I am seeking to construct an a priori language that is optimized to provide the greatest creativity for songwriting and poetry, in addition to being extremely general and productive in neologism, derivational morphology, and systems of nomenclature.

The first aspect of the language will be that it is inflected and fusional because it is easier for the mind to process (Source: Giuseppe Peano tried to remove all declension and conjugation when he attempted to construct a language (latino sine flexione), and Lancelot Hogben noted that the lack of grammatical forms made it harder to learn, using the words "[it had] too much grammar of the wrong sort, or not enough of the right").

However, this cannot be done by giving every word a fixed grammatical ending, because that will make it too regular and this will remove the "good kind of" disorder necessary for creativity. What does this mean? For instance, if I made it so that every noun has a regular suffix for declension attached to a root, then every noun will be at least two syllables long, which would make the spondee impossible ("Good friend, thou hast no cause to say so yet" - King John)

Surely, some smarter men have created metrics to analyze these aspects of language? If so, where are they?

  • 1
    Which aspects do you mean? Learnability is much easier to measure than creativity.
    – Draconis
    May 22 at 5:42
  • 3
    Depending on that, this might be a better fit for Conlang.SE, which allows more open-ended questions about things like "creativity" that aren't as objectively measurable.
    – Draconis
    May 22 at 5:48
  • 1
    @Draconis I'm sorry. My previous comment was rude. Here is what I meant to write: learnability is not a useful metric, because it doesn't matter how hard a system is to learn, provided that it is indeed a learnable system. A learnable system is one that takes a finite amount of time to learn. English spelling, Hebrew's indo-european vowelless abjad, and Japanese kanji are all unlearnable systems, because you never stop learning them, and not in a good way. What criteria are there to distinguish between learnable and unlearnable systems? May 22 at 7:07
  • 1
    @SirCornflakes Yes, that is a good observation, and the more "tricks" like that we can discover the more versatile and expressive the language will become. "We're trying to achieve world peace and understand the very structure of the universe using a language which was designed to tell one another where the best fruit was." - Terry Pratchett May 22 at 8:11
  • 1
    "Everything that we have so far seen to be true of language points to the fact that it is the most significant and colossal work that the human spirit has evolved -- nothing short of a finished form of expression for all communicable experience. This form may be endlessly varied by the individual without thereby losing its distinctive contours; and it is constantly reshaping itself as is all art. Language is the most massive and inclusive art we know, a mountainous and anonymous work of unconscious generations." -- Edward Sapir, Language, 1920.
    – jlawler
    May 22 at 15:07

1 Answer 1


What tools are there to measure aspects of language?

  1. What aspects of language are there?
  2. Which ones are measurable?
  3. How?

I am seeking to construct an a priori language

I understand “a priori” to mean something like “from the beginning” or “before”. What do you consider an “a priori” language? A language to only contain formulations relevant to a priori thinking, as in, a language which can only express pure reason with out any reference to empirically learned concepts like the external world?

that is optimized to provide the greatest creativity for songwriting and poetry

Nice. How so? What features of language do you think would support that?

in addition to being extremely general

As in, wide-ranging, with comprehensive span over a huge variety of possible concepts, and words for each of them?

and productive in neologism, derivational morphology, and systems of nomenclature.

Presumably because the nomenclature systems are themselves hierarchical in the way they classify things in the world, so the lexical items correspond by themselves being derived from words representing their containing or origin categories.

The first aspect of the language will be that it is inflected and fusional because it is easier for the mind to process

My understand of “inflected” means the words can change their form. “Fusional” means that the word isn’t discernibly decomposable into sub-word units reflecting its derivation process. So, you want derivation/transformation rules by which word spawn new words in a rule-based way, but you do not want it to be based on “agglutination”, where words are like bundles of semantic ingredients; you want it to be fusional, so that a defined word has a distinct form from any representation of its derivational process.

Although that is awesome, I am curious why you take it for granted that this is “easier for humans to understand”?

(Source: Giuseppe Peano tried to remove all declension and conjugation when he attempted to construct a language (latino sine flexione), and Lancelot Hogben noted that the lack of grammatical forms made it harder to learn, using the words "[it had] too much grammar of the wrong sort, or not enough of the right").

I see! Well, one common frame on what you said in the discourses that I have encountered in my own life, would be that there is a possible bias being exhibited, here, in which a person espouses an attitude that some (partially) personal, subjective assessment of something’s desirability, suitability, advantage, commendability, is a basis for a broader assertion that that thing actually lacks that quality inherently. All we know from that passage is apparently “Lancelot Hogben” did not like Peano’s language. I am not sure what it would take to generalize the momentary impression of one Lancelot Hogben to the palatability of uninflected (“analytic”) languages in general, but it might require us to hold his estimations in very, very high regard. I for one have never heard of him. Chinese is a common example of a language without or without much inflection, and I happen to find it really cool and interesting.

The other thing is, you seem to be saying that you want a language that has a lot of explicit grammatical markings as opposed to very few, but that is not the same as a language being “inflected” or “fusional”. I don’t know as much on the following topic as I would like, but I am pretty sure a common idea in the modern world is that the features of specific human languages vary, but we generally have very, very similar cognitive capacities, in a general way, and so differences in the specific features of this or that language might not have as heavy of an effect on thinking as you may imagine. I think that since natural languages evolve spontaneously and organically of their own accord, they tend to have a self-regulating constraint which causes them to tend towards similar overall profiles in terms of expressive capability, efficiency, and so on, even if with variation.

I think what you are actually interested in is not “inflectional” or “fusional” languages vs. whatever not-that is (analytic, agglutinative?). You are interested in grammatical explicitness. You want the components of a message/idea/act of communication to be unambiguously structured according to some rules of construction and also reception.

At that point, the idea may be slightly circular. Your language will choose to explicitly mark certain “grammatical” features. But which ones? We don’t really know why some ideas get codified into grammatical markers whereas in other languages they do not. For example, gender. Originally, we can imagine gender is a purely semantic notion. Some languages somehow made it part of the syntax of their language, not only the semantics. The words actually change their shape depending on if what they refer to is a boy or a girl.

My point is, I do not know if there is any intrinsically self-justifying list of informational topics that are “grammatical” as opposed to just “informational”, as in, semantic. In theory, any concept in semantics can become grammaticalized. You just have to turn it into a “function”, in a sense - make it “productive” - where instead of there being one single token in the language referring to that thing, take for example, “music” - you turn it into a function on other words, which can apply some rule to other words to add that idea to them: “pertaining to music, or not”. So, I can hereby create a new variant of English where it is grammatically required to inflect every single word as to whether or not it pertains to music in some way, with the letter [m]:

I[x] do[x] not[x] like[x] this[x] stupid[x] language[x] about[x] music[m] I[x] made[x].

I would say it has taken be a long time, and I am still working on it, in consolidating the intuition on a deeper level of the difference between syntax and semantics. Semantics is the world of meaning we are familiar with. Syntax is like a jigsaw puzzle where you make rules about the actual words themselves, how the letters get put together on the page.

You basically are saying you believe “language” requires rules of composition by which words are put together, in order to reflect their relationship to one another. But that you are worried this will actually suppress “creativity” - if the grammar is over-specified, there could even be a complete constriction on allowable formulations - making a language (interestingly! Like Newspeak in Orwell’s 1984) in which creativity or individual choice in self-expression literally impossible!

Ok: then if your idea for “creativity” means there being more choice in how you can say things - a kind of “multiplicity”, or as I think one person recently put it, a “hyper-productive” language - then I think I know a language just for you, because years ago I read about how Sanskrit has so many inflectional rules that it was said by a teacher that students learn how many different ways there are to say the same thing (I think).

So: I think you are looking for a language that is highly explicit and grammaticalized - like Ithkuil, it comes pre-loaded with tons of syntactic machinery for conveying loads of rich, semantic parameters, contexts, attributes, on top of any base message, of what you are trying to say - it is not a “free” language like free jazz or something, with minimally specifying or restrictive rules, no right or wrong in terms of word order, word choice, etc - but, somehow, you almost want the language to have a high level of redundancy in the sense of multiple choices. Maybe someone can think of an example in English where we have two semantically equivalent forms, I don’t know, “will” vs. “going to”, “Will you come?” = “Are you going to come?”. It also reminds me of programming languages which are strongly typed but implicitly. That’s a different (though fascinating) topic to draw inspiration from. There are some programming languages where different kinds of elements are not explicitly recognized as different types by the compiler (= on the syntactic level), and others which are strongly typed (like Haskell, a language I am currently very drawn to). But there are also languages like I think JavaScript and Swift where the elements have types but you can choose not to have to say them - they are optional. If you don’t say them, the computer “guesses” (sort of) what the type is, anyway.

So that means your language can be both heavily structured yet flexible by supplying tons of syntactic constructions but making them optional to be used! (Unlike Ithkuil, which requires all grammatical parameters to be present, in the sentence.) I wonder what natural human language might have some good examples of this? I can’t think of any just now.


In the introduction to his celebrated translation of Vidyakara's Subhāṣitaratnakośa, Daniel H.H. Ingalls describes some peculiar characteristics of the Sanskrit language.

He refers to the enormous vocabulary of Sanskrit, and also of the presence of a larger choice of synonyms in Sanskrit than any other language he knew of. Further, just as there exist a vast number of synonyms for almost any word in Sanskrit, there also exist synonymous constructions. In his elementary Sanskrit examinations he would ask his students to write in Sanskrit the sentence 'You must fetch the horse' in ten different ways. Actually, it is possible to write the sentence in Sanskrit in around fifteen different ways 'by using active or passive constructions, imperative or optative, an auxiliary verb, or any of the three gerundive forms, each of which, by the way, gives a different metrical pattern'.



Optional agreement in Santiago Tz’utujil (Mayan): https://www.degruyter.com/document/doi/10.1515/zfs-2020-2018/html?lang=en

Tangut verb agreement: Optional or not?: https://hal.science/hal-03893853/document

Optional agreement as successful/failed AGREE. Evidence from Santiago Tz’utujil (Mayan): https://www.jbe-platform.com/content/journals/10.1075/lv.20013.lys


The New Ithkuil Language: https://tnil.readthedocs.io/en/latest/

The forms and meanings of grammatical markers support efficient communication: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8670496/


  • 1
    That's a very well-done decomposition of my motivations, and the last paragraph is a correct summary. May 23 at 1:52
  • "a priori" <- in conlanging, a priori means "the wordforms came from my own imagination and were not borrowed from living languages" That's it. It does not mean what Immanuel Kant meant by a priori. FYI. May 23 at 2:35
  • "Nice. How so? What features of language do you think would support that?" <- The two features that would support these are 1. full usage of all phonotactical word forms (I don't know how to simplify this) 2. Grammatical affixes that provide diversity of short and long syllables for metrical variety. May 23 at 2:38

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.