How do polite verbs develop?

[So Haleza Grise]

I am curious about the development of special "polite" verbs - for example, the situation where the correct choice of the form for "eat" depends on whether I am speaking formally or informally.

Obviously there are morphological markers that some languages use to indicate that a verb is being marked for politeness (e.g. Japanese -masu), and I am curious about how that kind of marking could develop. What I am more interested in, however, are suppletive forms.

Is the process language-specific (as verbal morphology often seems to be) or are there general patterns?

Something I could imagine happening: semantic drift. A reverse of the process associated with Vulgar Latin mandūcāre "chew" --> to the general verb for "eat". If edo remained as a high-status verb for "eat" then I suppose you could have formal versus informal alternatives.

In English it might be something similar to using latinate terms for higher registers - "consume" for "eat" "expectorate" for "spit", etc.

Does anyone have some useful resources on this topic?

[Pabappa]

all I can say is that there are different kinds of politeness, and some languages mix them together and others keep them apart. e.g.

1) politeness based on status of the speaker and listener in a social hierarchy. If Im of the "one of my ancestors switched to the losing side during a major battle and had to flee across the river with you guys" tribe, and I am expected to be ashamed of that, I'll have to forever speak in a compromised language that lets everyone know Im of the lowest class, except when Im with others of my own group. Some languages just have this sort of thing codified in general so that all people are inferior to the royal family and have to speak like that wehenever theyre in the presence of the royals.

2) Completely serparate, is the use of a special vocabulary when speaking *about* someone higher on the social hierarchy. the stereotype in England is "the queen doesnt shit", etc, and some languages lexify this such that you are expected to use a whole other vocabulary when speaking about the royal family, or anyone conceivbed to be superior. This is regardless of whether those people are in the room aor not. e.g. for example, imagine if in English "people walk, but the queen perambulates". I took this based on Malay, someone here might be able to give more info. In some langs, there are pronouns that differ for the respectfullness of their referent.

Conceivably there might be some langs that have both kinds of politness rules, such that it's a matrix, and the number of speech registers might be as high as the square of the number of rungs on the social hierarchy.

3) A third type I dont know much about involves using different words for the same concept irrespective of who is being talked about. Most or perhaps all languages have taboos about bodily functions. This is different from the other two because its based primarily on consideration of the listener ratehr than the social status of either the speraker or listener. In English its almost inverted from type 1, since children are spoken to with the most euphemistic vocabularies of all , yet are expected to respect their superiors" (elders) in every other way.

So a language can tie this up with the others as well, and I believe that this happens in Japanese. The third type is probably less discvreet than the others since people will have different feeings as to what is a rude word or not, and speakers are expected to use their best judgment.

This can also apparently happen with nouns, such as Japanese https://en.wiktionary.org/wiki/御#Japanese <--- the word "rice" is now always marked with a politeness prefix, for all speakers of all backgrounds, while the affix is optional on most other words.

[Salmoneus]

I don't think there's likely to be any clear rules - taboo avoidance in general tends to be quite ad hoc.

Blust reports regarding Javanese speech levels:

- the ngoko (informal) vocabulary is almost always the inherited vocabulary (though not in absolutely all cases)
- the krama (formal) vocabulary is often templatically derived from the ngoko. Blust identifies five templatic alteration, three of which are different ways of replacing particular back or low vowels with high front vowels, which Blust attributes to universal sound symbolism. The other two, however, have no clear motivation, and one even goes against sound symbolism (so, final -a in ngoko is often final -i in krama, but final -i in ngoko can sometimes be final -os in krama). There is also an affix that is prefixed or suffixed to grammatical morphemes to make them krama
- in other cases, krama words represent borrowings (often from Malay) while ngoko words are native - but occasionally the opposite is true
- a few words seem to be doublets derived from the same parent language (due to now-lost dialects, perhaps?)
- there's also a small intermediate speech level, which seems to often be an 'abbreviation' of the krama

In Pohnpeian, by contrast (where there are three speech levels):

- most lexical differences are inexplicable and unpredictable
- there are some suffixes sometimes used in deriving royal honorific words from ordinary honorific words
- one such suffix derives from the word for "sky" (basically, if you ask a peasant if she likes her grub, and a nobleman whether he likes his repast, you ask the king whether he likes his heaven-repast).
- there are a few words where respect is shown by excessively lengthening a particular vowel (including greetings, so essentially 'whazzuuuuuuuuuuup' is a marker of deep respect)

Whereas in Samoan (where there are two levels):
- a sizeable number of polite expressions are transparent circumlocutions - so ordinary 'gun', but polite 'strong implement'
- again, it's the low-register terms that are inherited, and the upper register that are innovated (though there are exceptions). Many innovations are opaque, however
- unlike Javanese or Pohnpeian, Samoan high registers sometimes conflate terms that are distinct in lower registers. So there are different low-register words for "eye", "mouth" and "nose", but these all match the same term in the high register

In Tongan:
- "sky" is actually a replacement word for "face", "head", "eye", and "to bury", and "skysky" is a replacement word for "to warm sth. by a fire"

-----------------

Also worth bearing in mind:
- the general ideology of the distinction. Javanese speech levels reflect aesthetic considerations, and are motivated by the desire to seem cool: failure to respect the rules was punished by social disregard. In Micronesia and Polynesia, however, speech levels were motivated by fear: in the Tongan sphere, this is the worry that failure to show respect would, automatically, lead to being the victim of the powerful one's magical powers (whether the powerful one wanted this or not), whereas in Micronesia the concern was that failure to show respect would incite the ancestral ghosts of the powerful one to seek vengeance.


- the nature of hierarchy involved. Javanese speech levels reflect the general social status of the listener, whereas the oceanic systems are all specifically keyed to inherited social class (that is, in Pohnpei you use high registers with bankrupt kings and low registers with millionaire peasants, whereas in Java you use the high register with the millionaire and the low register with the homeless guy, regardless of their bloodlines).

- the triggering factors. Notably, in addition to the two main speech levels in Javanese, triggered by the status of the listener, there is a smaller specialised vocabulary triggered by the high status (or personal need to respect - eg parents, teachers) of the person being talked about, which is orthogonal to the main distinction [so, two schoolfriends chatting will use ngoko, but still use the specialised respectful words when describing teachers, say]. There's then a very small additional vocabulary used when the speaker is the subject of a verb and is showing deference to the object, used with a tiny number of verbs (like 'request'). Similarly, Samoan high register is usually triggered by the status of the listener, and sometimes of the person referred to, but the low register is then again triggered when the things referred to are the speaker's family or possessions.


----------------
--------------------
------------------------


So, I'd suggest higher speech forms may involve:

- loanwords (or loanwords from different languages)
- templatic substitutions (Blust also discusses Atayal, in which it seems as though the vocabulary of entire languages has been derived through a sort of totally unpredictable, semi-random pig-latin game, presumably from a male prestige secret language that has replaced the original vocabulary in most languages (but remains male-only in a couple))
- affixation
- replacements of particular terms with general terms
- opaque substitution by less fundamental synonyms
- kenning substitutions (as in poetic registers of Old English or, in abbreviated form, in Cockney Rhyming Slang)
- spelling pronunciations (Blust mentions the male-only use of "literary pronunciations" in Cham, signalling literacy and access to written texts)
- circumlocutions


--------
-----


And just for fun, two example sentences from Blust:

Aku wis maŋan səgane
Kula sampun nədha səkulipun
Both mean "I have eaten the rice". The first is from a teenage girl to her sister, and the second is from the same teenage girl to her uncle.

But:
Bapak wis dhahar səgane
Bapak sampun dhahar səkulipun
Again, both are from a teenage girl: the first to her sister, the second to her uncle; but here, she's saying "Father has eaten the rice". This shows the orthogonal lexical replacement systems. In the first pair, every word changes to show the speech level. But in the second, not only is the word for "father" invariate, but there is a third word for "to eat" used in both speech levels when the subject of the verb is worthy of respect (here, a parent).

[malloc]

I have always wondered how the suffix -masu in Japanese originated. Presumably it developed from an auxiliary verb that became permanently attached to the preceding verb stem. But how did the sense "this whole statement is uttered in a formal setting" originate?

[zompist]

There's a whole book on this, Penelope Brown and Stephen Levinson's Politeness: Some universals in language usage. I read it years ago, so I can't say I remember much of it, just that politeness strategies are super-common and easily get grammaticalized. And they do follow certain paths: humble forms referring to oneself (e.g. 'servant' > 'I'), grandiose or simply indirect forms referring to someone else (e.g. 'that side' > 'you').

Japanese -masu seems to come from mairu 'come/go' + suru 'do'. Brown & Levinson say it comes from a polite causative. In any case it's a lexical item that's become grammaticalized as a simple marker of politeness.

Shibatani (The Languages of Japan) notes that the honorific system in Japanese has considerably simplified since the shogunate. Honorifics go back to the Heian period, but reached a dizzying level of complexity in the Edo period.

I don't think the suppletive forms are hard to understand; after all, English is rich in register variations: "Whatcha eating?" vs. "Will you be dining in tonight?"

[So Haleza Grise]

Thank you for all of these! I will see if I can find the book.