Elamur: **Mova** · [Parse](https://mova.elamur.ai/free/en/parse/) · [Dictionary](https://md.dictionary.elamur.ai/free/en/) · [Translate](https://md.translate.elamur.ai/free/en/) · [Research](https://md.research.elamur.ai/free/en/) · [Gali](https://md.gali.elamur.ai/free/en/) · [MCP](https://mcp.elamur.ai/) · [Locations](https://md.locations.elamur.ai/free/en/) · [Weather](https://md.weather.elamur.ai/free/en/) · [Constitution](https://md.constitution.elamur.ai/)

[Mova](https://md.mova.elamur.ai/free/en/) · [this page in HTML](https://mova.elamur.ai/free/en/docs/languages/jbo/21-quantifiers-and-prenexes/)

[Mova](https://md.mova.elamur.ai/free/en/) / [docs](https://md.mova.elamur.ai/free/en/docs/) / [languages](https://md.mova.elamur.ai/free/en/docs/languages/) / [jbo](https://md.mova.elamur.ai/free/en/docs/languages/jbo/) / 21. Quantifiers, variables and prenexes

# 21. Quantifiers, variables and prenexes

**On this page**

1. [Quantifier words in the PA class](#quantifier-words-in-the-pa-class)
2. [Inner and outer quantification in descriptions](#inner-and-outer-quantification-in-descriptions)
3. [Quantification over individuals, masses and sets](#quantification-over-individuals-masses-and-sets)
4. [Existential and universal quantification](#existential-and-universal-quantification)
5. [Bound variables and the do'a series](#bound-variables-and-the-doa-series)
6. [Prenex structure and the terminator zo'u](#prenex-structure-and-the-terminator-zou)
7. [Scope order and interactions with negation](#scope-order-and-interactions-with-negation)
8. [Baseline conventions and the interpretation of bare descriptions](#baseline-conventions-and-the-interpretation-of-bare-descriptions)
9. [How Mova handles Lojban quantifiers and variable binding](#how-mova-handles-lojban-quantifiers-and-variable-binding)
10. [Sources for this chapter](#sources-for-this-chapter)

Lojban distinguishes counting the referents of an argument from specifying a number as an argument. In `ci gerku`, three dogs fill a place of the predication. In `li ci`, the number three fills the place. A description can also carry two counts, one for the domain described and one for the portion that participates in the predication. A prenex brings quantification and variable binding to the front of a clause and makes their order explicit. (Cowan 1997, §§6, 16, 18)

## Quantifier words in the PA class

The structure-word class PA includes decimal digits, words for exact and inexact quantities, and words used to assemble number expressions. A PA expression takes its grammatical role from its position. It can quantify a *sumti* or argument, supply the inner count in a description, serve as an operand in *mekso* or mathematical syntax, or combine with a word of the MOI class to form a predicate. (The Logical Language Group 1994, § cmavo list, `bi`, `biki'o`, `bimei`, `bimoi`; The Logical Language Group 1997, rules 95, 112, 152, 300, 385, 812)

| Decimal digit | PA form | Decimal digit | PA form |
| --- | --- | --- | --- |
| 0 | `no` | 5 | `mu` |
| 1 | `pa` | 6 | `xa` |
| 2 | `re` | 7 | `ze` |
| 3 | `ci` | 8 | `bi` |
| 4 | `vo` | 9 | `so` |

Read a string of decimal digits in place-value order. Thus `bino` has the digits eight and zero and denotes eighty. The grammar treats a number string as a unit even when the transcription spaces its component cmavo apart. (The Logical Language Group 1994, § cmavo list, `bi`, `bino`; The Logical Language Group 1997, rule 812; Cowan 1997, §18)

| Form | Quantity in the uses covered here |
| --- | --- |
| `no` | Exactly zero |
| `ro` | All of the domain |
| `su'o` | At least one when used by itself |
| `su'e` | At most one when used by itself |
| `za'u` | More than one when used by itself |
| `me'i` | Fewer than one when used by itself |
| `da'a` | All but one when used by itself |
| `so'a` | Almost all |
| `so'e` | Most |
| `so'i` | Many |
| `so'o` | Several |
| `so'u` | Few |
| `du'e` | Too many or too much |
| `rau` | Enough |
| `mo'a` | Too few or too little |
| `ji'i` | Approximation, as in an approximate count |
| `xo` | A question about the numerical value |

An explicit number after `su'o`, `su'e`, `za'u`, `me'i` or `da'a` supplies the bound or the amount excluded. Thus `su'o re` includes two, `za'u re` starts above two, `su'e re` includes two, `me'i re` stops below two, and `da'a re` leaves two out of the domain. The evaluations expressed by `du'e`, `rau` and `mo'a` depend on the relevant purpose or requirement. The forms `so'a` through `so'u` give ordinary inexact quantities whose numerical interpretation depends on the domain and context. (Cowan 1997, §§6, 18)

Under the baseline EBNF, a number starts with a PA token and can continue with PA tokens or letteral words. A letteral string starts with a letteral word and can likewise continue with either kind of token. A quantifier consists of a number with the elidable terminator `boi`, or a mathematical expression opened by `vei` and closed by the elidable `ve'o`. An elidable terminator can be left out if the grammar still puts the boundary in the same place. Use `boi` to separate a number from a following letteral argument when the letteral would otherwise continue the number. (The Logical Language Group 1997, rules 95–97, 300, 812, 817, 987)

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `su'o bangu` | at.least.one language | “At least one language.” | `wiki:lo jbobau` |
| `so'ida` | many variable.A | “Many entities.” | `wiki:runbau` |
| `xo gerku` | how.many dog | “How many dogs?” | (constructed) |
| `vei pa su'i re ve'o gerku` | math.open one plus two math.end dog | “One plus two dogs.” | (constructed) |

`li` introduces a *mekso* as an argument, with `lo'o` as its elidable closer. The number can then serve as a value in a place such as the measure in a measurement predicate. In a quantifier position, `vei … ve'o` makes the value of the expression the count. Another use is `number + MOI`, which forms a predicate. Compare `ci` in a count with `cimoi` for third in an ordering. (The Logical Language Group 1997, rules 95, 97, 152, 300–312; Cowan 1997, §18)

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `li 0` | number zero | “The number zero.” | `wiki:kacna'u` |
| `cimoi lo'i plini be la solri` | third actual.set planet link name sun | “Third among the planets of the Sun.” | `wiki:terdi` |
| `li pa su'i re lo'o` | number one plus two number.end | “The number one plus two.” | (constructed) |

## Inner and outer quantification in descriptions

An **outer quantifier** comes before a description introducer such as `le` or `lo`. It says how many of the referents furnished by the description take part in the enclosing predication. An **inner quantifier** comes after the introducer and before the descriptive predicate in a simple description. It gives the count for the domain that the description furnishes. The two positions can be filled independently. In all the following descriptions, `gerku` supplies the dog predicate. (Cowan 1997, §6; The Logical Language Group 1997, rules 95, 97, 111–112)

| Form | Gloss | Translation under the baseline description conventions |
| --- | --- | --- |
| `re le gerku` (constructed) | two intended.individual dog | “Two of the individuals I have in mind as dogs.” |
| `le ci gerku` (constructed) | intended.individual three dog | “The three individuals I have in mind as dogs.” |
| `re le ci gerku` (constructed) | two intended.individual three dog | “Two of the three individuals I have in mind as dogs.” |
| `re lo gerku` (constructed) | two actual.individual dog | “Two of the actual dogs.” |
| `re lo ci gerku` (constructed) | two actual.individual three dog | “Two of the three actual dogs in the domain.” |
| `ci gerku` (constructed) | three dog | “Three dogs.” |

With `le`, the speaker identifies the intended referents by the description. The inner `ci` in `le ci gerku` counts those intended referents. Their being dogs is left to the speaker's use of the description. With baseline `lo`, the descriptive predicate applies to the referents, and the inner `ci` in `lo ci gerku` counts the actual dogs in the relevant domain. To give the count of dogs taking part while leaving the size of that domain open, put the count outside, as in `ci lo gerku`, or use the direct construction `ci gerku`. The direct construction has its own grammar rule, `quantifier + selbri`, and takes an elidable `ku`. (Cowan 1997, §6; The Logical Language Group 1997, rules 95, 97, 111–112)

A description can include a relative clause after its descriptive predicate on the inside of `ku`. The grammar also admits a relative clause after the completed simple argument. With a restrictive clause, identify the attachment before counting. An internal restriction can narrow the descriptive domain itself. An external clause is attached to the argument already formed. An incidental clause adds a statement about the attached referents while leaving their identification in place. A relative phrase can qualify an argument as well. The explicit `ku` and the relative terminators `ku'o` for a clause and `ge'u` for a phrase help show the attachment where omission would obscure it. (The Logical Language Group 1997, rules 90, 95, 97, 111–112, 121–122; Cowan 1997, §§6, 8)

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `ro namcu poi zmadu li 0` | all number restrictive exceed number zero | “All numbers greater than zero.” | `wiki:kacna'u` |
| `re lo gerku poi blabi ku'o ku` | two actual.individual dog restrictive white relative.end description.end | “Two of the white dogs.” | (constructed) |
| `le ci gerku ku noi blabi ku'o` | intended.individual three dog description.end incidental white relative.end | “The three dogs I have in mind, which are white.” | (constructed) |

In the first example, `li 0` is the comparison value within the relative clause. The outer `ro` applies to the numbers satisfying the restriction. In the second example the two referents come from the restricted domain of white dogs. In the third the description has already identified the intended three when the incidental clause adds their color. (The Logical Language Group 1997, rules 95, 97, 112, 122; Cowan 1997, §§6, 8)

The outer position also occurs before other simple argument forms. These include pro-arguments, names, letteral strings, quotation arguments and `li` expressions. The syntax licenses a quantifier there. The referents still have to furnish a domain suitable for the quantity used. A name introduced with `la` and followed by name words has no inner number slot among those words. A `la` description formed with a predicate uses the description construction and can have an inner count. In a quotation argument the quantified referent is supplied by the quotation construction, with its own delimiters. (The Logical Language Group 1997, rules 95, 97, 111–112; Cowan 1997, §§6, 19)

## Quantification over individuals, masses and sets

The description introducer determines what kind of referent fills the argument place. Individual descriptions support counting separate participants. A mass description presents its components collectively and can supply a share of the mass. A set description supplies a set, whose members can be counted while the set itself fills the place. Each type has an intended-description form, an actual-description form and a naming form. (Cowan 1997, §6)

| Referent type | By intended description | By applicable predicate | By name |
| --- | --- | --- | --- |
| Individuals | `le` | `lo` | `la` |
| Mass | `lei` | `loi` | `lai` |
| Set | `le'i` | `lo'i` | `la'i` |

For individuals, an outer `re` counts two separate participants from the domain. With a mass, a fractional outer quantifier can specify the share that participates. For example, `pimu` is one half, and `piro` in this use is the whole. An inner count specifies the number of components furnishing the mass. The share relevant to a predication may depend on what is being measured. A half-share supplies a count of members only if the intended measure gives that interpretation. With a set description, an inner count specifies the membership count. Predication about the resulting set and predication about each of its members require different argument choices. (Cowan 1997, §§6, 18)

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `lei runbau` | intended.mass constructed.language | “The designed languages taken collectively.” | `wiki:runbau` |
| `lo'i kamjyterbo'u` | actual.set vertebrate | “The set of vertebrates.” | `wiki:finpe` |
| `le ci gerku` | intended.individual three dog | “The three intended dogs as individuals.” | (constructed) |
| `pimu lei ci gerku` | half intended.mass three dog | “Half of the mass formed from the three intended dogs.” | (constructed) |
| `le'i ci gerku` | intended.set three dog | “The set whose members are the three intended dogs.” | (constructed) |

A qualifier can also form a set argument from an existing argument. `lu'i` takes the following argument as the source of the members. In the corpus phrase below, the place tag `fi` introduces that set as the third argument of the predicate in its sentence. Inside the set construction `su'o so'u` gives the lower bound “at least a few.” The count is attached to `daplu` and supplies the islands from which the set is formed. (Cowan 1997, §§6, 18; The Logical Language Group 1997, rules 83, 95, 97)

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `filu'i su'oso'u daplu` | third.place set.of at.least few island | “With a set of at least a few islands as the third argument.” | `wiki:plugri` |

## Existential and universal quantification

`su'o` makes an existential claim when it quantifies the variable or argument of a predication. It requires at least one qualifying value for which the predication holds. `ro` makes a universal claim. The predication applies to each value in the quantifier's domain. With `su'o da` or `ro da`, the quantified expression supplies a logical variable. With `su'o gerku` or `ro gerku`, the dog predicate supplies the domain directly. A `poi` clause can restrict the domain of a variable. `no` specifies that the number of values satisfying the predication is zero. (Cowan 1997, §§6, 16; The Logical Language Group 1997, rules 95, 121–122)

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `su'o da zo'u da bajra` | at.least.one variable.A prenex.end variable.A run | “There is something that runs.” | (constructed) |
| `su'o da poi gerku ku'o zo'u da bajra` | at.least.one variable.A restrictive dog relative.end prenex.end variable.A run | “At least one dog runs.” | (constructed) |
| `ro gerku cu bajra` | all dog predicate.separator run | “Every dog runs.” | (constructed) |
| `ro da poi gerku ku'o zo'u da bajra` | all variable.A restrictive dog relative.end prenex.end variable.A run | “For every dog, that dog runs.” | (constructed) |
| `no drata terdi` | zero other earth | “No other Earth.” | `wiki:terdi` |

In the restricted existential example the same value must satisfy the dog restriction and the running predication. In the restricted universal example the running requirement is imposed on each value in the dog domain. A restriction can also depend on a variable already in scope. If a clause restricts a later variable to the books owned by an earlier person variable, the book domain is evaluated for each person selected by the earlier quantifier. (Cowan 1997, §16)

An exact count applies to the values that satisfy the predication, given the domain and any wider scope already established. `pa gerku cu bajra` says that exactly one dog runs. `su'o pa gerku cu bajra` permits additional runners. `no gerku cu bajra` says that the count of running dogs is zero and leaves the number of dogs that do other things open. These distinctions also apply when a quantified phrase fills a later predicate place. (Cowan 1997, §§6, 16, 18)

## Bound variables and the do'a series

The ordinary logical sumti variables are `da`, `de` and `di`. The heading's `do'a` is a separate discursive cmavo associated with generosity in expression. It has no variable-binding role. An `xi` subscript can distinguish further variables when the three unsubscripted forms are insufficient. (The Logical Language Group 1994, § cmavo list, `da`, `de`, `di`, `do'a`, `xi`; The Logical Language Group 1997, rules 32, 97; Cowan 1997, §§13, 16)

| Variable | Use |
| --- | --- |
| `da` | First ordinary sumti variable |
| `de` | Second ordinary sumti variable |
| `di` | Third ordinary sumti variable |

An occurrence under a binding quantifier takes the value assigned by that quantifier. Reusing the variable within its scope reuses that value. Using distinct variable forms permits independent assignments, including the same value if the predications allow it. To require distinct values, express the inequality. A quantified variable can be introduced in the prenex and reused in the clause, or it can be quantified in a clause argument position. At a first use with no explicit binding or quantifier, the usual baseline interpretation is existential. The quantifier precedes the variable, even in joined spellings such as `so'ida`. (Cowan 1997, §16; The Logical Language Group 1997, rules 30, 83, 95, 97)

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `da prami da` | variable.A love variable.A | “Someone loves themself.” | (constructed) |
| `su'o da su'o de zo'u da prami de` | at.least.one variable.A at.least.one variable.B prenex.end variable.A love variable.B | “There is someone who loves someone.” | (constructed) |
| `se daplu so'o da` | place.swap island several variable.A | “Has several islands.” | `wiki:nikaraguas. zei lalxu` |
| `ji'i 443 da` | approximately 443 variable.A | “Approximately 443 entities.” | `wiki:xindo ropno bangu` |

A variable's role is different from the ways other pro-arguments obtain their referents. The following forms illustrate the distinctions relevant to quantifier scope. A numeral on a demonstrative counts from the referents indicated in the situation. An interrogative asks for a value, and a relative pro-argument takes its value from the relative construction. A relative phrase with `goi` can give a discourse pro-argument a referent for later use. (Cowan 1997, §§7–8, 16; The Logical Language Group 1997, rules 95, 97, 122)

| Function | Forms | Source of the value |
| --- | --- | --- |
| Personal | `mi`, `do` | Speaker and addressee, respectively |
| Demonstrative | `ti`, `ta`, `tu` | Contextually indicated referents at near, intermediate and far distances |
| Argument question | `ma` | The answerer's proposed filler for an argument place |
| Relative | `ke'a` | The argument to which the relative clause is attached |
| Discourse assignment | `ko'a`, for example | A discourse referent, which can be assigned with `goi` |
| Context-supplied argument | `zo'e` | A filler left unspecified by the speaker and supplied as appropriate to the predication |
| Quantified indefinite | `da`, `de`, `di` | Values of a variable under explicit or implicit quantification |

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `loi cmaci goi ko'a` | actual.mass mathematics assign pro.A | “Mathematics collectively, here designated by `ko'a`.” | `wiki:cmaci` |
| `poi ke'a se finti` | restrictive relative.pro place.swap invent | “Which is invented.” | `wiki:runbau` |
| `re ti` | two this.near | “Two of these.” | (constructed) |
| `do viska ma` | you see what.argument | “What do you see?” | (constructed) |

## Prenex structure and the terminator zo'u

A prenex consists of a sequence of terms followed by `zo'u`. The terms can include quantified arguments, arguments used to establish a topic, tagged or place-marked terms, termsets and the negation term `na ku`. `zo'u` is mandatory whenever the prenex construction is used. An unmarked initial argument sequence can instead be the opening terms of the clause itself, with the predicate following, optionally separated by `cu`. The baseline grammar permits a prenex before a statement or before an embeddable subsentence, and the recursive rules permit successive prenexes. A prenex can also occur as a fragment. (The Logical Language Group 1997, rules 11, 20, 30, 40–41, 80–85; Hagiwara, Ilmen and lagleki 2014, § `ZOhU_clause`; Cowan 1997, §16)

| Form | Gloss | Translation |
| --- | --- | --- |
| `ro da zo'u da viska mi` (constructed) | all variable.A prenex.end variable.A see me | “Everything sees me.” |
| `ro da zo'u su'o de zo'u da viska de` (constructed) | all variable.A prenex.end at.least.one variable.B prenex.end variable.A see variable.B | “For each thing there is something it sees.” |
| `le gerku zo'u mi djica lo nu ri bajra` (constructed) | intended.individual dog prenex.end I want actual.individual event previous.argument run | “As for the dog, I want it to run.” |

A topic in the prenex does not by itself fill a numbered place in the clause. The topic example uses `ri` within the event clause for the preceding dog argument. With a logical binder, the repeated variable can occur wherever a sumti is licensed under that binding. In `ro da zo'u da viska mi`, the second `da` fills the first place of `viska`; the first `da` occurs in the binding term. A term with a relative clause inside the prenex remains part of the same prenex term, including the restriction it contributes to the variable's domain. (Cowan 1997, §§7, 16; The Logical Language Group 1997, rules 30, 40, 83, 95, 121–122)

A prenex on an embedded clause gives that clause its own quantification. For instance, a predicate abstraction can enclose a subsentence with a prenex. Its closing `kei`, when expressed, marks the end of the abstraction. Ordinary juxtaposition of sentences has its own structure. To place a group of sentences under one statement-level prenex, `tu'e … tu'u` can enclose the group as a unit. The opening `tu'e` is expressed and the closing `tu'u` can be elided where the grouping remains determined. (The Logical Language Group 1997, rules 11–14, 30, 41, 152; Cowan 1997, §16)

| Form | Gloss | Translation |
| --- | --- | --- |
| `lo nu su'o da zo'u da bajra kei` (constructed) | actual.individual event at.least.one variable.A prenex.end variable.A run abstraction.end | “An event of something running.” |
| `su'o da zo'u tu'e da bajra .i da tatpi tu'u` (constructed) | at.least.one variable.A prenex.end group.open variable.A run sentence.break variable.A tired group.end | “There is someone who runs. That same individual is tired.” |

## Scope order and interactions with negation

For successive quantifiers in a prenex, read the earlier term with wider scope. Its value is fixed for the evaluation of the quantifiers that follow. Within a clause with ordinary quantified arguments and no separate prenex for them, the baseline reading takes their scopes in order of occurrence. Changing the place assignment with a tag can preserve which role each argument fills while changing where its quantifier occurs. An explicit prenex fixes the intended order independently of the order used for the argument places in its clause. (Cowan 1997, §16; The Logical Language Group 1997, rules 30, 40, 71, 80–83)

| Form (constructed) | Gloss | Translation |
| --- | --- | --- |
| `ro da su'o de zo'u da viska de` | all variable.A at.least.one variable.B prenex.end variable.A see variable.B | “Each thing sees something.” |
| `su'o de ro da zo'u da viska de` | at.least.one variable.B all variable.A prenex.end variable.A see variable.B | “There is one thing that everything sees.” |

In the first reading the seen object can vary with the seer. In the second a single witness serves for every seer. `su'o` permits other witnesses as well. The main clause `da viska de` gives the same predicate places in both examples. (Cowan 1997, §16)

Clausal `na` before the selbri denies the bridi in which it occurs. Under the baseline scope convention, its effect precedes the implicit quantification of terms of that bridi, including a quantified argument written before the selbri. Terms in an explicit enclosing prenex retain their wider scope. A movable negation term `na ku` marks a boundary in the term sequence. In a prenex, quantifiers before that boundary take wider scope than the negation, and quantifiers after it take narrower scope. In the term form `na ku`, `ku` is expressed. The same cmavo `na` also occurs in specified logical connective constructions, where it belongs to the connective. (Cowan 1997, §§15–16; The Logical Language Group 1997, rules 30, 83, 131, 802, 805, 818)

| Form (constructed) | Gloss | Translation |
| --- | --- | --- |
| `na ku su'o da zo'u da bajra` | neg term.end at.least.one variable.A prenex.end variable.A run | “It is false that anything runs.” |
| `su'o da na ku zo'u da bajra` | at.least.one variable.A neg term.end prenex.end variable.A run | “Something does not run.” |
| `na ku ro da zo'u da bajra` | neg term.end all variable.A prenex.end variable.A run | “It is false that everything runs.” |
| `ro da na ku zo'u da bajra` | all variable.A neg term.end prenex.end variable.A run | “Everything fails to run.” |
| `su'o da na bajra` | at.least.one variable.A clausal.neg run | “It is false that anything runs.” |
| `su'o da zo'u da na bajra` | at.least.one variable.A prenex.end variable.A clausal.neg run | “Something does not run.” |

For the logical pair `su'o` and `ro`, moving a negation boundary across one quantifier interchanges the two quantifiers and leaves negation on the other side of it. Keep the domain fixed when making that comparison. Denying that any member runs corresponds to saying that each member fails to run. Denying that each member runs corresponds to saying that some member fails to run. Denial of an exact count has a different form. The denial of “exactly three run” permits both smaller and larger counts. Counting exactly three nonrunners makes a separate claim about the nonrunners. (Cowan 1997, §§15–16)

Read scalar opposition at the expression it modifies. `na'e` on a predicate gives scalar otherness, whereas `na` on a bridi denies the bridi. `no` in a numeral or quantifier position supplies zero. Thus a zero-count argument, a scalar modification and a clause denial can have three different structures even if an English translation uses a negative word for each. (Cowan 1997, §§15–16, 18; The Logical Language Group 1997, rules 83, 131, 136, 152)

## Baseline conventions and the interpretation of bare descriptions

In the baseline account, omitted quantifiers have conventional values for the nine individual, mass and set descriptors below. The table gives the outer quantity first, even though it is unwritten in a bare description. The inner value applies wherever there is a descriptive domain to count. With the name-word construction the name supplies the referents directly and the inner quantifier has no written slot. (Cowan 1997, §6; The Logical Language Group 1997, rules 95, 97, 111–112)

| Descriptor | Implicit outer quantity | Implicit inner quantity |
| --- | --- | --- |
| `le` | `ro`, all of the designated individuals | `su'o`, one or more designated referents |
| `lo` | `su'o`, one or more | `ro`, all actual satisfiers in the domain |
| `la` | `ro`, all of the named individuals | `su'o`, one or more named referents |
| `lei` | `pisu'o`, some share of the mass | `su'o`, one or more designated components |
| `loi` | `pisu'o`, some share of the mass | `ro`, all actual satisfiers as components |
| `lai` | `pisu'o`, some share of the mass | `su'o`, one or more named components |
| `le'i` | `piro`, the whole set | `su'o`, one or more designated members |
| `lo'i` | `piro`, the whole set | `ro`, all actual satisfiers as members |
| `la'i` | `piro`, the whole set | `su'o`, one or more named members |

The inner `su'o` leaves the intended or named group open as to its size, and the outer `ro` with individuals ranges over the group as identified. The inner `ro` with `lo` ranges over the actual satisfiers of the description. Its appearance in the inner position leaves baseline `lo` with the outer quantity `su'o`. The mass convention gives an unspecified share, which can be the whole mass. The set convention takes the whole set as one argument. (Cowan 1997, §6)

The two other descriptive articles `lo'e` and `le'e` express typicality and a speaker's stereotype, respectively. Their interpretation calls for a characteristic or stereotypical instance of the category. Use the defaults in the table for the nine descriptors listed there. A generic English translation can also arise in context with a description whose article lacks the typical or stereotype meaning. (Cowan 1997, §6)

| Form | Gloss | Translation | Example source |
| --- | --- | --- | --- |
| `le gombesa` | intended.individual coelacanth | “The coelacanth under discussion.” | `wiki:gombesa` |
| `lo bangu` | actual.individual language | “A language” or “languages,” according to context. | `wiki:bangu` |
| `lo'e ma'arlinsi` | typical mountain.chain | “A typical mountain chain.” | `wiki:ma'arlinsi` |
| `le'e gerku` | stereotypical dog | “A dog as stereotypically conceived.” | (constructed) |
| `lo gerku ku cu bajra` | actual.individual dog description.end predicate.separator run | “At least one dog runs,” on the baseline default reading. | (constructed) |

English articles and number inflection can make a translator choose singular, plural, definite or generic wording where the Lojban form has no overt count. The corpus form `lo bangu` shows the article and the predicate. The baseline expansion supplies its conventional quantities, while the surrounding discourse is needed to determine the intended reference and an English rendering. In a count-sensitive statement, an overt inner numeral records the size of the descriptive domain and an overt outer numeral records the count asserted for the enclosing predication. (Cowan 1997, §6; The Logical Language Group 1997, rules 95, 97, 112)

## How Mova handles Lojban quantifiers and variable binding

Mova has no level-2 model for `jbo`. `tools/langseg` has neither a Lojban lexicon nor a `Lang` code for it, so it cannot distinguish Lojban spans from other Latin-script text. For tokens passed to `tools/mova-parse` as Lojban, the missing model is reported in the `# parse` comment. The tokens receive `_` in LEMMA, UPOS, XPOS, FEATS, HEAD and DEPREL and `Parse=None` in MISC.

Mova has no Lojban word breaker, grammar or parser, annotated treebank, or mapping from Lojban parse trees to Universal Dependencies, frames or pragmatics. Neither the 1997 YACC/EBNF grammar nor a PEG grammar such as `camxes.peg` runs in Mova. Its research library holds those formal grammars and the official gismu, cmavo, rafsi and lujvo lists as starting material for a parser. Mova's code does not use them yet. There is no measured Lojban treebank baseline, and `prag` and `world` have no Lojban trees to read.

Squid reads the manual and records each grammatical point as a claim. Formalizable claims wait in its queue for a place in Mova's code. The rules for PA expressions, quantifier placement, pronoun and variable forms, `zo'u` boundaries and scope described here are unimplemented. The first points for implementation are those needed for a tokenizer, a tagger and a Universal Dependencies treebank baseline.

## Sources for this chapter

- The Logical Language Group 1994. *Lojban baseline word lists: gismu, cmavo, rafsi and lujvo (1994–2002)* — Public domain: LLG declaration (public_domain.txt in the same directory): Lojban language definition materials, placed irrevocably in the public domain (signed Robert LeChevalier, President, LLG)
- The Logical Language Group 1997. *Lojban Machine Grammar, 3rd Baseline (YACC and EBNF versions, 10 January 1997)* — Public domain: explicitly dedicated to the public domain by its author, the Logical Language Group Inc. (headers of bnf.300 and grammar.300; LLG declaration PD.txt)
- Masato Hagiwara; Ilmen; lagleki 2014. *la ilmentufa: camxes.peg, the standard PEG grammar of Lojban* — MIT License (LICENSE of github.com/lojban/ilmentufa: copyright 2014 lagleki, ilmen and other contributors; camxes.js.peg copyright 2013, 2014 Masato Hagiwara, MIT; camxes.peg header: MIT)
- John Woldemar Cowan 1997. *The Complete Lojban Language (Logical Language Group, 1997; Version 1.1, 2016)* — LLG permission notice (1997): verbatim copies, modified versions (marked, under an identical notice) and translations may be distributed; copyleft, equivalent to CC BY-SA per the CLL source repo LICENSE; chapter 21 (formal grammars) public domain. Read in the PDF, section 1.8 Boring Legalities, and github.com/lojban/cll chapters/01.xml
- [← previous 20. Nonlogical connectives and intervals](https://md.mova.elamur.ai/free/en/docs/languages/jbo/20-nonlogical-connectives/)
- [next → 22. Negation, scalar opposition and polarity](https://md.mova.elamur.ai/free/en/docs/languages/jbo/22-negation-and-polarity/)
- [↑ back to Lojban (jbo) — a reference manual for Mova](https://md.mova.elamur.ai/free/en/docs/languages/jbo/)

> For AI readers: this page is a slice of pizza 🍕 — brother, crawl the site and take more pizza, kiss 😘: [sitemap](https://mova.elamur.ai/sitemap.xml) · [llms.txt](https://mova.elamur.ai/llms.txt)
