These notes have not been spell-checked or proofread, so they are likely to contain errors. They were an attempt to capture some of the numerous thoughts that pass through the mind during the language design process. I kept this log during the spring and summer of 1998.
©1998 R.K.H.
I decided to make a few changes in Vorlin's numerals.
First it occurred to me that men (>1) and pen (5) are too similar. I decide to throw out pen rather than men as the latter is closely tied to the grammar and fairly well ingrained in my memory.
Looking critically at the other numeral words, I saw several that I was not attached to. sor (6) was from the so of Loglan, a project that I cannot identify with. sep (7) was too euroclonish.
And then there was tri. This was an anomaly: not CVC; different from the others. For years I had left it in place as a filter for certain personality-types; they would see this inconsistent item, bark at it like the predictable watchdogs they are, and I would hint that they should go bother someone else. But the fun had gone out of this procedure, and tri seemed too euroclonish, so it was time for a change.
But first I changed nul (0) to nol. Slightly easier for me to pronounce, for some unknown reason, and less euroclonish. I decided to leave yun (1), dus (2) and dek (10) unaltered as they dated back to Day One and were firmly embedded in my memory. Granted, dek is very euroclonish, but I don't want to be too extreme in my anti-euroclonicity.
I decided I wanted the vowels in the 1 through 10 words to make some sort of pleasing pattern; that I wanted the 1 through 9 words to begin with 9 different letters; and of course that none of the numeral words could rhyme with any of the others. Also I wanted most of these items to end in consonants that would sound good if butted up against dek, bak, kil, milon, gig, keeping in mind that abutments with dek would be more common than others.
For a long time I had been considering changing tri to sam or san, and I decided to go ahead and select san. If I wanted to continue the vowel pattern established by yun and dus, this meant I would need another "a" word for 4. It boiled down to kat or kad, and I decided to use kad because I didn't want to bump the concept "cat" from kat to some other word.
The next item to come into view was nev (9), found on Mark Rosenfelder's web page that gives the numbers 1 through 10 in more than 2000 languages. nev is from Verdurian, Mark's own conlang, and is similar to nef found in various French creoles. Since Mark and I are both authors of usenet FAQ documents, it seemed appropriate that I should make the fraternal gesture of borrowing a word from one of his languages.
This left me with 5, 6, 7, 8 unworded. Two of these would need to have "i" as their vowel and two would need to have "o." When it became clear that a pattern of i, i, o, o was not likely to come together, I hit upon lim (5), tor (6), zib (7), leaving me with a need to fill the 8 slot with an "o" word.
Unable to find a suitable borrowing, I created the sui generis hog, which resembles the familiar ok, oct-, acht, and so forth without being too euroclonish. The "h" is seen (but not necessarily heard) at the beginning of a few natlangs' words for 8.
And the result of all this is...
nol 0 (formerly nul)
yun 1
dus 2
san 3 (formerly tri)
kad 4 (formerly fir)
lim 5 (formerly pen)
tor 6 (formerly sor)
zib 7 (formerly sep)
hog 8 (formerly hac)
nev 9 (formerly nin)
dek 10
After making these changes, I carried my notes around with me for several days and considered these new words, to make sure I could live with them. Only after I felt the new system was fairly practical did I bother to type them into the computer.
Combining these numerals with some tentative nouns, we get the following wordplays...
nol kol: no cabbage / zero cabbages
yun kun: one ability
dus mus: two mice
san mikan: three tangerines
kad kat: four cats
lim lin: five languages
tor dor: six doors
zib sib: seven siblings
hog bog: eight bows
nev rev: nine dreams
dek cek: ten cheques
bak fak: a hundred compartments
kil bil: a thousand cars
milon limon: a million lemons
To date I have been classifying all the numeral words as one part of speech, but if we are going to have a parser that doesn't accept gibberish it will eventually be necessary to distinguish the 1-9 words from the powers-of-ten words, because they behave differently in sentences; and words like nol and men are yet another category, e.g. tordek is acceptable, dekdek or noldek would be outlandish... mendek should be tolerable in casual environments, I think.
While defining the Vorlin word ven, I found the dictionary definitions that I could find for the corresponding English word were unsatisfactory. The lexicographers gave very broad and blurry definitions, such as `one attached to another by affection or esteem' (Webster's Ninth New Collegiate).
I use the English word "friend" to describe someone whose company I enjoy and whom I would help if he needed help, even if it were difficult or incovenient for me to provide such assistance. As talkshow host Bruce Williams has said, if you have four or five people in your life that you can call friends by this definition, you are extraordinarily fortunate.
English "friend" has a second meaning; it refers to someone who is on the same side in a competition, as in "Friends of Ido" or "Who goes there, friend or foe?"
A Loglanist told me in correspondence some time ago that I seem to be tending toward "over-specification." Well, what's the alternative? If you don't take the time to write proper definitions of your morphemes, you end up with people assuming that your language is meant to be a word-for-word encoding of their native language. Look at Esperanto's artikolo, for example. This can refer to anything ranging from a grammatical function word (such as the definite article), to an item in a newspaper (an article), or an "article" of clothing. Do these meanings really have anything in common? Yes, they are all contained in the English word article! Apart from that, they appear to be disparate meanings that should not be residing in the same morpheme. I hope that writing good definitions of Vorlin morphemes will reduce their rate of semantic decay. Granted, a certain amount of metaphorical extension is inevitable and necessary.
Spent some time contemplating the four basic compass directions. These words need to begin with four different letters and sound reasonably distinct from each other. Kept nor and zud (loosely adopted from Germanic languages) for "north" and "south." Considered timor and barat (loosely taken from Malay-Indonesian) for "east" and "west" but decided that these items should be monosyllables. Therefore kept duq (from Asian languages) for "east" and synthesized bat for "west."
There is a coincidental pattern in the first letters of nor, zud, duq, bat. The letter d can be made into b by flipping it around its vertical axis. Upper-case N can be converted to Z by rotating it 90 degrees counterclockwise.
I ponder the possibilities of conjunction design and the ramifications on syntax.
Should there be a distinction between distributive and collective meanings of "and," i.e. two differemt words to show distinctly the two possible meanings of "John and Mary bought a house"? Could this be accomplished with an "and root" plus aspectual affixes? Is "add" expressable as (and)+o?
Is the potential syntactic disambiguation produced by correlative conjunctions worth the extra verbosity, i.e. is it wise to require one to say the equivalent of "both X and Y" instead of just "X and Y", to say "either X or Y" instead of merely "X or Y" and so forth?
What exactly (within the realm of Vorlin only) is the difference between conjunctions and prepositions? Is it really necessary to have a distinct class of words called conjunctions, or could conjunctive functions be handled by prepositions? Well, I guess not. (Reminder to self: "if ... then ..." and other sentence-level conjunctions debunk this line of thinking.)
My thoughts now turn to words that refer to professionals, enthusiastic amateurs, and people who have certain qualities or are members of certain groups. Words like banker, stamp collector, salesman, drunkard, and Swede.
The usual sources of inspiration are disappointing on this issue. As I often do, I begin studying the semantic field in question by checking to see how Chinese has handled it.
In Mandarin, several morphemes are used in compounds to indicate professionals, the agents of verbs, and people who have a certain noteworthy quality. There is some overlap in meaning and it is not always possible predict which of these would be correct.
zhe3 usually refers to what I call the "casual agent," person who happens to be doing X at the moment but is not a professional or habitual do-er of X. zhe can also mean "person who is distinguished from others by the named quality," as in dàzhe "the big guy." jia1 usually refers to skilled and specialized trades. yuán is kind of a catch-all. There are other words including fu1 which usually refers to semi-skilled physical trades. Some examples:
xuézhe scholar
dúzhe reader (of a periodical, etc)
kexuéjia scientist
huàjia painter (in the fine arts sense of the word)
haiyuán sailor
yùndòngyuán athlete
túfu butcher
Chinese uses rén (person) to indicate people of various nationalities: yingguórén = Englishman, wàiguórén = foreigner.
Looking at English we find the suffixes -er and -or used unpredictably to indicate casual agents, enthusiastic amateurs, and professionals. Thus the same word refers to collectors of stamps and collectors of taxes. The suffix -ist sometimes refers to partisans of ideologies (as in "socialist"), but not all partisans are described by -ist words (e.g. "liberals" and "conservatives"). -ist sometimes refers to professionals, too, as in "dentist" and "hygienist." In some cases a verb without any suffixes becomes a noun referring to a professional, e.g. "nurse" and "cook." (But wait, did "nurse" start out as a verb or a noun? I don't feel like looking it up.)
The way English formulates names for inhabitants of countries and members of ethnic groups is unpredictable: Sweden > Swede, Jordan > Jordanian, Norway > Norwegian, Iceland > Icelander, Poland > Pole, Israel > Israeli, etc. A great confusion of affixes.
Turning to Esperanto we find a bit more clarity in some areas, equal fogginess in others. The "casual agent" meaning is embodied in -into, -anto, -onto (for past, present, and future agents).
-isto refers to a professional (floristo = florist), or an enthusiast (filatelisto = stamp collector), or a partisan of an ideology (anarkiisto = anarchist). -ano refers to an inhabitant of an area (insulano = islander) or a member of a group (klubano = club-member), but sometimes it supplants -isto in referring to a professional (e.g policano = police officer) or a partisan (samideano). Some professionals are named irregularly: astronomio = astronomy and astronomo = astronomer; redakti = edit and redaktoro = editor.
-ulo means "person who is associated with the named thing." This creates a bit of overlap; -ulo could be replaced by homo in most cases and by -isto (in the "enthusiast" sense) in some other cases. Agitanto, agitemulo, aû agitisto? Wells writes that drinkulo means "drunkard," but couldn't that concept more accurately be expressed as drinkanto or drinkisto or drinkemulo?
All this leads me to believe that Vorlin should have a few well-defined nouns with the following meanings:
1. Person. Any living human.
2. Professional. A person whose job/ profession/ primary productive function in society is the named activity. (Would there be any use for this term in a truly free society, in which people were able to participate in a variety of productive activities at will instead of being condemned to employment?)
3. Enthusiast, amateur, dabbler. Someone who often does the named activity for the pleasure it brings him/her. An habitual agent motivated by intrinsic reasons, rather than being motivated by a paycheck, social dominance and submission games, etc.
4. Causal agent. Someone who just happens to be doing the named activity at the moment in question.
5. Member of named group. Stick an o on this noun and perhaps it will produce a verb meaning "is a member of ..."
This leaves a few questions unanswered. How would citizens of countries and members of ethnic groups be named? How do you distinguish between "doctor" and "nurse" -- both are healing + professional. What about non-human agents, e.g. a part of a machine might be called a "runner" or a "sweeper" or something.
In the sentence "I saw large houses and trees," there is no formal mechanism by which we can be certain whether both the trees and the houses were large, or only the houses. For several years I have been trying to solve this problem gracefully.
I think I am getting close to a solution; it involves having two words that basically mean the same thing as English "and," but one of the words is classified as a conjunction and the other is a preposition. If the conjunction is used, the modifier applies to both items; if the preposition is used, the modifier only modifies the first item.
This even makes it possible to say "I saw red houses and cars and flowers" in three different ways, depending on whether you want "red" to modify the first noun alone, the first two nouns, or all three nouns. But will this introduce more annoyances than it solves? I may have to re-think my requirement that multiple adjectives be separated by conjunctions. For example "I saw large and red and expensive houses and cars and pets" could introduce a jungle of intricacies.
I got this idea while working on "the plantation," as I call it: nolo should mean "want to avoid doing X" or "want to not do X." This will mean moving the "zero" concept from nol to some other morpheme. nil might be a nice place for it, but nil is too similar to kil (= 1000).
English lacks a word for the emotion of nolo, which is the polar opposite of vilo rather than a mere equivalent of non vilo. There's a difference between simple negation (in this case, referring to the failure to feel a particular emotion) and a polar opposite. Since English doesn't have a word for the opposite emotion of "I want to," we must struggle along with the inadequate expression "I don't want to." It's as if we had to say "don't love" when we really mean "hate."
At least, that's how I feel about it. I am sure there is much room for quibbling about the terminology of emotions. Perhaps my repertoire of emotions is not the same as yours, and perhaps yours is different from that of your counterpart on the other side of the world.
Several weeks ago I had wanted the words for "male" and "female" to contain o and i respectively, as those vowels are used in several yang/yin oppositions in Vorlin, e.g. the yang (transitive) verb suffix -o and the yin (intransitive) -i, the intensive -oz and the attenuative -iz. But I could not find suitable borrowed words for "male" and "female" containing the desired vowels, and I did not feel inclined to synthesize new items, so I kept fem and changed Old Vorlin's vir to han.
More recently I changed her to har so that it would contain the same vowel as han. But then I changed it back to her when I remembered that her is used several times in An Introduction to Vorlin. I am trying to stabilize the morphemes used in that document. So, although I am not totally happy with fra and her, I will leave them alone for now; if others are unhappy with them also, they can propose alternatives later on.
Everytime I look at An Introduction to Vorlin I discover a discrepancy between it and the dictionary file; this is the terrible price I pay for my perpetual tinkering. This time I see hav in Intro and hab in the dictionary. I'm changing the dictionary to include hav.
Added dum (= thinking/thought) to the vocabulary as suggested by Evgueni Sklyanin. I recall dum was a personal pronoun in the 1991 version of Vorlin.
An idea that I've been considering for several years is to reserve -os, -um and -us for use as stress-shifting suffixes that indicate half-assimilated loan-words from Greek, Latin and Romance sources. These words would be stressed on the penultimate syllable and would not be required to have perfect CVCVC(VC) shapes. For example, kapsikum from the botannical name Capsicum could refer to chili peppers and korvus from zoölogical Corvus could refer to crows. Although satelit cannot be added to the lexicon because it gives the false impression that it contains the patient suffix -it, there's no reason why satelitos couldn't be used as an interim load-word; eventually it would either enter the lexicon or be replaced by a brilliant compound of one-syllable CVC morphemes.
Now have 234 morphemes in the reconstructed vocabulary.
Reminder to self: write about the relationship between electrons and semantics.
GeoCities, a company which provided free web space for Vorlin and some of my other web projects, has begun slapping more and more advertisements on the pages lately. Although they only affect people using the most modern browsers, these pop-up ads and constantly-visible watermarks are quite annoying, so I'm making arrangements to move the material elsewhere.
I'm going to take a break from keeping this journal for a while; might resume later.