the process of making specific words / phrases from general ones

[Hyolobrika]

One process is "black bird" > "Blackbird" etc
Another is "person with white skin" > "White person" etc
In languages other than English how does this work in speech and in writing? I'm thinking particularly about the first process in Germanic languages. And does that morphological process always form a new word with a different meaning?
Native speaker input would be welcomed.

[zompist]

I'm not quite sure what you're asking. Blackbird is an example of a compound with a distinct lexical meaning— a blackbird is only one type of black bird. There are plenty of compounds with transparent meanings, easily taken from the components— e.g. a goatherd is a dude who herds goats. Your second example is also completely transparent.

Here's a page with dozens of German examples. Most are transparent; a few aren't.

Chinese and Quechua allow words to smooth together like this; French doesn't, it requires some syntactic glue: machine à ecrire, pays de Galles.

[Hyolobrika]

What I'm talking about is the morphological process. In written English, taking the space away. In spoken English AFAICT-only the change in emphasis.
I'm not talking about the relationship between the specific meaning and the general meaning (although we can)

[zompist]

OK. Those are two separate things. Cf. "White House", which has the initial stress but also a space. Whether we use a space, hyphen, or nothing at all is typographic. Generally as a term becomes more lexicalized it's more connected typographically.

As for initial stress, some initial-stress compounds are still transparent: chessboard, housewife, spearman, highlands, steamship, rainstorm, notebook, headband, bookshelf, etc.

[jal]

I think the initial stress is most remarkable when the first part of the compound is an adjective, as then you have a stress shift.


JAL

[Pabappa]

It's sporadic. E.g. in English, "spaceman" resists the shift to schwa, perhaps because it's a word used mostly by children. I think I've heard "fireman" and "policeman" both ways . Spearman would almost certainly have a schwa.

[Zaarin]

It's sporadic. E.g. in English, "spaceman" resists the shift to schwa, perhaps because it's a word used mostly by children. I think I've heard "fireman" and "policeman" both ways . Spearman would almost certainly have a schwa.

Homestar Runner frequently plays with the pronunciation of "-man" as /mən/ for humor. For myself, I have /mɪn/ as the reduced results in fireman and spearman (for me, /mən/ would be [mn̩], which is not the sound I have there).

[Linguoboy]

It's sporadic. E.g. in English, "spaceman" resists the shift to schwa, perhaps because it's a word used mostly by children. I think I've heard "fireman" and "policeman" both ways . Spearman would almost certainly have a schwa.

Homestar Runner frequently plays with the pronunciation of "-man" as /mən/ for humor. For myself, I have /mɪn/ as the reduced results in fireman and spearman (for me, /mən/ would be [mn̩], which is not the sound I have there).

"Spearman" with vowel reduction is liable to be confused with "spearmint" IMD.

In my morphology class, the professor pointed out that sandhi effects become more pronounced in compounds. His favourite example was "dru[k]store". "Ice tea" for "iced tea" is probably a case of this (followed by reanalysis as a productive compounding element, e.g. "ice coffee").

[Qwynegold]

In Swedish compounds differ from two separate words through intonation somehow. Here are some other changes that may happen, that I can think of right now:

• Nouns often get a genitive -s, e.g. papper+tallrik > papperstallrik (paper plate).
• Nouns in a few compounds get a weird, archaic change. This is some kind of fossilized stuff. E.g. flicka+barn > flickebarn (girl child); barn+mord > barnamord (child murder)
• Metathesis of -Cle and -Cre, e.g. äpple+juice > äppeljuice (apple juice)
• Nouns may lose a final -a or -e, e.g. pojke+kläder > pojkkläder (boys' clothes); matta+piska > mattpiska (carpetbeater)
• Nouns may replace a final -a with -o, e.g. kvinna+göra > kvinnogöra (women's tasks)
• Oftentimes no change is done to the noun (besides the intonation)
• Verbs lose their final -a, e.g. hyra+bil > hyrbil (rental car)
• Verbs may replace their final -a with -s, e.g. dröja (linger) + mål (case) > dröjsmål (delay)
• Verbs may replace their final -a with -o, e.g. göra (do) + mål (case) > göromål (chores)
• Orthographically, compounds are written together. A dash may be used if one of the words is an acronym or a compound that itself is not spelled as one word, e.g. TV+antenn > TV-antenn (TV antenna), chick flick + skådespelerska > chick flick-skådespelerska (chick flick actress); Preussisk blå + pigment > Preussisk blå-pigment (Prussian blue pigment)

Is this the kind of thing you were looking for?

[Linguoboy]

Another is "person with white skin" > "White person" etc

Actually, the process with white skin > white-skinned is even more interesting. The use of skinned here seems to imply the existence of a verb skin meaning "to have skin of a certain type". But no such verb exists; instead, the basic meaning of skin as a verb is "remove the skin from".

Diachronically, what we're dealing with is an adjective formant which happens to be identical to a past participial ending. But synchronically there's no way to tell whether fresh-skinned means "having skin which is fresh" or "skinned freshly" except context.

[Zaarin]

It's sporadic. E.g. in English, "spaceman" resists the shift to schwa, perhaps because it's a word used mostly by children. I think I've heard "fireman" and "policeman" both ways . Spearman would almost certainly have a schwa.

Homestar Runner frequently plays with the pronunciation of "-man" as /mən/ for humor. For myself, I have /mɪn/ as the reduced results in fireman and spearman (for me, /mən/ would be [mn̩], which is not the sound I have there).

"Spearman" with vowel reduction is liable to be confused with "spearmint" IMD.

Mine, too, but in this case context would usually make it clear: I'm not going to put a spearman in tea or resist the enemy cavalry in an RTS with spearmint. But the actual pronunciation is quite close [ˈspiɹ̠ˁmɪn] vs. [spiɹ̠ˁˈmɪnʔ~spiɹ̠ˁˈmɪ̃ʔ]. Stress is probably more useful for distinguishing the two than that word-final glottal stop. NB that spearman and spearmen are near-homophones but not quite: the second vowel in spearmen does not reduce: [ˈspiɹ̠ˁmɛn].

[circeus]

Chinese and Quechua allow words to smooth together like this; French doesn't, it requires some syntactic glue: machine à ecrire, pays de Galles.

I don't think this is an accurate characterization at all. In both English and French, noun-based compounds must conform to the rules of normal noun phrases prior to becoming set phrases. It's not the compounding that makes the resulting compounds different, it's the phrases they start out as in the first place.

In fact, french verb-based compounds do tend to smooth out. Many (most, probably) verb-noun compounds omit what would normally be a required determiner (although the meaning of one half of the compound is often eroded): garde-manger, porte-feuille, passe-montage, gratte-dos, rince-doigt, coupe-ongle, lave-auto, gratte-ciel, perce-neige etc. etc.

[zompist]

Chinese and Quechua allow words to smooth together like this; French doesn't, it requires some syntactic glue: machine à ecrire, pays de Galles.

I don't think this is an accurate characterization at all. In both English and French, noun-based compounds must conform to the rules of normal noun phrases prior to becoming set phrases. It's not the compounding that makes the resulting compounds different, it's the phrases they start out as in the first place.

Um, I don't quite know why you think you're saying something different than I did.

And sure, the VO compounds are simpler.

[jal]

[*]Verbs lose their final -a, e.g. hyra+bil > hyrbil (rental car)

Isn't this a case of using the verb stem instead of the full verb? That would be the same type of compounding as Dutch has (huren -> huurauto).


JAL

[Qwynegold]

[*]Verbs lose their final -a, e.g. hyra+bil > hyrbil (rental car)

Isn't this a case of using the verb stem instead of the full verb? That would be the same type of compounding as Dutch has (huren -> huurauto).


JAL

Hmm, it could be. I haven't actually studied this.

[Nortaneous]

In my morphology class, the professor pointed out that sandhi effects become more pronounced in compounds. His favourite example was "dru[k]store". "Ice tea" for "iced tea" is probably a case of this (followed by reanalysis as a productive compounding element, e.g. "ice coffee").

Food terms tend to drop -ed. Skim milk, ice cream, process cheese. Or temperature zones for food storage: you have the frozen zone for frozen shit, and the "chill" zone for refrigerated shit.

As for sandhi effects in compounds, consider the common AmE pronunciation of "newspaper" with, um, voicing assimilation.

[Travis B.]

As for sandhi effects in compounds, consider the common AmE pronunciation of "newspaper" with, um, voicing assimilation.

For me this is a case of frozen voicing assimilation. Synchronically my normal realization of /z/ that is not syllable-initial (and not preceded in the same word by an obstruent) or otherwise between vowels or semivowels is as [s] to begin with. However, that voicing assimilation took place in the past is betrayed by that the vowel of news in newspaper is short whereas the vowel of news by itself is long, even though they both normally end in [s] for me.

[Travis B.]

Another good example is the pronunciation of high school in NAE dialects with Canadian Raising of /aɪ/ where the vowel of high is raised (even though typically Canadian Raising does not operate across word boundaries).

[Pabappa]

I've never heard "noose paper" or "druck store", but i do have the raised vowel in "high school", along with a stress pattern like a typical compound word.

[Nortaneous]

Another good example is the pronunciation of high school in NAE dialects with Canadian Raising of /aɪ/ where the vowel of high is raised (even though typically Canadian Raising does not operate across word boundaries).

I don't have this.