Can words be "derived" on the fly?

Question

I recently starting learning Esperanto, and I am enjoying it so far. (I am enrolled in the Duolingo course.) The Wikipedia page for Esperanto says:

Esperanto words are mostly derived by stringing together roots, grammatical endings, and at times prefixes and suffixes. This process is regular so that people can create new words as they speak and be understood.

Is there some way to derive word from suffixes on the fly, as in, in a conversation? This quote says that the process for deriving words is regular - how is it done? If I don't know a word, can I derive it simply from a list of roots and endings? Does such a list exist? Thanks for any feedback!

Answer 1

Is there some way to derive word from suffixes on the fly, as in, in a conversation? [...] If I don't know a word, can I derive it simply from a list of roots and endings?

Yes and no, but mostly yes: If you have a concept in mind and don't know an Esperanto word for it, you can build one with Esperanto's word building system, that will usually be readily understood. Proficient speakers can (and often do) do this on-the-fly.

However

• the word you come up with for the concept will be a word for the concept, not necessarily the word for the concept. There might already by one or more well-established words (whether basic ones with their own root, or ones also built with the word building system) for what you're trying to express and using them might be considered more correct than what you've just built on-the-fly.

• some existing built words traditionally have a meaning that is more specific than what can be determined from their building parts. For example

  • mal·san·ul·ej·o can be analyzed as "place or facility (·ej·o) for people or a person (·ul·) who are the opposite of (mal·) healthy (san·a)" but is quite specifically a hospital and not just any generic facility for sick, ill or injured people.
  • tranĉ·il·o can be analyzed as "tool (·il·o) for cutting (tranĉ·i)", but is means specifically "knife" and not, e.g., "scissors".

  so if you by chance happen to build such a word without knowing it, you might be conveying a meaning more specific or slightly different than you intended.

This quote says that the process for deriving words is regular

It is mostly regular, and probably much more regular than in most national / ethnic languages. But don't take that as it being completely regular also for pre-existing built words. (See examples above.)

how is it done? If I don't know a word, can I derive it simply from a list of roots and endings?

There's roots, prefixes (beginnings), suffixes (endings) and word-type-indicators.

Lernu.net has a whole grammar chapter on the word building system and Duolingo has a "skill" focusing on just that topic. (See here for Duolingo's tipps on the skill.)

You'd usually begin with a root that conveys the base / main meaning of the word. You can attach prefixes before it and suffixes after it (usually removing the word-type indicator before adding a suffix) to alter its meaning. You can combine several such conglomerates to a single word, if appropriate. (Then the last / right-most conglomerate would give the base meaning and the conglomerates prefixed to the left would be modifying it.) Finally, you can attach a new word-type-indicator at the very end.

Let's play that through on the hospital example:

1. The topic here is health. san·a means 'healthy', 'of good health'. Let's start with that. (The ·a marks it as an adjective.)
2. Prefix mal· turns something into it's opposite, so to express 'sick' / 'ill' we can use mal·san·a.
3. The suffix ·ul· indicates a person with that trait. So a healthy person would be a san·ul·o, a sick or ill one a mal·san·ul·o. Note how we removed the ·a and added an ·o, to indicate that these are now nouns.
4. The suffix ·ej· indicates a dedicated place or facility for what comes before it. So the hospital can be expressed as mal·san·ul·ej·o. Note that we've removed the final ·o, added ·ej· and then added back a ·o at the very end, because the whole thing is again a noun.

Sometimes the base meaning doesn't have a root of its own, but there's an affix for it. In those cases, one can (ab)use the affix as a root:

• in·o - a female (i.e., a woman or girl, in specific contexts maybe a female animal)
• ej·o - a place for doing something specific, a facility

Sometimes this is also done when a root exists, to gain some brevity or a different nuance:

• mal·o - an opposite (synonym of kontraŭ·o, a contrary)

Does such a list exist?

Any good grammar book and many good dictionaries will contain a complete list of affixes (prefixes and suffixes) and of word-type-indicators. Comprehensive lists of roots might be harder to come by and there might not be an absolutely complete one, as Esperanto allows the addition of new roots for concepts where it's impossible or too impractical to express them with existing roots and the word building system alone.

In PMEG you can find the lists of affixes here.

Answer 2

Yes, this is quite common in Esperanto and people often invent words with suffixes that have a meaning in the context of a particular conversation.

For example, the word for to eat is manĝi. In an Esperanto congress you often have a manĝejo, which is the place for eating. -ej is a suffix that means “a place related to the root word”, so in this case a place related to eating. That word is already well known, but for example if the congress randomly also had a tennis court, you could happily call that the tenisejo, ie, the place for tennis, and just invent the word on the spot.

Other examples of suffixes could be:

• -em To describe a person that tends to do the action a lot. Like parolema is an adjective for someone that talks a lot.
• -ul Makes a noun for a person that does the action.

These two are often combined, for example parol·em·ul·o is a person that likes to talk a lot. You could make up words with this in the context of a conversation and you would likely be understood, for example you could say something like:

Mi tre ŝatas ŝakludemulojn

That would be something like:

I really like people-who-like-to-play-chess

You can find lists of these affixes on Lernu:

• Suffixes
• Prefixes

I wouldn’t recommend trying to learn them all in one go however. It’s probably best to try to learn them one at a time alongside other vocabulary.

Answer 3

Esperanto has a single root for noun/adjective/verb/adverb and applies an endings for those categories. And then there are several prefixes to precede root and suffixes to add to the root: -ebl = -able, lernebla = learnable. Also prepositions can function as prefix: sen = without, senbarba = beardless. Prefixes and suffixes are often listed separately in dictionaries.

The main difference is that the combining power is entirely regular, appliable to all feasible words, and the meaning exact. And there might be more composed combinations: polic-ist-o = police officer (fixed expression), polic-ist-ar-o = police community.

In effect this combinatory strength will allow expressiveness, active language, which you know is correct. The choice of formulation (sen barbo versus senbarba) is also a factor for language expressiveness.

One anecdote: a young child not knowing the word for chair, seĝo, used sid-il-o = tool/aid to sit. Totally correct and understandable.

Animals is a positive example, but only as national languages have a plethora of names. So -id (child), -in (female), vir- (male), -ej (place).

• ĉevalo = horse, ĉevalino, virĉevalo, ĉevalido, ĉevalejo, ĉevalisto

Rare words in daily life (cervo = deer & cervido etc.) will be easier to use.

There is no risk for (uncharacteristic) monotony by repetitiion, as the root will not be overly repeated and one can often drop the root (ino/ejo).

Answer 4

Let's recap how the vocabulary in Esperanto is formed.

There are prefixes which come in front of the root and there are postfixes, a.k.a. suffixes, which come after it. There is a limited list of the official pre- and postfixes, and most introductionary books list all of them. In addition there are some inofficial pre- and postfixes which are mainly used in certain special fields, like medicine, engineering, mathematics and so on. Already these pre- and postfixes have a certain character, nature, i.e. what kind of words they are meaningfully combined with. For instance you wouldn't combine the prefix bo- : meaning -in-law (for relatives) with a verb.

The biggest group consists of the roots. The first non-single letter, non-name entry in Plena Ilustrita Vortaro (PIV) is abak/o and the last zum/i. Between those are over 25000 entries (that's a number I found for an older edition). Not counting single letter, name and primitive particle (words like "as", "but" which do not flex in any way) entries, the entries in a good dictionary (like PIV) are marked for vortkaraktero : word character. For instance abak/o is fundamentally a noun, zum/i a verb and bel/a an adjective.

While in theory you can combine any prefix, even a couple ones with any root and any postfix or multiple postfixes, the sheer number of all theoritically possible combinations gets rapidly overwhelming. In practice you take max one prefix and max two postfixes, words with more are usually too hard to decipher, i.e. to understand what they mean.

The characters of the pre- and postfixes and the roots limit even more what combinations are meaningfull even in theory. There is a principle of sufficiency and necessity in Esperanto, it says that you should limit piling up pre- and postfixes to the mimimum to express the idea without it being mixed with another one. For instance martelo means hammer and when you use it you can obviously say like haki per martelo : to hit with a hammer, but to have a verb it is enough to say marteli : to hammer. There is no need to add more, since the "obvious" action with it is hammering. Of course that "obviousness" is open to interpretations.