(Originally appeared in Lognet 91/1)

Shoot the Engineer and Start Shipping Boxes

By Bill Gober

Following are comments on the Autumn 1990 questionnaire [with answers in italics by JCB].

Question 16: How could L6 be more expensive if published as a looseleaf notebook? It could be if it were printed in a short run with an expensive binding—the binder—compared with a 5" X 8" paperback printed in a longer run with a cheaper binding.

Question 18: What competitors cause worry? The Lojban people; see LN90/1:2.

Question 23: The [ethnic] declension already exists; it is simple and fairly handy; for these reasons, I favor keeping it. However, it is a declension and seems to me to be out of place in Loglan; if by removing this declension we can calm the clamor for more declensions, then I’m in favor of removing it. (See below for some suggestions I have relative to this declension.) No answer seems required, here.

Question 25: What about the nationality words? How do we say 'adult male Spanish human'? In this case we have a human animal. How do we put da’s predicate in the animal declension? Kalroa ce mendi ce spani humni if we don't decline human animal predicates, spani humno if we do.

Question 26: If the general declension must be implemented, make it as general as possible, in keeping with Loglan’s spirit. No answer needed.

I think that adding declensions to Loglan will have a tremendous negative effect on the language. Others have argued the technical pros and cons of declensions in Lognet; I won’t rehash those arguments. The negative effect that I foresee is an organizational one: to add declensions is to invoke another Great Morphological Revolution. I do not exaggerate; consider the changes that declensions will bring:

• The general declension, taken to its limit, will change half the primitives of the language; the animal declension adds to the changes. Perhaps 400 or 500 primitives, all told? My impression is that from L4&5 to L1, the number of primitives added or recast was lower than this. [Quite right; there were about 200.—JCB]

• Several times that number of complexes will disappear, because their metaphors will now be covered by declined primititves. Regardless of whether this change is positive or negative, it’s still huge.

All affixes, bar none, and therefore, all complexes, bar none, will have to be re-evaluated, and new complex-making schemes developed:

(1) Many of the current vowel-final affixes will now have the wrong vowel; all affixes will have to be checked for this problem, and new affixes made when the problem is found.

(2) Many new affixes will have to be made, because the language will have 1600 to 2000 new primitives, and we’ll want affixes for as many as possible. Right now I can say mismaogru for [a] group of mixers; since with declensions I would use the primitive ?miksi instead of mismao, I’d like a 3-letter affix to use in making the complex for 'group of mixers'. There may not be enough 3-letter affixes available to meet the demand: only 1965 are possible, and 884 are already in use.

(3) Nor will it work to use the primitive itself as the affix: since the final vowel of a primitive will now bear semantic weight, we can no longer drop the vowel and hyphenate—?miksygru is out. ?Miksigru resolves into mik+sigru, millionth part of a cigarette. There must be either some means of using a primitive non-finally in a complex while keeping the primitive’s final, meaningful vowel, or some completely new scheme for generating short affixes.

Taken together, these three entail a complete reworking of the complex-building system...which was the crux of the GMR.

Implementation of declensions in Loglan will require a major revamping of the language, which will take it off-line and keep it from being used. The GMR took ten years; can this Declensional Revolution be finished in any less time? Regardless of how long the DR would take, what problem would it resolve? Less than a year has passed since the Going Public Again, and the logli are few in number; the language hasn’t been used enough to discover any problems so major!

The premise of the GPA was that Loglan is ready for use; if it’s ready for use, why is it still being fiddled with?

In Lognet 90/3, JCB assures us that Loglan is no longer a plan for a language, but a language; you fiddle with plans, you speak a language.

In L1, Loglan is called the first of the engineered languages; I’m an engineer by profession, and we have a saying: 'In the life of any project, the time comes when you have to shoot the engineer and start shipping boxes.' Meaning that, if permitted to, the engineer will tinker with the product indefinitely...and no one will ever get to use it.

Let’s not tinker with the language—let’s use it! [As announced elsewhere in this issue, the Keugru's decision on the declension proposal will be ready for announcement in the next issue.—JCB]

Too-Closely Packed Primitives: I’ve discovered some primitives that are too-closely packed. I stumbled upon the pair cirna and sirna while using MacTeach2. This discovery prompted me to write a program to look for others. Apart from the ethnic triplets (merka/merki/merko, et al.), which I expected, I found the following pairs:

canse  sanse cirna  sirna

curdi  surdi tulpa  tulpi

(Tulpa is from the sample borrowings in L1; tulpi from the primitives.) [*Tulpa is a misprint; the word is tulpi. The other three pairs are genuine fugitives from the unpacking bench. I'm not sure there is much we can do about them now. Perhaps see which final vowel change would lose the least recognizability? But Djori Gober certainly deserves our thanks for "apprehending" them.—JCB]

A Suggested Extension to the Ethnic Declension: I suggest that the ending -e be allowed on an "ethnic" word with the meaning of 'is a territory inhabited/frequented by (the people in question)'. E.g., frase = 'French territory'.

The greatest immediate use I see for the e-ending is in natural language translation. Currently, to translate the English 'French river' I must say [me]la Frans, vrici or vrici je la Frans. Usually, I won’t want to say fraso vrici because the river is generally independent of French culture. The same is true of frasa vrici and frasi vrici. The river is "French" because French people live around it; this is what the e-ending conveys that the current ethnic endings can’t.

Thus, in the Loglan tradition of going beyond the natural languages by accommodating them, the e-ending has a further charm: it lets us talk compactly about landscape features in terms of the people who live around them. For instance, the English 'The rivers of Indian Country flow clean and pure' translates as Lo rinde vrici ga kroli go klini ze punfo. We can talk about the streets of a Chinese neighborhood as lo junge trida. And a logli in New Orleans, when asked where the best restaurants and clubs are, can reply vi la Frase.

A Way to Make the Ethnic Declension Analytical: The ethnic declension is unanalytical: the final vowel of an ethnic word is, in effect a morpheme. In no other part of Loglan does the final vowel of a primitive or borrowing bear semantic weight.

On the other hand, it is handy and mnemonically useful for these closely related ideas to be so alike in form: a frasi is a member of fraso and speaks frasa; the alternative terms that are currently available would be frasypeu, frasykultu, and frasyleu.

Is there a way to split the difference? To have analyticity and mnemonic similarity? One way would be to use a series of affixes that differ only in their final vowel. These would not be abbreviations of any primitive, but would be created ex nihilo for the sole purpose of reproducing the ethnic declension using affixes.

Here are the affix-series that are still available:

cpa   cpi   cpo   cpe

doa   doi   doo   doe

gia   gii   gio   gie   giu

je   jei    je   jee jeu

joa   joi   joo   joe   jou

nua   nui   nuo

vra   vri   vro   vru

vua   vui   vuo   vue   vuu

zba   zbi   zbo   zbe

zea   zei   zeo   zee   zeu

zia   zii   zio   zi   ziu

zoa   zoi   zoo   zoe   zou

zua   zui   zuo   zue   zuu

zva   zvi   zvo   zve   zvu

My favorite is vua etc., for these reasons:

(1) It’s vowel-rich.

(2) It has all five vowel-endings...valuable for future growth.

(3) For the vowel-endings that are currently in use, [the affix] can be pronounced as a monosyllable [so] the stress doesn’t have to shift between the words that use different affixes in the series.

This has the added charm of enabling us to make ethnic terms out of any predicate, including place or descriptive predicates.

Some examples:

(1) On the east coast of North America, archeologists have discovered the remains of an extinct people, which they’ve named the Red Paint People (for the red ochre found oozing out of their burial mounds). Under this scheme, in Loglan we call them redpinvui; their culture is redpinvuo; and (to use my earlier suggestion) their territory is redpinvue. Note that a member of the affix-series differs semantically from the existing affix with nearly the same meaning. Redpinpeu means someone associated with red paint, while redpinvui means a member of the Red Paint people; similarly for the others: the ethnic series affix says something about the Red Paint people, while its counterpart says something about red paint.

(2) Kirk Sattley reports (LN90/1:18-19) that someone on the BIX computer net has declared Loglan the official language of the Mars colony [See my remarks in the opening. —Tisra]; so what will the colonists call themselves? The first logli astronomer is liable to claim marsi for [things that pertain to] the planet itself, thus seizing the most obvious predicate for a colonist. Unless the colonists insist on borrowing 'Mars' a second time to get an ethnic predicate for themselves (and I haven’t a clue what that second borrowing might look like), they can call themselves marsyvui; this makes their dialect of Loglan marsyvua, their culture (including hoverboarding down Olympus Mons?) marsyvuo, and their enclaves in other worlds marsyvue.

(3) The Loglan for Luna, Earth’s moon, is lunra. In the Loglan version of The Moon is a Harsh Mistress, the Loonies will be lunryvui. Well, maybe not throughout the book. At first, most Loonies didn’t think of Luna as home (they considered themselves exiles from Earth), in which case, lunrypeu would accurately reflect their attitude. Later, once they had declared independence and gone to war, they definitely considered themselves different from the Terrans—now it’s lo lunryvui versus lo terlyvui.

As ethnic word groups are currently made, every one is a single-source primitive or a borrowing. In the examples, borrowing is difficult (as for Mars, Luna, and Earth) or ill-advised (as for the Red Paint People); an ethnic word group couldn’t have been made, or at least, a good one couldn’t be. If we’re to keep using the ethnic declension (and I certainly would like to), we need to adopt some scheme like the ethnic affix series; otherwise, we’ll be able to make ethnic word groups only for the peoples of the present...but not for those of the past and the future. [Let me encourage you to submit this idea to the Academy.—JCB]

Copyright © 1991 by The Loglan Institute. All rights reserved.