💾 Archived View for yujiri.xyz › spem captured on 2023-01-29 at 02:35:18. Gemini links have been rewritten to link to archived content

-=-=-=-=-=-=-

Spem, the philosopher's conlang

This is a conlang I'm making with the quite serious intention of getting people to speak it. Yes, it's a crazy uphill battle, but the amount we have to gain from a better language is staggering.

Update: the Spem project had a deep flaw which is that the names of everything were based on nothing more than my own subjective associations with sounds. I decided to reapproach the association profiles of each sound, as well as the set of included sounds, with more reasoned design. As this involves redoing most of the vocabulary, I decided to give the language a new name and separate documentation, while leaving Spem up here in its current state until the new iteration, Thaya, is ready to replace it.

The phenome and alphabet

Dictionary search

The goal is that you should be able to learn the language mostly just by reading the dictionary, but a few concepts get dedicated articles:

Demonstrative- and quantifier-type compounds

Default identifiers

Degree modifiers

Articles that aren't strictly about learning the language:

Design insight: the tradeoffs of parts of speech

The phonetic mapping: each sound's ideographic profile as I perceive them

Open question: variable dereferencing?

Philosophical accuracy

The language you speak has an enormous impact on the way you think, and likewise the language a society speaks has an enormous impact on its culture. I'd go as far as to say *most* prevalent harmful ideas can be traced to our languages suggesting them (note that most of these points apply to most or all natural languages):

**Polymorphic "good"**. "Good" refers to many completely different concepts, such as morality, pleasure, and skill. (The related words "should" and "must" can also refer to probability.) The conflation of morality and pleasure doesn't just suggest utilitarianism, it lets people write things like the tvtropes turn-the-other-cheek propaganda piece "PayEvilUntoEvil" and people can get behind the garbage. Making such wrong and destructive ideas sound true would be much harder if morality and pleasure weren't named the same thing.

https://tvtropes.org/pmwiki/pmwiki.php/Main/PayEvilUntoEvil

**Synonyms facilitate circular reasoning.** When there are many ways to communicate the same thing, it's easy to see a difference where there isn't one, and people can make statements of the form "X is right because X is right" very persuasively by calling X two different things. Even honest people who were raised with bad ideas can do this unintentionally.

**Commands.** Commands as a language construct allow you to pressure someone to do something without distinguishing between: you think they morally ought to, you think it's in their interests, or you just want them to. Commands have a lot of harmful psychological power because of this.

**"Redeem" being transitive.** To redeem yourself means to do noble things that balance your past sins. Someone else's actions can't make you a better person, so this verb is meaningless when used non-reflexively. Yet a religion followed by *billions* revolves around the idea that Jesus's death redeemed *us* by paying back Adam's debt... all because the verb "redeem" being transitive suggests this incoherence.

**No distinction between causation and deduction.** The word *because* communicates both the causal relationship of events and logical deductions. "Because X happened, Y happened" and "Because X is true, Y is true". And I have definitely seen people abuse this confusion to create bizarre sophistry like a time I argued with a metaphysical materialist and he argued that all of my arguments were appeal to consequence fallacy because I was refuting his ideas by showing that they entailed absurd and untenable consequences.

**Emotion vocabulary is useless.** We h ave several dozen words that designate emotions, and most of them are poorly defined or conflate multiples, which breeds a culture of not understanding how people work. We need this vocabulary section to be devised by someone who really understands human emotions. And that's not me yet, but it will be eventually, and the same can't be said of most people.

Emotion philosophy

Esperanto doesn't address most of these problems.

Ease of learning

It would be a massive benefit if children learned to communicate faster. One reason is that they could learn other things and mature faster, but a more important one is that a person uncapable of sophisticated communcation is much less likely to be treated as a person. Hence children and animals being the two most brutally oppressed groups.

General design philosophy

**As much as possible, have a one-to-one mapping of concepts to lexical constructs.** If two sentences use the same concept, they should use the same grammar; for example "You should do X" and "It's okay to do X" are both making statements about morality so the sentence shouldn't need to be restructured. It should only need to switch out "should".

Another example is causation. English says "I made X do Y", but Spem says "I caused that X do Y". The Spem grammar is more intuitive and more flexible - sometimes in English when the make-verb construction doesn't seem to cut it we end up saying something like "I made it so ..." which sounds a little awkward. The Spem grammar works anywhere.

**Speed is important, and so we should use our phenome thoroughly.** Probably most one-syllable words should be words, and most two-syllables. The most common words should be the shortest. Three-syllable words should be a rare exception. Part of the reason speed is so important, besides the obvious, is that if the language is slow then people will be motivated to take shortcuts that fudge their meaning, which harms philosophical accuracy. I certainly do that every day even in English (which is by far the fastest of the languages I know).

**No homophones.** Of course there'll be no homophones, but I also want to minimize "multi-word homophones", where all the *individual* words are unique but two words in a row can sound the same as an unfortunate third word.

I'm ambivalent about whispering. Whispering mostly or entirely removes the distinction between voiced and voiceless consonants, so that would mean that to truly avoid homophones we'd have to also not have any words only distinguished by voicing? I'm not certain whether whispering makes it completely undiscernable, but it seems like something to avoid, except that it's such a hamper.

**If the thing is similar, so is the representation.** Related words have most of their sounds in common. This is difficult to uphold constantly, and it's obviously unfeasible to uphold the reverse (if the representation is similar then so is the thing). But all the interrogative pronouns, for instance, share a starting letter or two. Words that are opposites usually have the same structure and just switch one or more of the sounds for "opposite" sounds (for example *ee* versus *oo*).

**Phonetic associations are not arbitrary.** Our associations with certain sounds - for example, how *l* is a "soft" sound and *ah* is a "wide" sound - are not completely arbitrary; they're based on a combination of what we physically do to the produce the sound and natural sounds that are similar. I believe these associations should be upheld in the language; giving sounds consistent profiles of assocations makes the vocabulary intuitive.

**Beauty.** Spem should be beautiful. This will make it not only enjoyable to use, but more enticing to learn. (Phonetic appeal was one of the things to drew me to Japanese.) We should avoid having words that are likely to be used in sequence sound ugly together.

Obviously these goals clash often and it's not always clear how to prioritizing them, but I tried to list them in order of descending importance.

Broad concrete choices

**No plural.** The difference between one and two is not special and does not deserve to be treated differently from the difference between two and three. Number is unspecified by default because it's often not essential to the intended meaning of your statement; how many times have you had to write '(s)' after a word so it applies to both singular and plural? We have words like many and several, and might add a one-syllable word for "two or more", and the use of this or "one" before stuff won't sound unnatural because we'll be used to it being standard whenever the number actually matters.

**No distinction between nouns and adjectives.** Substantive adjectives or placeholder words are used quite often in English: "the poor", "the green one", et cetera. In Spem any adjective can be used this way by default and it won't need a noun like "one" in that second example. Note that while "the poor" works well, if you don't want to use "the" you need a noun like so: "poor people". You shouldn't have to say "people".

Basically there are "entity words" or descriptors in Spem that just pile together to describe an entity. The entity has whatever traits the descriptors specify, and all descriptors are created equal.

**SVO word order (subject-verb-object), like in English.** SVO is the best word order because the verb is a different part of speech so it makes sense to use that as a natural divider between subject and object. Especially given the above policy, having the subject and object next to each other could cause serious problems.

Japanese is a language that generally follows **SOV**, and that's part of why it needs particles like は (wa), が (ga), and を (o) to denote the subject and object of sentences (I know Japanese doesn't have subjects don't pile on me Japanese teachers I'm simplifying it for the 外人). These particles are all one syllable but needing 1 or 2 of them in most complete sentences adds a *lot* of syllables at the end of the day.

**No affixes.** Pretty much all words in Spem are standalone; things like tense and the -tion conversion are accomplished with separate words. The distinction is somewhat meaningless, but I think it helps clarity to write them with a space so that they can't be mistaken for a different word. Secondly, in languages that use suffixes for type conversions and stuff, they usually end up needing exceptions for words where the regular pattern sounds too ugly to be used or is hard to pronounce. I think making them technically detached words helps avoid this (it affects how we accent the word intuitively).

Adjectives before or after nouns?

Although I don't plan to grammatically distinguish the two I think we should still have a custom for it. If there's a custom then whether the custom is followed can be used to convey additional information, such as reversing the order being used to emphasize.

I've developed the de facto standard of adjectives-first, but I'm open to having my mind changed.

Discussion about Spem mostly happens in a public Matrix room:

https://matrix.to/#/#spem-dev:tchncs.de

contact

subscribe via RSS