(This is an article that originally appeared in Lognet 92/2. Used with permission of The Loglan Institute, Inc.)

Lepu Nardu Na Le Satci
(The Difficulty at the Beginning)
by James Jennings

In 1977, when I first read Loglan 1 (3rd edition, 1975), I found a short section called "Abstraction with po, pu, and zo". I remember thinking long and hard on these operators. I wasn't quite sure what subtle use one could make of them. Eventually, I gave up. After all, they didn't seem important.

Over the next decade and a half, I followed Loglan as a spectator. I tried to translate a phrase or two, and read as much as I could understand of the journal La Loglentan (aka The Loglanist), but mostly I just observed. I eagerly read the 4th edition of Loglan 1 when it came out in 1989, and then I observed some more. I received the questionnaire from The Institute that asked among other things if I would like to help edit the new dictionary. I circled "maybe".

I was a volunteer.

When I received Robert McIvor's dictionary sample, I panicked. It was full of pu and po. I looked in the new Loglan 1 under "Abstraction with po pu zo" (page 126). There were only three pages. I looked under "Abstract description with lopo lopu lozo" (page 191). Again, there were only three pages. I read the 6 pages. I read the dictionary sample. I read them all again. And again.

When I had finished commenting on the dictionary sample and had mailed it off, I found that I could write in Loglan. Not well, mind you, but something had clicked. I have come to believe that abstraction operators are psychologically the most important feature of Loglan. They are not like anything in English and you can't do without them. It is a pity that Loglan 1 didn't have more examples.

So let's do some examples. First a quick review.

Loglan 1 pages 126-129 give these examples. The "manhood" example seems a little strange. It isn't easy to figure out what an "event of being a man" is and the suggested "manhood" is certainly not a common English usage. The fact is, for English speakers at any rate, certain predicates can be fitted with a particular abstraction more easily than others. It is true that any abstraction can be applied to any predicate, but some combinations are easier for the beginner.

Suppose you wanted to generate lots of po examples. Try looking for predicates that are more "event-like": Words that describe things that happen briefly. Words of the form "X does...".

Note that in these translations we said things like "(a particular instance of) destruction", not just "destruction". If we left out the "a" in the English, we would need to rephrase the Loglan in terms of a mass description like lopo hutri.

Things that happen over a longer period of time or which don't have a sudden onset are a little harder to figure out. In these cases, you can call the po form a "state" predicate instead of an "event" predicate.

For pu examples, look for predicates that are "property-like": Words that make claims that are independent of time. Words of the form "X is ...". Predicates that seem to "name" things, as in "X is a...", are more difficult. Abstractions of these predicates seem unnatural in English. The abstraction operator zo isn't used much, so if you want to ignore it at first, there shouldn't be any problem. In fact, I had a little trouble thinking up interesting examples. The ones I did find were of the "ingredient" type: These predicates all describe amorphous things, things which require a measuring device, like a tablespoon or a ruler, to be counted or doled out. The zo.constructions, on the other hand, refer to the measurements themselves. English perhaps comes close with the notion of a "carat", which is the proportion of gold in a "gold" object. Thus in Da zo de aurmo, X could be the numerical measurement in carats of Y's "purity" as a gold object.

Examples where the property is not easily measured often make sense but aren't very satisfying in English.

Zo is probably most useful in quantitative or scientific writing. You can find some examples in the translations from Scientific American that appear in the back of Loglan 1.

Once you are comfortable with the above kinds of examples, you can try other combinations. It will sometimes require a lot of imagination.

Next I want to talk about grammar. The abstractions as used above have a very short "scope". They tend to modify the next predicate only. Note how the following phrases parse. If we want the abstraction to have a longer scope, we need to do something special. In the last example, ge has grouped the predicates to its right into a single expression, forcing pu to apply to all of bilti ckela. The same sort of things happens when predicate phrases are used in descriptions. The next example is very, very important:

When lo and pu (or po) are separate words, pu only modifies the next predicate. When lo and pu are joined together, they acquire a very long scope. Words like lopu and lopo don't just modify the next predicate; they can swallow entire sentences. Here is an example with lopo: There are two dangers when using abstract descriptions like lopo.
  1. Because lopo is so powerful, you must take care that it doesn't swallow too much. If the lopo-phrase comes just before the main predicate, it can be ended by the main predicate's tense marker. If the phrase is the last term of the sentence, it is ended by the end of the sentence. Here is an example that uses both techniques. Otherwise, you will have to end it with a terminator of some kind. Gu will work but guo is better since the sole use of guo is for ending abstract descriptions like lopo.
  2. Because the two-word phrase lo, po and the single word lopo are grammatically very different, you must be very careful to pronounce them distinctly. When you intend the short scope, you must pause (or write a comma) to separate the words (as in /lo.poPREda/). It might also help, when you want a single word, to put stress on the first syllable (as in /LOpoPREda/). [Actually, a stress before a pause often has the effect of calling attention to it, as in /leTO.MREnu/ and /LO.poPREda/.JCB]
How do you learn to see the Loglan abstractions in English text? If you're looking up single words, the Loglan dictionary will help. For example, "death" is listed as lopo morto. For larger chunks, you have to learn to see the phrases and sentences that have been swallowed whole.
  1. See Spot run.
  2. See (Spot runs).
  3. See (the event of Spot running).
  4. Vizka lepo la Spat, prano.
  1. I like drinking coffee.
  2. I like (drinking coffee).
  3. I like (events of me drinking coffee).
  4. Mi fundi lopo mi hompi lo skafi.
The last two examples had "event-descriptions" as arguments. Here is an ordinary description with short-scope po.
  1. With a sudden leap, he was gone.
  2. With a sudden (event of jumping), he was gone.
  3. Tie le sudna po volti, da pa godzi.
Here is an example with lezo.
  1. It is longer than the ocean is wide.
  2. It is longer than (the ocean is wide.)
  3. It is longer than (the amount by which the ocean is wide.)
  4. Da langa lezo le mursi ga kubra.
Lastly, I'd like to show you a nifty example where a simple English statement seems to call for a lepu phrase in Loglan. Alas, I can't find one. As near as I can tell, English property abstractions are all single words. Whenever I've thought of an example of a lepu phrase that seemed to translate into natural English, it was easy to rewrite the Loglan without the lepu. Of course, I could be wrong. If anyone comes up with a good lepu example, I'd love to hear it.

I hope that these examples help you get over the hump. Try playing with them. Invent your own. Have fun. And remember,

Lopo pifkao ga madzo lopu purfe.


jennings@halcyon.com