that the subjects of study are human beings. As such, it can be very difficult to dispassionately study them, and even when you think you have blocked out prejudices and biases, your results might still offend people, and could be taken by others as validation of their own prejudices. Some arguments for equal complexity of languages invoke this argument, though the obvious counterargument is that we don’t have to consider complexity good or bad. In any case, it’s still difficult to measure.

          The whole story on my “regular irregularity” probably can’t be adequately summed up in a comment. I’ll see if I can work up a blog post on the subject in the future.

          • AlucardNoir

            A. Tx for taking my request so seriously, should you do it I will be looking forward to it.

            B. Why would I be interested in an engelang podcast. The main problem with engelangs is that they are subject to the same reinvention of language by children as any language is, not to mention the same naturalistic tendencies to change as normal languages, I’m listening to your podcast both so I can learn new things and so that I can built a language that can resist or attempt to resist such changes, what’s the point in building the “perfect language” if it will stop being perfect in a 100 years. (and I’m looking here both at the auxlang Esperanto and at Classical Arabic)

  3. Panglott

    I had been thinking that Chukchi had both vowel harmony and a five-vowel system. But I looked it up and I was mistaken: Chukchi has the five cardinal vowel that participate in vowel harmony, along with a neutral schwa that does not participate in vowel harmony, six vowels all told.

  4. Qwynegold

    David’s explanation of Finnish vowel harmony was confusing. I would explain it like this:

    There are three groups of vowels: The front vowels Y, Ö, Ä, the back vowels U, O A, and the neutral vowels I, E. Front and back vowels can’t both occur in the same word, while neutral vowels can hang out with any vowel.

    Suffixes that contain a non-neutral vowel come in two versions, one with a back vowel and one with a front vowel. For example the partitive case suffix is either -(t)a or -(t)ä, and you have to choose the right version when adding the suffix to a word. The front-back pairs of vowels are: Y-U, Ö-O, Ä-A.

    If a word consists of only neutral vowels, then it will take front harmony.

    Now there are some words that break vowel harmony, mostly loan words. One example is parfyymi (perfume) which has both a back A and a front Y. The rules seem to be unclear what to do in this case. The general rule is that the first non-neutral vowel decides which harmony the rest of the vowels have, so some people would make the partitive form of that word parfyymia (which would break vowel harmony twice: A > Y > A). But I think one should instead consider the last non-neutral vowel, so I say parfyymiä (which breaks vowel harmony only once).


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.