Reading the very amusing threads on this forum about the purity of language and the opposition against introducing new roots to the dictionary, a thought vilely ambushed me:

Do neologisms actually exist?

No doubt, we start to use new words (drone, blog, lol, investment bubble, fracking) continously, but: how new are they really?

Most neologisms are either:

• Similes or metaphors or metonymies: a blitzkrieg is a war carried out as fast as a lightning. A drone (the Obama ones) is a flying object without brains that is commandeered by another entity; much like its entomological name-giver
  
• Expansion (or shifting) of the original meaning: an "investment bubble" is not a liquid formation held together by its surface tension, but it is as fragile.
  
• Acronyms or contractions or apocolocyntoses: Laser, blog, vlog, grep, scuba.
  
• Loans from other languages.
  
• Composed from existing words (roots): helico-pter = screw wing; mob-ile tele-phone = move-able far-voice.

But when do we actually have completely new words? From scratch?
I mean, a new arrangement of sounds, unharkened so far, which is not analyzable?

I really don't grok it in fullness.

Such coinages are indeed relatively very rare. I don't believe that when people say "neologism" they're thinking only of entirely "new" words - otherwise you will indeed be speaking of so little a set as to be not particularly useful.

One process that sort of toes the line between coinage from existing roots and coinage ex nihilo is neologism derived from sound similarities. For example, the English chortle is (if I recall correctly) a mashup of the words snort and chuckle. Shart is a somewhat less couth example. (Interestingly, I'm not sure I know of any cases of this sort of word formation being responsible for any non-obscure Esperanto words.)

Quark is a good example of a meaningless sound that was applied to a new concept because it sounded right to somebody. Genericized trademarks are another good source of words, like "to google" (which one can certainly do without Google). The previous meaning of that sound (googol) was also a good example of a "hey, let's go with this sound for that" neologism.

(And yeah, there's Heinlein's contribution.)

Eltwish:Such coinages are indeed relatively very rare. I don't believe that when people say "neologism" they're thinking only of entirely "new" words - otherwise you will indeed be speaking of so little a set as to be not particularly useful.

One process that sort of toes the line between coinage from existing roots and coinage ex nihilo is neologism derived from sound similarities. For example, the English chortle is (if I recall correctly) a mashup of the words snort and chuckle. Shart is a somewhat less couth example. (Interestingly, I'm not sure I know of any cases of this sort of word formation being responsible for any non-obscure Esperanto words.)

Quark is a good example of a meaningless sound that was applied to a new concept because it sounded right to somebody. Genericized trademarks are another good source of words, like "to google" (which one can certainly do without Google). The previous meaning of that sound (googol) was also a good example of a "hey, let's go with this sound for that" neologism.

(And yeah, there's Heinlein's contribution.)

I've started a list of words that COULD be English words (based on the phonotactic rules of the language) but aren't (as far as I can determine from looking at Merriam-Webster online). I've got quite a number; started with the nucleus vowel, e.g. (short) a, added a consonant (or a cluster) to the end then started adding phonemes to the beginning. Started out with -ab(b) and listed from b all the way to z-, skipping those that were already words (bab, brab, chab, hab, mab, pab, etc) then went on with -ack (brack, cack, chack, etc).
So far i've done -ab, -ack, -ap, -ape, -eep, -ipe. I should start to be more methodical, i.e. go back and pick up -abe, -ake, -ad, -ade, -af(f), -afe, etc..
Anyway, I wonder how such a list for Esperanto phonotactics would look. And maybe instead of bringing in needlessly heisted from other languages roots, one could go to this list if a root showed it was needed.

orthohawk:

Eltwish:Such coinages are indeed relatively very rare. I don't believe that when people say "neologism" they're thinking only of entirely "new" words - otherwise you will indeed be speaking of so little a set as to be not particularly useful.

One process that sort of toes the line between coinage from existing roots and coinage ex nihilo is neologism derived from sound similarities. For example, the English chortle is (if I recall correctly) a mashup of the words snort and chuckle. Shart is a somewhat less couth example. (Interestingly, I'm not sure I know of any cases of this sort of word formation being responsible for any non-obscure Esperanto words.)

Quark is a good example of a meaningless sound that was applied to a new concept because it sounded right to somebody. Genericized trademarks are another good source of words, like "to google" (which one can certainly do without Google). The previous meaning of that sound (googol) was also a good example of a "hey, let's go with this sound for that" neologism.

(And yeah, there's Heinlein's contribution.)

I've started a list of words that COULD be English words (based on the phonotactic rules of the language) but aren't (as far as I can determine from looking at Merriam-Webster online). I've got quite a number; started with the nucleus vowel, e.g. (short) a, added a consonant (or a cluster) to the end then started adding phonemes to the beginning. Started out with -ab(b) and listed from b all the way to z-, skipping those that were already words (bab, brab, chab, hab, mab, pab, etc) then went on with -ack (brack, cack, chack, etc).
So far i've done -ab, -ack, -ap, -ape, -eep, -ipe. I should start to be more methodical, i.e. go back and pick up -abe, -ake, -ad, -ade, -af(f), -afe, etc..
Anyway, I wonder how such a list for Esperanto phonotactics would look. And maybe instead of bringing in needlessly heisted from other languages roots, one could go to this list if a root showed it was needed.

But your proposal would turn Eo from a posteriori to a priori (or the other way round, I never remember which is which).

Eltwish:The previous meaning of that sound (googol) was also a good example of a "hey, let's go with this sound for that" neologism.

Interesting enough the word googol (the 100th power of 10) was coined by a 9-years-old. Or at least wiki says so.