A lot of languages, both natural and constructed, mark things like part of speech, gender/noun class, tense, and what not with suffixes. But what about prefixes?

I think that would make more sense for a head-first language. But the issue with just adding vowels to the beginning of all words makes things look repetitive, and also adds an additional syllable.

An alternative would be to have the initial consonant instead be a prefix. Though this would mean that every root in the language would have to start with a vowel (unless the language was very liberal with initial consonant clusters).

I don't know of any language that does this. Its common for languages to have the rhyme of a word mean something, but I've never heard of the onset of the word meaning anything.

And yes, I'm aware of the 'consonant mutations' of the Celtic languages, but those are triggered by agreement. They can't indicate anything on their own, as far as I know.

I'm not saying that all conlangs have to be like natlangs, but in my experience, there's some things that languages do where there's a very good practical reason why they do them. If you come up with something that no natlang does, there may be a good reason for that. Humans for example seem to favor placing the subject of the sentence at the beginning. If you say try to place the verb first, you end up with a lot of weird problems. Of course, there are verb-initial languages, but they're rare and tend to feature very exotic grammars (Tagalog comes to mind).

  • 1
    Affixes can be more than one phoneme... think of the English prefix dis- which can attach to words which start with both vowels and consonants.
    – curiousdannii
    Jul 15, 2018 at 6:48
  • What's your actual question here? Whether there are any conlangs with lots of prefixes?
    – curiousdannii
    Jul 15, 2018 at 6:49

3 Answers 3


Many languages mark things like those you mention at the beginning of a word. Noun class in Swahili and the other Bantu languages comes to mind. It doesn't require all roots to begin with a vowel; the prefixes also include vowels (as in Kiswahili).

Russian has prepositions (not prefixes) such as в, which are pronounced as part of the onset of the following word (and it does sometimes cause initial consonant clusters), in addition to case. There's no reason why a conlang couldn't have the same as a prefix for gender, case or tense.

You might also want to look at the WALS chapters Prefixing vs. Suffixing in Inflectional Morphology, Position of Case Affixes, and Position of Tense-Aspect Affixes for lists of languages that have these features as prefixes.

  • In Bantu, prefixes not only mark noun classes (and thus adjective agreement), they also do most of the inflection of verbs. Oct 20, 2018 at 21:01

The closest in design is John Wilkin's Philosophical language. The words are formed according to a scheme not unlike Dewey Decimal Classification: Every letter of a word adds some more specific information to it. In this scheme, the first letter already carries a lot of information (giving the basic factorisation into classes) and the additional letters add specifics.


In the conlang "Ro", each letter represents a specific thing. The first letter is a clue to the meaning of the rest of the word.

For example,

gebrac means an inch, and the first syllable "ge" means measurement.

radac means boy, and the first syllable "ra" means person.

The initial consonant can therefore change the meaning since each letter has a meaning.

rebec means think, ribec means remember, but zibec means repeat.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.