Yes of course, advokato. One can be strangely blind to typing errors.

There is at least one fixed expression, in which the parts of the word are ignored in favor of a given meaning. We've already talked about it. This is the case where

1: arbo = tree
2: -aro = collection of
3: arbaro != collection of trees
(where != is the symbol for "is not equal to")

It means forest/woods, and that is it. Yes, a forest IS a collection of trees, but there are also collections of trees that are not forests.

This certainly follows no established rule. If I wish to refer to a non-foresty collection of trees, I cannot use the word whose parts mean "collection of trees". So I have to use something else, instead. This is a case of usage trumping logic.

If I recall correctly, you (sudanglo) were in favor of that.

Riot, do you really feel that the limitation of arbaro to a wood (or forest) or lernejo to a school falls in to the same category as the many usage-determined meanings of the turns of phrase of the natural languages.

With a basic knowledge of Esperanto, it is no great leap to predict that arbaro means a wood.

Such words are usually more specific than any aro da. The same goes for ejo-words, which are usually more tightly circumscribed in their application that any old place where the activity takes place.

Could you, on the other hand, knowing just some English predict 'fully-fledged' or that we say 'piss poor' and not 'wee poor' or 'urine poor'

Hypoboreus, as I understand it, a pidgin is a mixture of languages with perhaps some inherent contradictions, whereas a creole is a regularised form with its own consistent rules, that may emerge with a second or subsequent generation of speakers.

Esperanto can't be counted as a pidgin on two grounds.

1. It does not have two parent languages that were mixed up.

2. It has its own consistent rules different to those of other languages.

And it started life with those rules. It didn't have to wait for a second or third generation of speakers for those rules to emerge.

sudanglo:Riot, do you really feel that the limitation of arbaro to a wood (or forest) or lernejo to a school falls in to the same category as the many usage-determined meanings of the turns of phrase of the natural languages.

Irrelevant. Esperanto is not a natural language, as you perfectly well know. It doesn't matter how illogical one or more natural languages may be; we are talking about Esperanto rather than those others.

"Arbaro" is, of course, only one example of words whose meanings are independent of their parts*. As you have pointed out, there are a whole bunch of words whose meanings are significantly more circumscribed than someone who simply read the rules of Esperanto would expect.

In another thread, you argued that Esperanto should be rule-based in a way that natural languages cannot be; that a word should mean what its parts suggest it means. To me, that is logical. Regarding my example, however, you seem to feel that for some words it's perfectly fine to ignore their part-derived meanings in favor of usage-determined ones.

I'm having trouble figuring out a consistent base for your arguments, at least in this regard.

Does a word mean what its constituent parts indicate it means, or not? If not, what possible rules could we come up with to describe these words that don't follow established Esperanto patterns while still mimicking their forms? Do we call them "exceptions" and tell students that they'll "just have to memorize" their special meanings? Yuck. I'm actually surprised that you, as someone who generally seems in favor of solid rules and descriptive models, are not bothered by this example.

I'm not against giving words additional meanings to the ones their parts suggest. But I do have a (small) problem with denying meanings to a word that its parts clearly indicate.

----------
* In the sense that the parts indicate a meaning that has actually been removed from a word.

that a word should mean what its parts suggest it means.

Yes, I think that is a principle in Esperanto, in the sense that the meaning shouldn't be alien to the components.

Esperanto avoids expressions whose meanings aren't suggested by the elements - it avoids idioms.

But that an X-aro may not encompass all aroj da X-oj is not inconsistent with this principle. If an X-aro was use to designate something that could NOT be seen as some aro de X-oj that would not be good Esperanto.

The phenomenon of circumscription of meaning works at the level of even the simplest compound eg root+finaĵo.

The finaĵo 'o' (tio kio estas) can mean:

a person (lernant-o);
an act (rigard-o);
a state (fart-o);
a result an action (not-o);
the manifestation of a quality (fort-o);
an abstract quality (feliĉ-o) .. and so on.

But with words like X-o the meaning is not alien to the elements.

There is no difference in principle between limiting a lernanto to a person and limiting arbaro to a forest.

Yes, those are all good examples.

I'm not complaining about this situation here because I want to change it. Esperanto is what it is, and there's no magic wand anyone can wave to make it different.

The reason I bring it up is because you've been talking quite a bit recently about how Esperanto, unlike the natural languages, should be able to be described by simple rules. By being artificial, Esperanto is special, and should not be treated as just another natural language.

But my example (and yours) point out a trap for beginners, precisely because we encourage them to follow rule-based patterns. These sorts of words only pretend to follow Esperanto's word-building rules - they look like they follow an established pattern, but they don't.

So when you are casting around for a word you don't know, and you rely on your word-building skills to create a word that expresses your intended meaning, you should be able to be confident that what you build, as long as it is according to the rules, is valid.

"Arbaro" is a trivial example, and I don't mean to hammer on it specifically, but stick with me here for a moment anyway. Maybe I am someone from a desert country without forests, or whatever. I've just learned Esperanto, and I see a set of ten or so trees over in a clump, and, using my newly minted word-building skills, decide to call that clump an "arbaro". Because that's logical. I (the person in my example) have never seen the word used before, so I have no reason to assume that I am using it incorrectly.

But I AM using it incorrectly if I refer to that clump as an arbaro. It's not an arbaro. Ten trees do not make a forest. It's logical, but it's WRONG.

Esperanto is supposed to avoid these sorts of traps, by virtue of its special, artificial nature. Here, though, it is behaving very unspecial and natural-language-like.

"Arbaro", and other similar words, show that you can't have full confidence when following the word-building rules; that what you've built actually means what you intend it to mean.

There is no rule that can cover these constructions. Again, I'm not saying we should (or even can) change this - I'm just saying "here are some examples that you cannot codify with any generalized rules". In these examples, the rules as they exist only confuse the issue, as they lead the beginner astray until he's "just memorized" the special list of words with narrowed meanings. Until he's memorized that list, there's a bunch of little language landmines lying around for him.

So when you are casting around for a word you don't know, and you rely on your word-building skills to create a word that expresses your intended meaning, you should be able to be confident that what you build, as long as it is according to the rules, is valid.

Yes, I don't think I can disagree with that. But that underlines the importance of getting the mechanisms of the language properly described in the lerno-libroj.

And in saying what I have just quoted, Riot, you bring out one of the ways that Esperanto differs from the natural languages.

The natural languages are cluttered with expressions that at one time had some transparency, but now are used as fixed expressions with the origins of their original use lost in the shrouded mists of time (eg learning the ropes, member of the upper crust etc).

Your hypothetical desert guy who knows nothing of arbaroj - has never learnt any geography at school, or watched American TV imports, or read illustrated European fairy tales in his youth - does not sound like the sort of guy who would ever want to or need to learn Esperanto.

But your average Esperantist, wondering what to say for a wood, who comes across arbaro, thinks 'brilliant!', I don't need to learn a special root for that.

He comes to the language with an idea of what a wood or forest is, and expects Esperanto to have some way of expressing that idea. Meeting 'arbaro' he makes the association.

He doesn't then get bogged down with philosophical considerations as to how many trees make a wood, any more than he would with regard to the word in his own language.

In other words, his pre-existing knowledge of the world and the things that need to be named, conditions his interpretation of compounds in Esperanto,

He doesn't have to separately learn the limitations of such compounds because he brings with him ideas of what the limits should be.

Continuing the theme:- you might argue that in different countries what counts as a matenmanĝo is different.

The Dutchman with his plate of sliced cheeses and meats, the Frenchman with croissants and jam, the Englishman with his cooked hot food.

But this does not make the learner of Esperanto suppose that a biscuit eaten at one minute to noon is a matenmanĝo because he ate it in the morning (matene manĝis ĝin).

Even if a matenmanĝo has different manifestations in different cultures, the learner already knows that a matenmanĝo is more circumscribed that any act of eating before mid-day.

I think that being surprised that "arbaro" translates most readily into "wood" or "forest", is possibly due to a misunderstanding of the motivations and usage of word-building in Esperanto.

Not least because, the most useful and sensible compound words in Esperanto, seem to have this feature that the final compound has an ever-so-slightly more special meaning than the summed meaning of its components. You usually have this scenario where compound word X made up of smaller words A and B is a kind of AB, but is more specific than just an AB.

E.g.
lern-ejo = school (not just "learning place")

The most often times when this phenomenon does not occur, is when there is no such specialised term in the language you're translating to.

I really feel that the argument I made previously on this page makes clear why compounds like this add expressiveness to Esperanto, and says why it's not desirable to interpret the purposeful compounding of terms simply as its literal parts. Even if blackberry was a bad example.

Esperanto takes advantage of how humans can deal with untangling ambiguity in speech. To make the word "wood" it is not necessary to say "medium-sized-collection-of-trees". You can get away with just "tree-collection".

This comes from the declaration of a leading Esperantist:

“Principo de neceso: En konstruon de vorto oni devas enkonduki ĉiujn vortradikojn, sufiksojn, prefiksojn kaj finiĝojn necesajn por elvoki klare kaj plene la ideon reprezentotan de tiu vorto. Principo de sufiĉo: Se la ideo tiamaniere esprimita estas jam bone kaj nekonfuzeble komprenata el la kunteksto sen iu el la sufiksoj uzataj, tiu ĉi sufikso povas esti elĵetata kiel neutila kaj nenecesa”.

Which I think is:

"Principle of necessity: In the construction of a word, one must introduce all root-words, suffixes, prefixes and grammatical endings necessary for clearly and completely evoking the idea to be represented by that word.
Principle of sufficiency: If the idea in the way expressed is already well understood, and unlikely to be confused in the given context, without some of the suffixes used, then these can be thrown away as useless and unnecessary"

This makes Esperanto extremely practical, usable and lightweight.

At the same time, these compounds lose their expressive power, if part of the ambiguity resolution taking place is to decide if the user was simply using the word to mean its literal parts. It's a result of using the smallest representation possible (principle of sufficiency) for a new word that interpreting with just literal components, diminishes the meaning of the compound.

Take a look at section 7 here, it has some interesting information on word building, including that quote.