New auxlang: Atlas

[tommus]

OK. Great. Atlas started off being very regular and I believe every little thing that can be done to make it even more regular is a very positive step.

As for the issue of "o" and "bai", I completely understand the usage. It is very clear in your examples. What I am confused with is why you show it in the Grammar in the way that you do. You show it like this:

Quantity o, bai (approximately), bix (more, more than), mei (less, less than)

Why wouldn't you show it like this?

Quantity o (exactly), bai (approximately), bix (more, more than), mei (less, less than)

[Rodiniye]

OK. Great. Atlas started off being very regular and I believe every little thing that can be done to make it even more regular is a very positive step.

As for the issue of "o" and "bai", I completely understand the usage. It is very clear in your examples. What I am confused with is why you show it in the Grammar in the way that you do. You show it like this:

Quantity o, bai (approximately), bix (more, more than), mei (less, less than)

Why wouldn't you show it like this?

Quantity o (exactly), bai (approximately), bix (more, more than), mei (less, less than)

I took for granted that people would understand that "o" is the standard preposition for quantity, but obviously it needs a bit more explanation!

I am updating the dictionary with more animals/trees etc. and a new small text but it will not be ready until tomorrow I think.

I will probably do a video weekly with worldwide news of the week. Probably that will start next weekend.

Thanks Tommus! Anything else you see please let me know. So glad that a few problems of a young project have been caught early thanks to people like you.

I am starting to love this project more and more everyday. I think people will see how easy it is to learn and form new words with the examples introduced tomorrow.

[tommus]

Here is the latest version of the popup dictionary:

Atlas Popup Dictionary Version 4.0 (English and Spanish)

The main change is the addition of all nine of the verb tenses for all verbs. Here is an example of what the dictionary entries look like for the verb "acetes" - "accept". In an ideal world, the dictionary would have the English words to fully describe the Atlas verb in all its 9 tenses. But that is a huge amount of work and is something that would come later. For now, I present the translation plus the tense in brackets. That is what is in the text dictionary and that is what pops up in the application. Because Atlas is super regular, and not supposed to have any exceptions, ideally all words including verb tenses can be constructed even if they maybe don't make sense in actual speech and writing. So I generate all possibilities for all verbs. Like this:

acetas accept [non-complete present]
acetat accept [non-complete past]
acetaz accept [non-complete future]
acetes accept [complete present]
acetet accept [complete past]
acetez accept [complete future]
acetis accept [habitual present]
acetit accept [habitual past]
acetiz accept [habitual future]

The future tense worries me a bit. I'm not sure how the future tense works. I'll try to translate all 9 tenses for "accept".

acetas accept [non-complete present] - I am accepting
acetat accept [non-complete past] - I had been accepting
acetaz accept [non-complete future] - I will be accepting
acetes accept [complete present] - I accept
acetet accept [complete past] - I had accepted
acetez accept [complete future] I will have accepted
acetis accept [habitual present] I usually? accept
acetit accept [habitual past] - I usually? accepted
acetiz accept [habitual future] - I will usually accept

In any case, all these cases of all verbs have been generated. A total of 1134 to add to the dictionary.

[tommus]

Some additional questions for Rodiniye about the Atlas Grammar.

Can you generate nouns from verbs?

For example, you have 3 verbs for work: kaqes, aikes and kames. I don't know what the differences are. But let's use kaqes for an example. Can I make the noun "work" like this: "kaq" + "u" to make it a concrete noun "kaqu", or perhaps an abstract noun "kaq" + "e" = "kaqe"?

I think there may be a gray area between concrete and abstract nouns. Concrete would be something made out of material such as "car". Abstract would be something you can imagine like" idea". What about "work". It is something you can see but perhaps cannot touch. What about "rain". "Water" is something you can touch, but can you touch "rain" or only the water in the rain? What about a "storm"? A "hurricane". An "illness". What about an "obstacle". If it is something concrete that is stopping you, then it would be concrete. But if you said that lack of time was an obstacle to learning something, does that "obstacle" then become abstract? In Atlas would you spell those two "obstacles" differently?

On page 11 where the Grammar talks about the interfix "i", there is an exception noted:

*-i drops when the resulting word can be pronounced without the vowel.

A language that is extremely regular and doesn't have exceptions shouldn't have exceptions. Bad for people and bad for computers. Why couldn't

Ze cit sebun* – he-she ate (used to eat) apples.
*-i dropped because “cit” can easily be pronounced.

simply be "Ze ciit sebun". Perhaps that then becomes an extended "i" in pronunciation.

On page 11 about Verbs. I think it would be good to state right there that verbs must have both aspect and time. I think it becomes obvious from the examples, but I think it should be stated up front.

The more I learn about Atlas, the more interesting it gets. The only things I don't like to find are exceptions.

[Rodiniye]

@ Tommus

First of all, difference between concretes (-u) and abstracts (-e) is that the first ones can be touched (things, objects) and the second ones can't.

Sometimes it is very easy to understand as you said: idu (hand) is clear, as ideie (idea).

Sometimes choosing one or another can completely change the meaning of the word:

kames means work, referring to make an effort or do an activity. kamu would be the result of that work, the object, whereas kame would be the process, the "work" or "effort" put into it.

I take advantage of the situation to specify that "kam" means what is described above, "kaq" means work in terms of an activity paid for, what you do for living, whereas "aik" means work in terms of function, something that is functioning, working (aikes).

Some areas might be a bit grey. Yes, is rain something you can touch? well somebody could say it is the water you touch, and rain is the process. Can you see the street? What is a street? Well streets don't exist, it is just a group of buildings, roads, etc together. So some people will say that streets cannot be touched (you touch the buildings), but they exist. In theses cases, it is up to the speaker to choose their way to express how they see it, and I think it can easily be understood in both ways.

Illnesses as you mentioned are straight forward to me. Illnesses are no more than processes and you cannot really see them (you can see their effects however), so in this case I would go for -e for sure (bioke).

See how this distinction between -u and -e can have very different meanings sometimes. If you take the word for "heal"; you can get:

gere(healing)
geru(drug)

Or another example ("skrib")

skriba (writer)
skribu (pencil)
skribe (writing)

So there you go. As for the verbs:

acetas accept [non-complete present] - I am accepting
acetat accept [non-complete past] - I had been accepting or I was accepting
acetaz accept [non-complete future] - I will be accepting
acetes accept [complete present] - I accept
acetet accept [complete past] - I had accepted or I accepted
acetez accept [complete future] I will have accepted - or I will accept.
acetis accept [habitual present] I usually? accept
acetit accept [habitual past] - I usually? accepted - or I used to accept
acetiz accept [habitual future] - I will usually accept

"usually" is not normally needed in English, but it is good here for clarification purposes.

Differences between "had been accepting" or "was accepting" would be established by context.

As for the interfix "i" yes you are not the only one so that will be changed today.

You are a star! I will add your dictionary in the dictionary section of the blog too.

Thanks!!!

[Rodiniye]

And new post in the blog about word formation in Atlas! I think it is pretty interesting so that you can see how word formation works and how we can build a language with only 500ish roots, so I will copy it here!

(remember for more info: www.atlas-language.blogspot.com

Text:

Some people have been wondering whether a language with only 500ish roots (525 at the moment to be precise) can actually be efficient.

The secret, as some of you might know, is that Atlas combines these 525 roots (adding as well prepositions and numbers sometimes) in order to create all words. Some people have been wondering whether this system can actually work (and how it works), and here is a good example.

I have chosen animals, for instance, in order to show how easy and intuitive word formation is, and how the 500ish roots system works.

The "animal" root for animal is hewan. There are a few hewanan (animals) that have their own words, so they form roots by themselves: kan (dog), neko (cat), tuig (giraffe), hat (elephant), pex (fish), zunz (bird), serp (snake), hax (insect) and only a few others.

Based on some of these roots, we create all animals. The most used ones are hewan, pex, hax, zunz. Even the root mer (sea) can be used in order to be referred to animals (as we will see).

So take the word hewan, adding other roots to its left we get:

xin (line) + hewan = zebra
ats (warm, hot) + hewan = camel
guard (keep, save) + hewan = kangaroo
sar (cold) + hewan = penguin
vit (quick) + hewan = cheetah
direb (tree) + hewan = squirrel
lent (slow) + hewan = snail
henk (change, modify) + hewan = camaleon
wuts (tail) + hewan = monkey
its (person) + hewan = chimpanzee
dorm (sleep) + hewan = bear
suiv (follow) + hewan = sheep

So you see how quickly we can produce new words, take for instance the root zunz now (bird)

vis (see, sight) + zunz = falcon
dar (catch) + zunz = eagle
noc (night) + zunz = owl
mer (sea) + zunz = seagull

or some with toot hax (insect):

vlig (fly) + hax = fly
sang (blood) + hax = mosquito
hen (one hundred) + hax = centipede
warn (color) + hax = butterfly
ner (black) + hax = beetle
ats (hot, warm) + hax = cockroach
ruq (red) + hax = ladybird



Even with mer (sea):


tang (leg) + mer = octopus
det (behind) + mer = crab
zen (value, cost) + mer = oyster
verm (close) + mer = clam
bais (white) + mer = squid
qenz (orange) + mer = prawn

So! you can see how they actually work. See how the complementing word is related somehow to the meaning. For instance, vithewana (chetaah), because these animals are very quick, or suivhewana because sheep normally follow each other (it is one of their characteristics).

Of course this system is applied for everything in the language, not only animals, but we might see some examples of that at some other point. They work, however, in the same way.

So this is how we form words in Atlas, and you can see how easy it can get. I am sure that after reading this, you remember a at least the composition of a few animals. If I tell you which is the quick+animal construction referring to, would you know?

[jeff_lindqvist]

If Rodiniye and Tommus joined forces and gave a talk at the next Polyglot Gathering, I'd be in the audience.

[tommus]

Rodiniye has a new section on the Atlas website about making new words and asked people to suggest new words. So I tried a couple. That short process produced some comments and questions.

beaver (tree eating animal): direbociashewana
shark (dangerous fish): hazaripexa
great white shark (big white dangerous fish): besibaisihazaripexa

My reaction with these words, and even with the shorter animal names that Rodiniye has just added is this: Everything is getting run together. It is resembling German and Dutch where words often get composed of one or many adjectives plus the key word. Often that makes for long or even very long words. Such words are difficult for people (especially learners) and computers to parse and figure out. I am still finding that a challenge in Dutch. But in Atlas, I believe it could become much worse. Atlas is a language of simplicity. So I propose that words (especially complex nouns) be handled in the elegant way that longer numbers are handled in Atlas.

7258: sietkel-doihen-piatdas-uit

so now:

beaver: direbo-cias-hewana
shark: hazari-pexa
great white shark: besi-baisi-hazari-pexa

To me, that is a whole lot easier to write, and to read and understand. Pronunciation does not change in sound, but also becomes much easier to say. Then if you need to add an adjective to an existing word, it gets separated from the hyphenated word (unless it becomes a fundamental word by itself. So:

fast great white shark: xin besi-baisi-hazari-pexa

Now, a fundamental question. Since Atlas is a language where you can put together parts to make new words, will there be a "fixed" dictionary? For example:

Rodiniye added the new word for camel as:

ats (warm, hot) + hewan = camel (full word = atshewana)

But someone else may not know the word for camel and thus makes up the word:

iaxu (sand) + hewan = camel (full word - iaxuhewana) [I don't know the word for "desert" but that would be even better.]

So the questions are:

1. Will there be a rigidly-fixed vocabulary for Atlas?
2. If not, I could see "regional" dialogues would grow up, using different words, such as the Brits driving lorries and the Canadians/Americans driving trucks. Is that a danger?
3. Even if there is a fixed vocabulary, will people still tend to make their own, especially for the rarer words?
4. With no fixed vocabulary, except for a few hundred root words, does Atlas need a large fixed vocabulary? As long as everyone knows all the roots and the modifiers, is it OK to have all manner of different names for the same thing, as long as everyone understands? So I could have a "sand animal" and Rodiniye can have his "hot animal" and someone else can have his "desert animal". Context will undoubtedly help out too. And if the hyphens are used, as I am suggesting, people can easily keep track of the accepted roots and accepted modifiers, and learn them completely. That would go a long way to making Atlas easy to learn.

[Rodiniye]

1. Will there be a rigidly-fixed vocabulary for Atlas?
2. If not, I could see "regional" dialogues would grow up, using different words, such as the Brits driving lorries and the Canadians/Americans driving trucks. Is that a danger?
3. Even if there is a fixed vocabulary, will people still tend to make their own, especially for the rarer words?
4. With no fixed vocabulary, except for a few hundred root words, does Atlas need a large fixed vocabulary? As long as everyone knows all the roots and the modifiers, is it OK to have all manner of different names for the same thing, as long as everyone understands? So I could have a "sand animal" and Rodiniye can have his "hot animal" and someone else can have his "desert animal". Context will undoubtedly help out too. And if the hyphens are used, as I am suggesting, people can easily keep track of the accepted roots and accepted modifiers, and learn them completely. That would go a long way to making Atlas easy to learn.

1. It is unavoidable that once words are in the dictionary, they should be used in that sense. However, if they are not known, the speaker is free to try and make his/her own word of course. If speakers were tending to use other words instead of the ones in the dictionary (the "camel" example) then they could be changed of course. Atlas is designed to be a dynamic language.

2. Not really. It could happen that people tend to use two different words for the same concept. It happens as well in natural language.

3. Possibly for new words, but for common words I would expect people to use the dictionary or to propose new words or even changes. If there is a strong evidence that something would have to be changed, then so be it.

4. That is the key. So let's say root words are the most common. From them, we can possibly get 2000-3000 very common words too. But we all known a language has got a lot more than 3000 words. The idea is that people should stick to the 500 basic roots, then try to stick to the words in the dictionary, and make words up for:
a) what they do not know
b) what is not in the dictionary

As you mentioned, if you do not know "camel", you could try to make it up, and probably it could be understood by either luck or context. The intention is that you could refer to the dictionary later on and find the correct combination, if there is one. If not, keep using it or even propose your new word.

---

As for the creation of words, let me point out!

1) only basic roots used (the last one taking the appropiate suffixes). So in your example, hazari-pexa would be hazarpexa.
2) I will establish this system for words longer than 2 basic roots. I had this in Rodinian already. I think 2 roots are ok to handle, but it would definately help with longer words. So direbo-cias-hewana would be direb-ci-hewana (only basic roots), wereas direbhewana (squirrel) would still be without it.

PS: We have already seen here what you were discussing. "shark" is in the dictionary zanpexa (tooth-fish), but you used hazarpexa, which I would have totally understood.

[tommus]

once words are in the dictionary, they should be used in that sense. However, if they are not known, the speaker is free to try and make his/her own word of course.

OK. I think that is perfectly reasonable. It probably means that the accepted Atlas dictionary should be established as soon as possible so that learners can learn all this basic vocabulary before they start making too many new words that might be different. And although you say that Atlas is a dynamic language, there should be a main period of vocabulary development (perhaps one year) during which the main vocabulary is "processed" (proposed, debated, adjusted, and finally, permanently set). After that, newer vocabulary can continue to grow, perhaps subject to an annual review and acceptance as permanent. Maybe there can be an annual "Atlas Day" when new vocabulary gets officially approved.

As I said before, the more that I learn about Atlas, the more I like. I think the future looks very good!

In the spirit of the Atlas vocabulary becoming very logical and very easy to learn and remember, "processed and improved" by ideas and suggestions, I would encourage all those who are becoming interested in Atlas to become involved. This involvement can be here on LLorg, or on the Atlas website, or elsewhere. I would guess that the interaction here on LLorg could become very dynamic because of the very wide spectrum of language learners from all over the world in so many languages.

I hope Rodiniye will be very understanding if I propose a couple of what I think are improved words (again, words that may be easier to learn and remember).

Current vocabulary word: zanpeha = shark (toothed fish). I wonder if toothed fish more logical points to piranha, which is basically all teeth? My suggestion for shark would be hazarpexa = dangerous fish. I recognise that not all sharks are dangerous, and all sorts of fish have teeth, etc. so it is all very debatable. It really comes down to which form most readily associates the word with the meaning.

The Atlas website constructs camel as "atshewana" (hot animal) whereas the Atlas dictionary shows "xota" = camel, from the Iranian language. Xota would not have any advantages to remembering unless you understand Iranian, whereas a derived word such as "desert animal", or even "hot desert animal" or "humped animal" would be much easier to learn and remember. As yet, I don't think Atlas has the root words for desert or humped. So in this case, there are competing options, between a native and a derived word. I would generally favour derived words, again because of their ease in understanding.

My earlier suggestion for beaver (tree-eating animal) is not the best. Tree-chewing would perhaps be better but there is not yet a word for chewing. Flat-tailed animal might be even better for beaver.

So because of the multitude (and I really mean many, many) possibilities for derived words, there really should be a dynamic period of discussion and interaction amongst Atlas devotees before too much of the Atlas vocabulary is cast in concrete. The more logical the words, the easier they will be, and the easier Atlas will be.