Using a Diary as a Conlang Creation Tool

One thing I've done while developing Kílta is keeping a diary in that language. This is a very boring diary. I often talk about the weather, but you would be amazed how much vocabulary, and especially grammar, you can pull out of that notoriously uninteresting topic of conversation. It turns out I also talk rather too much about the state of my digestion, but I will, in the interest of decency, not include any example sentences about that.

When starting it is very important to stick to very simple material. This diary is not for musings on the eternal verities — at least, not yet — but is a language proving ground. If all you have to do is say "it rained today" then you're more likely to develop a regular conlang diary habit. If you start off trying to do a Situationist analysis of funerals in the video game Animal Crossing you are immediately going to run into serious troubles. Save this for much later. At the start, and possibly for several years, what you'll be saying in the diary will be fairly simple and mundane.

If you have a conculture in development, you could also record events from that persepctive, or by mapping your local circumstances onto an appropriate concultural situation, but again staying quite simple at the start.


You do need a few things in place before you can start the diary. Basic decisions around nouns, pronouns, verbs, and argument structure should be in place and settled. If you're still dithering over how to form a simple past statement, it's not yet time. By preference, you will not be making radical revisions to earlier decisions once you start on the diary. One way to get better in your conlang is to go back and read old entries, and this will not work if the language is completely different from week to week. Minor tweaks don't matter, though, and are expected.

If your language can handle the first 25 examples of Gary Shannon's syntax tests, and these sentences aren't expected to change radically in the future, you are more than ready to start the diary phase.

Here is my first diary entry for Kílta:

Ënni nama tëtti asíko rui.
today sun a.bit shine.PFV MIR
The sun's shining a bit today!

In óhësto.
and cold-INCH-PFV
And it's gotten cold.
The first sentence is perfectly correct Kílta years later. In the second sentence I have attached the inchoative suffix -ëst- to the adjective óhin cold. That is no longer how this is handled in Kílta, though it remains understandable.

Here's the scond entry. I must have had a cold or something:

Ha ittán kwilë kárchirë.
1SG still too.much cough.IPFV
I'm still coughing too much.

In ënni nama asíko.
As before, the first sentence remains perfectly correct Kílta, though these days I would nearly always use the topic marking for the subject in this sort of sentence. In looking through the diary, it only took about a month before the regular use of the topic marker showed up.

And the second sentence is only trivially different from the first line of the first entry. This is good. Repetition helps internalize the language.

Next Steps

You can easily — and probably should — spend a few weeks mostly discussing the weather. Well, if you live in Southern California some other similarly minor topic should be chosen (I went to work/school, it was boring, etc.). In any case, repetition of simple patterns and meanings are a good start.

Do go back and read earlier entries. Resist the urge to correct them if you see minor issues. I only allow myself repairs of obvious spelling errors. That said, do not treat the diary as a sacred relic. If you start off on a sentence and it all goes wrong, cross it out and start again. There's no need to keep it pretty.

A next good step is sequencing. How do you join clauses and maintain a flowing narrative? Narrative coherence is not often a high priority for conlangs and it really should be addressed sooner than later. Any larger texts you undertake later will be much easier.

While I had created the core converbs in Kílta near the very beginning — it was was a design decision I settled on at the start — I didn't really start using them for sequencing until about three months in. Once I started that, though, it became far more common. Here's the first clear case of sequencing with the general converb I can find:

Táni ahëkarur óhu nët, ilivët, ënni nama asíkirë.
yesterday air.PL cold.PL be.PFV.CVB, rain.PFV.CVB, today sun shine.IPFV
Yesterday the wind was cold, and it rained, and today the sun is shining.

A few days later, I wrote this:

Mëtaula chimár, ha tumun si cholo.
mëtaula chim-ár, ha tumun si chol-o
storm come-IPFV.CVB, 1SG thunder ACC hear-PFV
A storm is coming and I hear thunder.

These days I'd probably use the inchoative with the last example, cholësto I'm starting to hear thunder. Over time Kílta expression has gravitated to using the inchoative more often than a literal translation of it would make for good English style. This is a subtle detail about the language of a sort that is often not captured in conlang grammars.

Once I used the converbs a few times in the diary, I started using them more consciously. Because converbs were on my mind regularly, I also made sure to work them into my thinking about how Kílta would handle all sorts of grammatical matters. My use of them in the current grammar is, I believe, more consistent because I had them more at the front of my mind for the diary.

Discoveries: The Hodiernal Future

Certain patterns of use you discover in the diary can start to slide their way into becoming not just common phrases, but requirements of grammar. For example, I invented a future early, [V-at no (re)]. That is, the infinitive plus the copula is the core. The particle re is sort of an irrealis and is used in all futures that are predictions. In effect, the only time re is omitted in positive future statements is when the speaker is announcing a firm intention. As a result, re is very common in futures.

So, I already have re as a fundamental part of the future construction. I noticed myself once or twice using re with the adverb ro later (the similarity in form to re is purely an accident). After a quick look, it was clear [ro V re] was being used as a hodiernal future, that is, a future referring to things that will happen later today.

Here's the first example, which showed up a few weeks before I started using converbs for sequencing regularly:

Aronëtin no, in ro ilivo re.
cloudy be.PFV, and later rain.PFV re
It's cloudy and it'll rain later.

Notice that just ro and re are doing all the work. The infinitive and copula are not used here. The association of re with the future was enough to generate this new construction. I was happy to read years later that it's not unusual for a natural language's futures and moods to have more kinds of constructions than aspect or tense.

Kílta's current (summer 2020) future tense situation is even more complex, in part due to this hodiernal future discovery. These sorts of things can be a more natural way to break out of completely schematic and compositional conlanging, without resorting to all the techniques of the historical method.

Keep Going

At this point, you just keep going. Keep reading older entries. Keep slowly introducing new vocabulary and grammar. Keep looking at what you're doing and consider how it might generate particular forms of expression. For example, if you look at the Kílta grammar and look in the Partitive Quantifier Construction section, you'll notice that it has started to encroach on other semantic territory, again due to me just wanting to express something slightly different with it in the diary.

There will be times when it doesn't seem like much new is happening from the diary, but I believe you'll find it more than worth the effort on those occasions when it does produce new possibilities for you.

Additional Benefits

Copyright (c) 2006-2021 William S. Annis