Lojban Cheatsheet

This is a little summary of Lojban grammar, covering just basic concepts.

Pronunciation

The most important differences to English pronunciation:

• Vowels are pronounced more or less like in Spanish.
• The apostrophe ' is pronounced like h in "hat" (voiceless glottal fricative).
• The letter c is pronounced like sh in "ship" (unvoiced postalveolar fricative).
• The letter j is pronounced like voiced s in "vision" (voiced postalveolar fricative).
• The letter v is pronounced like v in "valve" (voiced labiodental fricative).
• The letter x is pronounced like ch in "Bach" (unvoiced velar fricative).
• The letter y is pronounced like a in "America" (schwa).
• The letter z is pronounced like z in "zoo" (voiced alveolar fricative).

Definitions

When dealing with Lojban, it's basically unavoidable to use Lojban's own terminology instead of standard linguistic terms. So here are some nonofficial definitions for common concepts.

Parts of speech

Lojban words (jbovla) can be sorted into three groups:

• brivla: words specifying a relation, similar to verbs:
  • gismu: Lojban's ~1,400 root words (CVCCV/CCVCV + 1-4 extra rafsi [short forms] each)
  • lujvo: compound words made from rafsi
  • fu'ivla: loanwords in one of four stages of assimilation (formerly called le'avla)
• cmavo: Lojban's ~1,000 particles (CV/CV'V/.V/.VV) organized in ~100 classes (celma'o), e.g.:
  • gadri: articles (LA, LE series)
  • sumti cmavo ("pro-sumti") are pronouns that act as sumti (KOhA series: mi, ti etc.)
  • bridi cmavo ("pro-bridi") are anaphoric pronouns referring to bridi (GOhA series)
  • namcu: numbers (PA and MOI series for cardinals and ordinals, respectively)
  • lerfu: letterals (BY, LAU, TEI, and XI series)
  • modals (BAI series)
  • vocatives (COI series)
  • attitudinals are interjections, indicating emotion (UI series)
  • evidentials for evidentiality (UI series, too; so sometimes called attitudinals as well)
  • …
• cmevla: proper names (a name is called cmene in Lojban)

Syntax

Lojban is heavily based on predicate logic, so it has its own terms for syntax:

• bridi: a predicate consisting of a relation and optional arguments, basically a sentence
• selbri: the relation of a predicate – a verb phrase (can be a brivla or tanru)
• sumti: argument of a predicate – a noun phrase (description/pro-sumti/name/quote/number)
• tanru: a verb compound made of two or more brivla (and possibly cmavo)
• mekso: a mathematical expression (MEX)

Grammar

A few basic syntax rules:

• The basic word order of a bridi is sumti-selbri-sumti.
  • The particle cu comes before the selbri, if we can't identify the selbri from context.
  • Default place order can be swapped with se (x1/2), te (x1/3), ve (x1/4), and xe (x1/5).
  • Or, places can be tagged explicitly with fa (x1), fe (x2), fi (x3), fo (x4), fu (x5).
  • Places can be skipped with zo'e. (Starting a sentence with a selbri implies x1 = zo'e.)
• Before a bridi, you put .i to properly separate it from the previous sentence.
  • There are more connectors, like .ije, which joins sentences cumulatively ("and").
  • New paragraphs can be indicated with ni'o, which replaces ".i".
  • The contrastive conjunction ku'i still requires ".i" at the beginning of a sentence.
• Quantifiers appear before sumti.
  • With descriptions, they can also appear after the article to specify set cardinality.
• "pro-sumti" can stand alone, but sumti made from other words require a preceeding article:
  • le and lo: definite and indefinite descriptions, terminated by ku, if needed.
    • lei and loi: plural versions of le and lo
  • la: names, also terminated by ku, if needed.
  • lu, le'u, zo, zoi: quotations
  • li: numbers
• "pro-sumti" can also be used as possessive pronouns after articles.
• Nominalizers (abstractions) nu (event), za'i (state) etc. must be terminated by kei if needed.
• Tense particles come before the selbri, but are optional:
  • pu: past
  • ca: present
  • ba: future
• Locative particles act like tense particles and come before the selbri, too (after tense).
• Aspect particles go between tense particles and selbri.
• Verb compounds (tanru) are left-grouping:
  • "fenki logji bangu" is "insanely logical language", not "insane logical language".
  • In general, tanru are preferred over lujvo. (?)
• Attitudinal compounds are additive, and the more important ones should come first.
  • The particle pe'a marks the previous word as figurative speech.
• Relative clauses start with poi, contain the relative pronoun ke'a and end with ku'o.
• Information not covered by existing places can be added to a relation with Lojban "modals":
  • Modals are not to be confused with what linguists call "modals"!
  • Modals start with fi'o and are terminated with fe'u, if needed.
  • The are shorter modal tags from semla'o BAI which can be used without "fi'o".

Further reading

• The Lojban Reference Grammar (one page per chapter)
• The Lojban Reference Grammar (one page per section)
• Diagrammed Summary of Lojban Grammar
• vlasisku (Lojban dictionary search engine)
• Lojban parser (graphical output)
• Summary of Attitudinals
• Robert Baruch's introductory series on YouTube

---

Copyright © 2021 by Thomas Heller [ˈtoːmas ˈhɛlɐ]