Properties

We have seen, here and there, instances of Lojban expressions of properties. Lojban treats properties as abstractions, introduced by ka. There is nothing controversial about that; properties are things you can talk about (sumti), which involve relationships and characteristics (selbri.) So if xendo means 'kind', for instance, le ka xendo refers to 'kindness'.

The thing about properties, though, is that they are properties of something. They are associated, not just with a selbri, but with a particular place of the selbri. For example, kindness is not just le ka xendo, but the property of someone displaying kindness — as a characteristic of that someone. In other words, not just le ka xendo, but le ka ___ xendo, where ___ stands in for that 'someone'.

As a further example, consider influence and susceptibility. Both involve the relationship expressed in Lojban as xlura:

x1 (agent) influences/lures/tempts x2 into action/state x3 by influence/threat/lure x4

So the Lojban for influence is le ka xlura. And the Lojban for susceptibility is... le ka xlura? Strictly speaking, yes: both properties involve the same bridi, xlura.

But obviously, we can't have the same expression for both influence and susceptibility; we have to have a way of highlighting the place in the bridi we are interested in. Though the two properties involve the same bridi, they focus on different places of that bridi. Influence is the property associated with the x1 of xlura, the influencer. Susceptibility is the property associated with the x2 of xlura, the influencee. So how do we say that in Lojban?

Lojban's solution to this problem is fairly similar to Lojban's approach to questions, as it turns out. Remember in Lesson 13 that the search for extraterrestrial intelligence was, in terms of Lojban, a search for the value to fit in the slot

leka ___ terdi bartu pensi

By the same token, influence is a property of things that fit into the x1 place of xlura; so you can think of influence as leka ___ xlura. If we know that mi fits into the slot, we have 'my influence'; if we know that la fred. fits into the slot, we have 'Fred's influence'. And susceptibility is a property of things that fit into the x2 place of xlura; so you can think of susceptibility as le ka xlura ___ (or le ka ___ se xlura.)

Lojban has a word for that slot associated with properties. It isn't ma, because you're not asking someone what fills the slot; you're just pointing out that there's a slot there that can be filled. It isn't ke'a either, because ke'a refers back to something you've already expressed as a sumti (though you might think of a relative clause as a property belonging to that sumti.) Property slots get their own KOhA cmavo, ce'u. So:

Lambda Note, Part 1: If you:

  • did Computer Science at University, and you didn't skip Theory of Computation in third year just because it had all sorts of strange Greek letters and ivory tower mathematics in it;

  • did Computer Science at University, and skipped Theory of Computation in third year, but hacked around with LISP a lot anyway;

  • did Linguistics at University, and did not run screaming from the Formal Semantics elective in third year (if you were even offered it) just because it had all sorts of strange Greek letters and more mathematics than you were used to (i.e. none);

then it will mean something to you that ce'u is a lambda variable, and that

le ka ce'u xlura da de di

corresponds to

λx.xendo(x,da,de,di)

The rest of you (which includes 90% of all programmers and 99% of all linguists) can go ahead and forget I ever mentioned this.

If you cast your mind back to Lesson 7, you'll remember that we split up the abstractions Lojban uses into two main types: events, using nu, and facts or propositions, using du'u. A property, as introduced by ka, is still what we called there a reification. That means it's just like du'u: it's something you hold in your mind about what happens in the world, rather than something that objectively happens in the world. The difference is, ka has an empty slot, occupied by ce'u; and you're interested in the ka-clause only inasmuch as you're interested in what fills the slot. On the other hand, du'u-clauses don't necessarily have any such slot — although they can.

Note: This means that, when you get down to it, there is no real difference between ledu'u ce'u xendo and leka ce'u xendo. But as we discuss below, there is a real difference between ledu'u xendo and leka xendo: by default, ka is assumed to contain ce'u somewhere (since it is a property of something.) No such assumption is made for du'u: ledu'u xendo is normally assumed to be just ledu'u zo'e xendo; the fact that someone is kind, rather than the property of someone being kind.

Most usage of ka in Lojban fits this pattern of 'filling a slot' straightforwardly. This is particularly the case when a ka-asbtraction is required in the place structure definition of a gismu: a ka-clause is required, because by its definition the gismu involves that slot. So with sisku 'seek', you search for ka-clauses, to find what will fill the slot. With karbi 'compare', you compare things to see how well they fit the slot. Or alternatively, the gismu by definition fills that slot, by relating the property to the value satisfying it. For example,

What happens when you find the value that fills the slot? Then — and here Lojban parts ways with English — you no longer have a slot; so you no longer have a property. You've gone back to du'u. If mi mansa do leka ce'u pensi, that's the same as saying do djuno ledu'u mi pensi. A property applying to a known entity is no longer a property at all in Lojban, but a fact — or (if you no longer have to reify it) an event.

Be careful here: what English (and in fact, most traditional usage) calls properties are often actually considered just states in Lojban — that is, something that happens in the world, but without anybody lifting a finger. Being a runner (also known as 'running') is hard work; so we're happy to think of it as an event: nu bajra. But being happy (also known as 'happiness') is something that just happens, without any work; so we're inclined to call it ka gleki. But that's misleading. English distinguishes between running and happiness grammatically, because run is a verb and happy is an adjective. But verbs and adjectives don't mean anything to Lojban (or to many other languages), so there's nothing to say you can't say nu gleki instead. Much of the time, in fact, that is precisely what you should be saying. As a rule of thumb: if you wouldn't say ka bajra in a sentence, don't say ka gleki either.

Note: For instance, is illness a quality in the sentence "Fred's illness is more debilitating than George's"? Let's use running instead. If we translated more debilitating as a single lujvo, rubri'amau, would we say leka la fred. bajra cu rubri'amau leka la djordj. bajra? No; we'd likely say lenu la fred. bajra cu rubri'amau lenu la djordj. bajra. In fact, there is a quality involved in the sentence, if you expand it out fully — but it's not the illness, but the debilitatingness: lenu la fred. bilma cu zmadu lenu la djordj. bilma kei leka ce'u rinka lenu zo'e ruble "The event of Fred being ill exceeds the event of George being ill in the quality of causing someone to be weak."

Tip: In older Lojban, you'll often see phrases like leka mi gleki for "the property of me being happy." That's because we used to not know any better (ce'u is a recent addition to the language), and were treating Lojban properties pretty much the way English does. The proper way to say this in Lojban is lenu mi gleki, or ledu'u mi gleki. Alternatively, if you want to emphasise that the property "x1 is happy" is being applied to you, you can say leka ce'u gleki kei poi ckaji mi — a literal translation of "the property 'x1 is happy' as applied to me".

Lambda Note, Part 2: The infinitesimal number of you that know about lambda calculus are by now thinking this is a pretty lame way of implementing beta-reduction. All I can say to that is, if you want LISP, you always know where to find it...

Sometimes you'll want to speak of properties of applying to two entities at once. For example, the cop wants to know who talked about the heist, and to whom:

le pulji cu djica lenu djuno ledu'u makau tavla makau le nu jemna zercpa.

In that case, he's looking for both x1s and x2s to fill in his ka-property:

le pulji cu sisku leka ce'u tavla ce'u lenu jemna zerle'a To put it more formally, he is seeking pairs {.abu, by.} such that the proposition .abu tavla by. is true.

Tip: By default, two different instances of ce'u are two distinct entities. So the example given is not saying that the police are looking to someone who talked to themselves about the heist!

The main use for multiple instances of ce'u is our old friend simxu: if we want to speak about reciprocality, we are very much interested in which two places are related through that reciprocality:

mi ce do simxu leka ce'u tavla ce'u lenu jemna zerle'a

There are some reciprocalities that can be distinguished nicely in this way: simxu leka draci fi ce'u ce'u is a situation where people take turns writing plays for each other, while simxu leka draci fo ce'u ce'u is a situation where people take turns performing plays for each other.

Note: The quantity abstractor, ni 'the amount by which...' can also take ce'u. Had we actually looked at ni in this course at all, this piece of information might have been slightly more useful to you.

Vocabulary

ckire 

x1 is grateful/thankful to/appreciative of x2 for x3 (event/property)

mamta 

x1 is a mother of x2; x1 bears/mothers/acts maternally toward x2; [not necessarily biological]

Exercise 2

Express the following qualities in Lojban, using ce'u explicitly in all cases.

  1. Gratitude

  2. Similarity to Arnold Schwarzenegger

  3. Motherhood

  4. Having a mother

  5. My similarity to Arnold Schwarzenegger

  6. Being a place where people get anxious; creepiness, (one interpretation of) hauntedness (Hint: Use sumti tcita.)