New auxlang: Atlas

[Rodiniye]

I will finally have to update the webpage tomorrow. I will probably do a new post on permanent and non-permanent verbs, since I think it is hot topic at the moment! Countries will come next (probably towards the end of next week).

Was just going over the word "begu" now (aubergine/eggplant), why is that a -u and not an -o?

Both would be possible. However, "-o" ending means the described reality is alive. When using "-u", you are assuming it is "dead", or "taken from its plant", etc. So "begu" would be the aubergine you have on the table, "bego" would be one which is still growing.

Same for instance with "itsa" (person). "itsu" would be a corpse.

Same with food. "kuka" is chicken, the animal. You call it "kuku" when it is on the table ready to be eaten.

Here are examples for the next few derived roots:

I like it!

aikez >>> will be functioning, will be working

So here are those 48. I'd like some feedback, especially on whether I am getting the verb tenses right.

I know it is just a way of "marking" it in the dictionary, but remember that non-permanent actions can be translated by future simple in English too, as in:

"tomorrow I will go to Milan" - Wi vesez zu Milanu hoa".


Verbs do not always have direct translation. However I agree that "continuous" forms can be used in the dictionary as a preference in order to mark that the action is not permanent.

What I would do is:
- for permanent verbs (ending in "is"): translate with the simple tense, perhaps adding "be used to":

wi vesis: I go/I am used to going

- for non permanent (ending in "es"): translate it with a continuous tense plus a simple tense:

wi veses: I am going/I go

That's how I would do it I think!

[tommus]

wi vesis: I go/I am used to going

I think for this first dictionary of "derived roots" that we should stick to a single translation of the verbs for two main reasons: it will take less time to do all these words, and when the popup dictionary is used, the popup translation will be a bit shorter and more quickly understood. Also, in a probable future function of the popup dictionary (to translate whole sentences), it looks awkward to have two translations. I suppose in this case, just one could be displayed. In any case, time to complete the first version is the main issue I think. And the user can probably figure out equivalent meanings.

[crush]

What I would do is:
- for permanent verbs (ending in "is"): translate with the simple tense, perhaps adding "be used to":

wi vesis: I go/I am used to going

- for non permanent (ending in "es"): translate it with a continuous tense plus a simple tense:

wi veses: I am going/I go

I think "tend to" or "regularly" might be clearer, eg. "I (regularly) go".

For veses, i think just the "to be going" form is fine. In English, if you say "I go" it's generally a truth or something you do regularly, eg. "I go to the gym every Thursday." Can you think of an instance where "veses" could be translated as "go" rather than "going"? In the past tense it makes sense to have both forms:
wi veset: I went (once), I was going
.. but not in the present tense, in my opinion.

[tommus]

I think "tend to" or "regularly" might be clearer, eg. "I (regularly) go".
For veses, i think just the "to be going" form is fine. In English, if you say "I go" it's generally a truth or something you do regularly, eg. "I go to the gym every Thursday." Can you think of an instance where "veses" could be translated as "go" rather than "going"? In the past tense it makes sense to have both forms:
wi veset: I went (once), I was going
.. but not in the present tense, in my opinion.

These are a few of the reasons I still think that Atlas would be simpler, yet just as powerful and understandable, if there was a single form rather than permanent/non-permanent. Many users will not make a distinction or will put supporting words and context around the verb to illustrate if it is specifically permanent or non-permanent. For example:

wi vesis: I go
wi addem vesis: I normally go, I usually go

[tommus]

I think we may still have an issue with adverbs. The grammar says there are some fixed adverbs: duzu (still), zuen (already), haiek (only), cast (often), itzei (too, as well), nithe (neither), shedon (almost), dou (even), dahi (again), an (on), av (off).

And non-fixed adverbs: " ... they are derived from basic roots, adding the morpheme –em: vitem (quickly), lentem (slowly),"

However, we discussed several times here on the Forum that adjectives would have "ik" endings. So all the "derived roots" I am currently working with have "ik" endings. Are there now two kinds of non-fixed adverbs, those ending with "em" and meaning a translation with "ly", and other adverbs ending with "ik" that have some other meaning? This needs to be clarified.

[Rodiniye]

However, we discussed several times here on the Forum that adjectives would have "ik" endings. So all the "derived roots" I am currently working with have "ik" endings. Are there now two kinds of non-fixed adverbs, those ending with "em" and meaning a translation with "ly", and other adverbs ending with "ik" that have some other meaning? This needs to be clarified.

Is it possible that you are looking at an older version of the grammar? the new one (31st august) includes different adverbs and "-ik" ending for the regular ones. The other ones are fixed forms, 3 letters, ending in -a (zia=already, iza=too...)

Anyway, there are no adverbs ending in -em any longer. That ending was moved in order to mark connectors, as discussed previously. The list of connectors will be pretty much closed with all (or most) of the possible connectors ending in -em (wanem=to finish with, ekeem=first of all...), etc.

wi addem vesis: I normally go, I usually go

*should be: wi addik vesis as mentioned above.

I understand the reasoning behind it but let's stick with Atlas as it is. Adverb usage in order to mark these kind of things would not last long I believe. It is easier to have a system with 2 really easy aspects that encouraging people to use different adverbs for every different situation. We could say the same with time. Why bother having present, past and future? You could use adverbs. The reality is that most languages end up marking these things in languages, be it through suffixes, adverbs or other particles, and not having suffixes is not always the easiest option. Compare having "z" as a future to English having "will", "be going to" or present continuous usage marking future.

This is what I believe. We went through a period in which all changes were made, let's now stick to it and modify only what's really necessary. I really think it's good as it is!

I think for this first dictionary of "derived roots" that we should stick to a single translation of the verbs for two main reasons:

That's fine

[tommus]

Is it possible that you are looking at an older version of the grammar? the new one (31st august) includes different adverbs and "-ik" ending for the regular ones.

Now I'm confused as to which one I was looking at. On 31 August, I downloaded this grammar:

20170831-Atlasgrammar-v5.pdf

At the moment, the grammar on the website is:

20170817-Atlasgrammar-v5.pdf, and inside the pdf it says "v5 - 31aug17".

In any case, I'm glad that the new grammar has "ik" for adverbs, plus the other changes.

[tommus]

I am working on the "derived roots". I have completed about 500 of the over 8,000, so it is going to take some time.

I put a little program together to get an idea of what "derived words" should be "derived" next. So I took a list of the 1000 most common English words and compared it with the Atlas words available so far. There are about 427 of the 1000 that don't yet exist in Atlas. I think this is part of the reason I found it so difficult to write that short paragraph about the hurricane. There really are a lot of very common words that we don't have in Atlas yet. Some of these missing words will probably be created with the completion of the "derived roots" but many of them won't. For discussion purposes, here is the list of the 427 words:

Code: Select all

able
about
according
action
actually
address
affect
agency
agent
ago
ahead
along
also
always
analysis
anyone
anything
apply
argue
arrive
article
artist
assume
attorney
authority
away
ball
bar
beat
benefit
best
better
beyond
billion
board
both
box
bring
budget
camera
campaign
candidate
capital
card
career
case
center
central
century
certainly
chance
choice
church
citizen
civil
claim
clearly
coach
collection
college
commercial
community
compare
concern
Congress
consider
consumer
continue
control
could
couple
course
court
crime
cultural
current
customer
deal
death
debate
decade
defense
degree
democratic
describe
despite
detail
develop
development
different
dinner
director
disease
doctor
down
drop
drug
early
economic
edge
education
either
else
employee
enter
entire
environment
environmental
especially
ever
every
everybody
everyone
everything
evidence
exactly
example
executive
exist
expect
expert
explain
fast
federal
figure
fill
final
finally
financial
focus
foreign
former
forward
fund
general
generation
government
great
ground
growth
gun
guy
heat
herself
him
himself
home
hospital
hour
huge
husband
identify
impact
improve
including
increase
indeed
individual
industry
instead
institution
interesting
international
into
investment
involve
item
its
itself
job
just
kid
kind
kitchen
knowledge
large
late
later
lawyer
lead
leader
least
level
likely
local
loss
lot
magazine
main
maintain
major
majority
many
market
matter
may
maybe
media
medical
mention
message
might
military
model
modern
movement
movie
much
music
must
my
myself
nation
national
nature
nearly
necessary
network
never
none
note
nothing
now
officer
official
oh
oil
ok
once
onto
option
others
our
owner
page
painting
parent
participant
particular
particularly
partner
patient
pattern
people
perform
performance
perhaps
personal
physical
policy
political
popular
population
position
positive
possible
prepare
president
pressure
pretty
prevent
private
probably
production
professional
professor
program
project
property
protect
provide
purpose
quite
race
radio
raise
range
rate
reality
realize
really
recent
recently
reduce
reflect
region
relate
religious
remain
represent
Republican
require
research
resource
respond
response
responsibility
reveal
rich
rise
rock
room
school
science
scientist
score
season
security
seek
senior
series
serious
serve
set
several
sexual
shake
share
shot
significant
simply
since
sing
single
site
social
society
soldier
somebody
someone
something
sometimes
soon
sort
source
southern
special
specific
speech
spend
staff
stage
stand
standard
start
statement
station
stay
step
stock
store
structure
student
stuff
subject
successful
such
suddenly
suffer
suggest
support
sure
talk
task
technology
ten
tend
test
than
thank
their
them
themselves
then
theory
thing
though
thought
throughout
thus
tonight
top
total
tough
trade
traditional
training
treatment
trial
trip
trouble
truth
TV
type
unit
until
upon
usually
various
victim
view
violence
visit
vote
wear
western
what
whatever
whether
which
whom
whose
why
wide
wife
will
wish
within
wonder
worker
would
writer
wrong
yard
yet
your
yourself

[crush]

Many of those words are already in the main roots, i believe. For example:
able - should be related to "das" (dase = skill)
about - might be translated with "de"?
action - might be duse (act of doing)
also - iza (too)
couple - duea (a couple as in people), due (two)
death - more (mores - die)
decade - there's no specific word for ten, so i can only think of eke-wue-nane
fast - viti (vite - speed)
him - zi(wer)
home - batu (house) or perhaps (ialbatu - family-house)
hospital - bokbinu (sick-building)
hour - some variant of qe?
its - de zi
itself - zi mem
job - kaqe (work)
large - besi (big)
etc.

[tommus]

Many of those words are already in the main roots, i believe.

Yes. I agree. However, I am thinking about these from a computer translation point-of-view. For example, if a user double clicks on the English word "able", how does the computer know to make an association with "skill"? Likewise for also (too), decade(?), hour (qe), etc. And people will also have difficulties coming up with a possible alternative if the word doesn't appear in a dictionary. So I think most of these have to be fleshed-out in a dictionary. Otherwise, picture how someone goes about writing some Atlas text. They want to translate a word that isn't in a list or in a popup dictionary application. So they think of some synonym to the word they need. They have to keep trying alternatives in hope of finding one that has an Atlas translation. And by then, it may not be very close to the word they want.

And these are only words from the 1000 most common. You need many more than that. If we had a list of the 5000 most common English words, how many would exist in Atlas? So the dictionaries will have to be populated.