It seems to me that sometimes one uses a root that is really an adjective in building a word, but if the pronunciation is hard, it seemed to me that PMEG states I should put in o. Example found in ESPDIC: lignoglavo even though it really is ligna glavo or wooden sword. In this case, is the o simply used as connecting roots that are difficult to say together, but does not actually say anything about the nounness of the root lign?
In standard Esperanto, you do not have complete freedom to glue an adjective to a noun to get a new noun. For example, you can say flavĉapela virino ("yellow-hatted woman", adj + noun → adj) but it does not follow that you can say Ŝi aĉetis flavĉapelon instead of flavan ĉapelon.
In the case of lignoglavo, the word is considered to be derived from glavo el ligno rather than directly from ligna glavo. Both roots are categorised as nouns in the Baza Radikaro Oficiala.
While it is usually a safe choice, there is no rule that only -o- can appear between elements; if you feel the need for a vowel, it should be the appropriate ending of the element before it. So for example you have dubasenca (from duba senco + -a), multekosta (from multe kosti + -a), or even vivipova (from povi vivi + -a; this only really happens in combination with devi, povi, voli). The accusative ending -n of a noun should disappear however.
Years ago, I read a description of a common error in Esperanto. I have not been able to find this article again, but it called this error "Hidden Elision of the Adjective Ending."
Hidden Elision of the Adjective Ending is when you take the -a off of an adjective and stick it to the front of the noun and pretend that the meaning has not changed. The term comes from this process: if we say bruna urso, we could attempt to elide the -a by saying brun' urso -- but since this is not permitted in Esperanto to have the adjective with no ending, some people will try to "hide" this by pushing the two together: brunurso. Whatever brunurso means, it is not the same as bruna urso.
The word lignoglavo is interesting. This word is found in PIV and defined either as a toy sword for children, or as a special kind of ceremonial Japanese sword. The only use of it in the Tekstaro is in Robinsono Kruso and appears to be a translation of the term "sword of wood". I suspect the Japanese usage is simply a literal translation of 木剣 (bokken). Certainly the swords in Robinson Crusoe are not children's toys since they can take off a man's head.
Certainly this makes sense as glavo el ligno (c.f. lignopeco) but these also strike me as isolated examples. This is another example why ESPDIC is really not a good resource for new Esperanto speakers. I'm told its stated goal is to document all usage - whether good or bad. If you had some other examples, it might be worth discussing those in more detail.
- Nova edzino - a new wife
- Novedzino - a bride
Usually, the meaning changes when you tack the adjective up front.
Yes. -O can be added for clarity or omitted if it is easier. In printed text, you can also add a hyphen for clarify. Just like any language, there are different ways of saying things.
- Lignoglavo = Wooden sword.
- Lignglavo = Wooden sword.
- Ligno-glavo = Wooden sword.
- Ligna glavo = Sword with an attribute of being wood.
- Glavo de ligno = Sword made of wood.
The roots may be joined together directly, or with an epenthetic (linking) vowel to aid pronunciation. This epenthetic vowel is most commonly the nominal suffix -o-, used regardless of number or case...
Kantobirdo (a songbird) versus birdokanto (a birdsong). Velŝipo (a sailship) versus ŝipvelo (a ship sail).