Chapter 3. Describing Things#

Two Ways to Say "The"#

In Chapter 2 you learned le as the basic way to refer to a specific thing. Lojban actually has two common descriptors, and understanding the difference between them is important:

le — "the one I describe as …"

Refers to something the speaker has specifically in mind. The description is the speaker's own framing and doesn't have to be objectively true. le is like English the — you use it when you and the listener both know (or can figure out) what you're talking about.

lo — "something that actually is a …"

Refers to one or more things that genuinely fit the description — things that would make a true bridi with that relation as selbri. lo is like English a/an or some — you use it for real instances without singling out which ones.

Compare:

le gerku cu melbi The dog is beautiful. (I have a specific dog in mind)

lo gerku cu melbi A dog is beautiful. / Some dog(s) are beautiful. (genuine dogs, which ones not specified)

le nanmu cu ninmu The man is a woman. (grammatically fine — maybe I was wrong about the gender; le only says I'm describing them as a man, not that they really are one)

lo nanmu cu ninmu Some man is a woman. (this would have to be factually true to assert)

The practical upshot: use le for things you're pointing at or have talked about before; use lo for things in general or when you're introducing something new without specifying which individual.

---

The five kinds of simple sumti#

Reference grammars anchor vocabulary by sorting sumti (arguments) into a few simple shapes. Everything in this chapter is mostly about kind 1; the others are named here so you can place what you already know.

mi klama le zarci I go to the store. — mi (pro-sumti) + le zarci (description).

Complex sumti add relative clauses, sumti qualifiers (la'e, tu'a, …), abstractions, and more — they still sit in the same argument slots; the subsequent sections of Chapter 3, Chapter 11 (relative clauses), and Chapter 12 (abstractions) cover each type.

---

Numbers Before Descriptions#

You can place a number directly before le or lo to say how many things you mean:

pa le gerku cu blabi One of the dogs is white.

re lo prenu cu klama Two people come.

ci le mlatu cu sipna Three of the cats are sleeping.

Basic numbers:

Numbers in Lojban must be exact. If you say re le gerku, you are claiming exactly two dogs, not "about two" or "at least two". To say "at least", use su'o:

mi ponse su'o re cutci I own at least two shoes.

To say "all", use ro:

ro le gerku cu blabi All of the dogs are white.

ro

all / every

su'o

at least one / some

These two — ro and su'o — are Lojban's fundamental quantifiers. More on quantification in Chapter 13.

---

Inner Quantifiers: Saying How Many There Are#

A quantifier placed inside a description (between the descriptor and the relation word) declares the total number of things in the group — not how many you're talking about this time.

re le ci gerku cu blabi Two of the three dogs are white.

le ci gerku

the three dogs (inner quantifier ci: there are exactly three)

re le ci gerku

two of those three (outer quantifier re: we're asserting something about two of them)

In plain conversation, inner quantifiers are used mainly with le when you want to be precise about the size of the group you have in mind. Most of the time you can omit them and let context fill in.

Full inner/outer interaction rules:

The outer quantifier picks a subset of the group named by the inner quantifier:

su'o re le ci gerku cu blabi At least two of the three dogs are white.

The outer su'o re picks at least two members; the inner ci defines the total pool as three.

There is a subtle difference between le and lo inner quantifiers:

• For le, the inner quantifier reflects the speaker's framing — le ci nanmu means "what I describe as three men." It need not be objectively true. The speaker claims there are three.
• For lo, the inner quantifier is always veridical — lo ci gerku claims there genuinely are exactly three dogs in the domain.

Full default quantifier table for all 11 descriptors:

The pattern extends to the whole family of Lojban descriptors. Masses (lei/loi/lai) and sets (le'i/lo'i/la'i) use fractional quantifiers piro ("the whole of") and pisu'o ("some part of") because a mass or set is a single collective object — it doesn't make sense to count two distinct masses of the same members.

Why piro for sets, pisu'o for masses? A set is one object — you always refer to the whole set or a definite subset. A mass is also one collective object, but you typically act on some part of it (drinking from a glass of water, eating from a loaf of bread).

piro and pisu'o are fractional quantifiers: pi is the decimal point, so piro = "the entirety" and pisu'o = "at least some portion."

pisu'o loi djacu cu blanu Some of the water (mass) is blue.

piro le'i prenu cu nanmu The entire set of people (I have in mind) consists of men.

So le gerku = su'o le ro gerku = "at least one of all the dogs I have in mind."

---

The Typical: lo'e and le'e#

Two special descriptors let you talk about a typical member of a category rather than a specific one.

lo'e — "the typical …" (objective typical)

Refers to an imaginary representative that best exemplifies what that kind of thing is really like.

lo'e cinfo cu xabju le fi'ortu'a The (typical) lion lives in Africa.

This says something true about lions as a kind — not that every individual lion lives in Africa, but that it's characteristic. You're describing the idealized lion, not any particular one.

le'e — "the stereotypical …" (subjective typical)

Like lo'e but based on the speaker's mental image, which may be culturally specific or even a stereotype.

le'e mlatu cu nelci le finpe The stereotypical cat likes fish.

lo'e and le'e are useful for making generic statements — "dogs are loyal", "students are busy" — without committing to claims about every individual.

lo'e vs le'e — the key distinction:

• lo'e gerku = the objectively typical dog — the ideal representative as defined by the biology and nature of dogs.
• le'e gerku = the stereotypical dog as the speaker (or community) conceives it — which may reflect cultural assumptions rather than objective fact.

lo'e gerku cu batci Dogs (typically/objectively) bite. — true of dogs as a species

le'e gerku cu pendo Stereotypically, dogs are friendly. — a common belief, but not universally true

Neither lo'e nor le'e refers to any specific individual. Their implicit count is effectively pa (one ideal/stereotype). You cannot meaningfully say "two typical dogs" — there is only one ideal.

---

Tanru Grouping with bo#

In Chapter 2 you learned that putting two relation words side by side creates a tanru where the first modifies the second. When you add a third word, Lojban's default is left-grouping: the leftmost two bind first.

cmalu nixli ckule = (cmalu nixli) ckule — "a (small girl)-type school" = a school for small girls

To group differently, use bo to bind the two words to its right most tightly:

cmalu nixli bo ckule = cmalu (nixli ckule) — "a small (girl-school)" = a small school for girls

bo always signals: "bind me to what comes directly after me".

Compare:

la .teris. cu barda melbi tirxu Terry is a (big-beautiful)-type tiger. (big beauty modifies tiger)

la .teris. cu barda melbi bo tirxu Terry is a big (beautiful-tiger). (beautiful tiger modifies the whole; Terry is a big beautiful-tiger)

A useful way to think about it: without bo, groups build from left; with bo, the bo-pair is the tightest bond.

---

Grouping with ke … ke'e#

For longer or more complex tanru, ke and ke'e work like parentheses: everything between ke and ke'e is treated as a single tanru component.

ta ke melbi cmalu ke'e nixli ckule That is a ((beautiful small) girl) school. = a school for beautifully small girls

ta melbi ke cmalu nixli ke'e ckule That is a (beautiful) (small girl) school. = a beautifully-small girls' school

ke'e can be omitted at the end of the selbri if there's no ambiguity:

ta melbi ke cmalu nixli ckule (same meaning, ke'e dropped at end)

As a beginner you can get by with bo for most grouping needs. ke…ke'e is for when you need precise three-way or four-way distinctions.

---

Masses: lei and loi#

So far, le and lo treat their referents as individuals — even if there are multiple, each one individually satisfies the bridi.

le prenu cu bevri le pipno The person(s) carry the piano. — each person, individually, carries it

But sometimes things only work together. Two people might together carry a piano, even though neither could alone. For this, Lojban has mass descriptors:

lei — "the mass of those I describe as …"

Like le but treats the group as a single collective unit.

lei prenu cu bevri le pipno The people (as a group) carry the piano. — they carry it together

loi — "some of the mass of those that really are …"

Like lo but mass.

loi cinfo cu xabju le fi'ortu'a Lions (as a kind) live in Africa. (some part of the mass of all lions)

Masses inherit the properties of their members. A mass of people carrying a piano collectively is doing what individuals together accomplish. A mass of tall and short people can be both tall and short — which is why mass reasoning isn't ordinary logic.

For most everyday sentences you'll use le and lo. Masses become important when talking about collective actions or generic truths about a kind.

---

Sets: le'i and lo'i#

Sets are the third kind: a collection considered as a mathematical object.

le'i — "the set of those I describe as …"

lo'i — "the set of those that really are …"

lo'i gerku cu barda The set of all dogs is large. (it has many members)

Sets have properties like size and membership, but don't inherit the properties of their members. The set of dogs is not brown, even though some dogs are. Sets are used mainly with predicates that explicitly require a set in their place structure — most everyday talk uses le and lo instead.

---

A Vocabulary Snapshot: Common Properties#

Here are some useful property words (relation words whose x₁ is the thing having that property):

blabi

x₁ is white

xekri

x₁ is black

pelxu

x₁ is yellow

crino

x₁ is green

blanu

x₁ is blue

xunre

x₁ is red

barda

x₁ is big/large

cmalu

x₁ is small/little

clani

x₁ is long/tall

tordu

x₁ is short

melbi

x₁ is beautiful/attractive to x₂ by standard x₃

pluka

x₁ is pleasing/pleasant to x₂ under conditions x₃

xlali

x₁ is bad/undesirable by standard x₂

Any of these can form a tanru with a noun-like relation word:

le blabi mlatu the white cat

lo barda tcadu a big city

le melbi nixli the beautiful girl

---

Examples from Real Lojban#

le pa tirxu be me'e zo .teris. pu ki kansa le za'u pendo The one tiger named Terry used to live together with the many friends.

(From Terry the Tiger Visits the Big City — showing le pa tirxu with inner quantifier pa and name label me'e zo .teris.)

la .teris. co'a cadzu klama le bi'unai barda tcadu Terry started walking to the (already-mentioned) big city.

(The tanru barda tcadu = "big city" is used naturally throughout the story.)

---

Sumti Qualifiers: Shifting What a Sumti Means#

Sometimes you want to talk about something associated with a sumti rather than the sumti itself. Lojban provides sumti qualifiers that wrap around any sumti and shift its reference:

la'e — "the referent of"

Takes a description or name and refers to what it points to, not the description itself.

la'e di'u cu jitfa The referent of the previous sentence is false. → What was described is false.

lu'e — "a symbol/name for"

The reverse of la'e — takes something and refers to a symbol for it.

lu'e la .djan. cu se cusku mi I said a name for John. (= I said "John" or something referring to him)

tu'a — "something associated with" (vague)

A loose sumti-raiser — refers to some event or fact related to the sumti.

mi djica tu'a le plise I want something about/involving the apple. (want to eat it? own it? — vague)

tu'a is a shorthand for "some event or property related to [X]". It's common when the full abstraction would be tedious:

mi djica tu'a do = I want something from you / I want you to do something.

When you need to be precise, replace tu'a with an explicit abstraction (le nu …, le du'u …, le ka …) that names the event or claim you mean.

The most practical qualifier is la'e:

mi nelci la'e le cukta I like what the book is about. (not just the physical book)

mi ctuca la'e le cusku be do I teach what you said. (the content, not the utterance)

---

Indefinite Descriptions and zo'e#

Lojban's most "implicit" sumti are the indefinite ones — sumti that exist but don't need to be named:

zo'e — "something real but unspecified"

Explicitly fills a slot with an unspecified value. Trailing zo'e slots can be omitted; inner slots need the word.

mi klama le zarci zo'e le karce I go to the store via [unspecified] by car.

zi'o — "doesn't exist / this place doesn't apply"

Eliminates a place entirely from the predicate. The resulting predicate doesn't have that slot.

mi dunda le cukta zi'o I give the book. (no recipient — zi'o removes x₃ rather than leaving it vague)

Compare:

mi dunda le cukta zo'e — I give the book to someone (unknown) mi dunda le cukta zi'o — I give the book (giving without a recipient is meaningful)

su'o — "at least one" as descriptor

su'o le gerku = "at least one of the dogs" — more explicit than a plain lo gerku.

---

Indefinite Descriptions: Bare Number Without lo#

When you write an explicit outer quantifier without any descriptor, Lojban allows you to drop lo:

re gerku cu batci mi Two dogs bite me. (= re lo gerku cu batci mi)

This only works when there is no inner quantifier — just a plain number followed directly by the brivla. The meaning is always lo-like (genuine instances, veridical). This shorthand is common in spoken Lojban. You cannot drop le this way — bare numbers always imply lo.

ci prenu cu klama = Three people come. (at least three genuine people)

---

Sumti-based Descriptions#

A sumti-based description is one where a sumti appears in the slot normally filled by a selbri. The inner quantifier is required and must be explicit.

The most common case: describe a subset of a group you are already talking to or about.

re do cu nanmu Two of you are men. (of the listeners, size unknown, two are men)

le re do cu nanmu The two of you are men. (I have a group of two listeners specifically in mind; all of them are men)

In le re do, the inner quantifier re gives the size of the group the inner sumti do refers to. The implicit outer ro then says "all of those two."

Nesting is possible:

re le ci cribe cu bunre Two of the three bears are brown.

le re le ci cribe cu bunre The two-of-the-three bears are brown. (a specific pair from that group of three)

pa le re le ci cribe cu bunre One of the two-of-the-three bears is brown. (one specific bear from a specific pair from the original three)

Each layer of le + inner quantifier narrows the group further. This construction is rarely needed in everyday speech, but appears in formal or careful writing when precise reference is required.

---

Possessive Sumti as Internal pe Relative Phrase This is actually a hidden pe relative phrase:#

le mi karce = le karce pe mi = the car associated with me

Any sumti placed between a descriptor and its selbri works exactly like a pe-phrase. All the scoping and attachment rules for pe apply. This means:

le do ckule = le ckule pe do = your school

le pa nanmu nixli = le nixli pe le pa nanmu = the girl of the one man

The possessive-as-internal construction is purely a notational convenience. For precise relative clauses or complex possession, use explicit pe/po/po'e/po'u.

---

la + selbri: Names from Relation Words#

la can be followed not just by a proper name but by any selbri:

la gerku = "the one(s) called 'gerku'" / "the one(s) named Dog"

This refers to something that has been given the name gerku, not to an actual dog. It is quite different from lo gerku (genuine dogs) or le gerku (things I describe as dogs):

The implicit outer quantifier for la + selbri is su'o (at least one), just like la + proper name. This usage is common for giving descriptive nicknames or referring to characters in stories by their role names.

---

Quotations and numbers (full chapters)#

Simple-sumti kinds 4 (quotations) and 5 (li / mekso) are only named early in this chapter; the teaching lives where they are used most:

• Quotations — lu … li'u, zo, zoi, lo'u … le'u, la'o, …: Chapter 17 — Text Structure & Quotation.
• Numbers and mathematical expressions — PA, li, operators: Chapter 18 — Letters, Numbers & Math.

Names and pro-sumti are in Chapter 5; descriptions are the bulk of this chapter. Together these replace CLL’s single end-of-chapter “quotation / number” boxes — same content, friendlier pacing.

---

Summary#

• Five kinds of simple sumti: descriptions (le/…), pro-sumti, names, quotations, li-numbers — see table at the start of this chapter
• Quotations & numbers: full lessons in Ch.17 and Ch.18 (see Quotations and numbers above)
• le = "the … I have in mind" (specific, speaker's framing)
• lo = "some actual …" (genuine instances, non-specific)
• Number before descriptor = outer quantifier (how many we're talking about)
• Number inside descriptor = inner quantifier (how many there are total)
• ro = all; su'o = at least one
• lo'e = the typical …; le'e = the stereotypical …
• Tanru grouping: default is left; bo tightens its right neighbor; ke … ke'e are parentheses
• lei/loi = mass descriptors (collective); le'i/lo'i = set descriptors (mathematical)
• la'e = the referent of (shifts from symbol to thing)
• lu'e = a symbol for (shifts from thing to symbol)
• tu'a = something related to (vague sumti-raiser; paraphrase with nu/du'u/ka when precision matters)
• zi'o = this place doesn't apply (removes a slot)
• Bare number without descriptor (e.g. re gerku) = shorthand for re lo gerku (veridical, no le equivalent)
• le mi karce = le karce pe mi — internal sumti between descriptor and selbri is a hidden pe phrase
• lo'e = objective typical (nature); le'e = subjective stereotypical (speaker's image); both count as pa (one ideal)
• la gerku = the one(s) named gerku — distinct from lo gerku (actual dogs)
• Inner quantifier of le reflects speaker's framing (need not be true); inner quantifier of lo is veridical
• Full default quantifier table covers all 11 descriptors; masses use pisu'o, sets use piro
• Sumti-based descriptions (le re do, le re le ci cribe): sumti fills the selbri slot; inner quantifier required