High Valyrian Verbal Prefixes
A number of prefixes may be added to High Valyrian verbs. These elements seem to have originated as inflectional, but have gradually drifted towards derivational. By the time of AV, they are definitely derivational, but in classical HV they are still, arguably, inflectional.
In the examples we've seen so far, the prefixes appear to come in a set order. For the time being we will assume this order is mandatory, until we see evidence to the contrary, and organize the page on this principle. Prefixes will be listed in order of proximity to the verbal root.
Type I: jo-
- a·jo·memēbagon to continue to advance, to rage on
- jo·hegagon to continue to slaughter
- jo·mozugon to continue to drink
Type II: Valence markers
There are three prefixes of this type. They are mutually exclusive.
- a-, h-, s-, z- the instrumentive (also called "instrumental-passive.")
- i-, j- the oblique applicative
- u-, v-, b-, o- the locative applicative
The instrumentive prefix has four allomorphs: a-, h-, s-, z-. As we have very few examples, the distribution can only be guessed at at this point, but it does appear that word-initially before a vowelh- is used, and before a consonant s- or z- if it forms a licit cluster, a- if not. Here are all the forms known:
The instrumentive promotes an instrument to the subject. Thus:
- Vala egromy rōbir ezīmza. — "The man splits the fig with a knife." → rōbir egry hezīmza. — "The knife splits the fig."
If we wish to include the original subject of the sentence, it is shifted to an adpositional phrase with ondoso:
- Valo ondoso rōbir egry hezīmza. — "The knife of the man splits the fig."
It is likely that the instrument so promoted by the instrumentive prefix must immediately precede the verb, as occurs in our sample sentences above.
Some verbs require an instrumentive if the subject is inanimate (even if it is not, strictly, an instrument). For example:
- Mirri ōdria uēpi dōrī drējī zgiēñisi, sepār hen mībājȳr udīr ānogrosa anehussi. — "Some old wounds never truly heal, and bleed again at the slightest word."
- Vīlībāzma ajomemēbza — "The war marches on."
This is not, however, true of every verb:
- Lo jention mirre nūmāzme ēza... — "If leadership has any meaning..."
This form is functionally very similar to that which linguists refer to as the "instrumental voice" or the "instrumental passive." Indeed, the latter term is what David J. Peterson himself uses. However either of these terms runs the risk of being confused with the active and passive voice (ezīmza "he splits," ezīmaks "he is split"), and in fact the High Valyrian instrumentive is orthagonal to the category of verbal voice: it is quite possible for a verb to be both instrumentive and passive at once:
- rōbir ezīmaks. — "The fig is split (by someone)." → rōbir hezīmaks. — "The fig is split (by something)."
Or, with the instrument/agent explicitly stated:
- rōbir valo ondoso ezīmaks. — "The fig is split by the man." → Egromy rōbir hezīmaks. — "The fig is split with a knife."
As with the active, it is likely that when the focus of a passive sentence is on an instrument, rather than an agent, the instrumentive prefix is required. It is not known whether it is optional, or forbidden (but presumably it is not required) when both an agent and an instrument are explicitly stated.
Adding this prefix sometimes gives a word an idiomatic meaning.:
- tepagon "to give" → s·tepagon "to share."
- māzigon "to come" → *a·māzigon literally "to return" if amāzinon is any indication.
- mazverdagon "to create" → a·mazverdagon "to recreate"
- sikakson "to be born" → asikakson "to be reborn"
The oblique applicative takes the forms i- and j-. Full details of the distribution are not known: we have seen j- in j·iōragon "to accept," and j·epagon "to ask." However, DJP has made it clear that the latter should properly be iepagon, but it is commonly pronounced jepagon by modern speakers. Very likely it is i- if the word begins with a consonant, or a vowel with which it may form a licit rising diphthong, but j- if the word begins with i or u.
The oblique applicative promotes an indirect object, or possibly another oblique argument of the verb, to the direct object:
- *Belmurtoti vestrās kesīr pōnte jiōrinna... — "Say to the slavers I will receive them here..." → Belmurtī ivestrās kesīr pōnte jiōrinna... — "Tell the slavers I will receive them here..."
It is probable that the object so promoted must immediately precede the verb.
With an adpositional prefix
- *Yno syt vīlībilāt? — "Will you fight for me?" → Yne sytivīlībilāt? — "Will you fight for me?"
Adding this prefix sometimes gives a word an idiomatic meaning:
- derēbagon "to gather" → iderēbagon "to select"
- iōragon "to be standing" → jiōragon "receive (someone)"
The case of iderēbagon is particularly interesting. Since it literally means "to gather for," the applied object is not the thing chosen, but the person for whom it is selected. The thing chosen is in fact the indirect object. Thus:
- *Vala rōbra Aerot derēbza. — "The man gathers figs for Aerys." → Vala rōbrȳti Aeri iderēbza. — "The man selects figs for Aerys."
Thus, in the passive iderēbagon won't mean "be selected," but rather "be selected for; to be the recipient of a selection":
- Aeri iderēbaks. — "Aerys is selected for."
Note as well that, because one might just as well say *Vala rōbra Aero syt derēbza, one can also say sytiderēbagon "to select for," with the "for" explicitly stated:
- Vala rōbrȳti Aeri sytiderēbza. — "The man selects figs for Aerys."
- Aeri sytiderēbaks. — "Aerys is selected for."
The locative applicative takes the forms u-, o-, v- and b-. Full details of the distribution are not known, but very likely it is:
- u- word-initially before a consonant (e.g. unektogon), or a vowel with which it can form a licit rising diphthong (no examples yet); or medially between consonants (e.g. henujagon).
- v- word-initially before i or u (no examples yet), or medially between vowels (e.g. bēvilagon).
- b- after a prefix that ends in a consonant, before i or u (e.g. hembistan).
- o- after a prefix that ends in a (e.g. vaoresagon)
The locative applicative promotes an adpositional phrase with spacial reference to what is effectively an indirect object. The applied object will be in either the dative or the genitive, based on phonetic conditions: if the verb begins with a vowel, the dative is used (-ot in the singular), if a consonant (including h), then the genitive (-o in the singular). Thus:
- *Guēso gō iōrilen. — "I was standing under a tree" → *Guēso viōrilen. — "I was standing under a tree."
- *Gerpe hen guēsē nektotan. — "I cut a fruit off the tree." → *Gerpe guēsot unektotan. — "I cut a fruit off the tree."
The applied object, if it is explicitly stated, must immediately precede the verb. Furthermore, no word in the appropriate case may immediately precede the verb unless it is the applied object.
Note that in the examples above, there is no way to reconstruct what the original adposition wasin the non-applicative sentence. Viōrilen could mean "I was standing under," "I was standing on," or even "I was standing inside;" the only clue here is context: one typically stands under trees, not on top of them or inside them. Even this is not reliable: if the speaker was a sentry on watch in a forest, he may very well have been standing on the tree. (It is much harder to imagine what else unektotan could mean in the sentence above!)
With an adpositional prefix
It is possible to make a locative applicative less ambiguous by adding a prepositional prefix. The following prefixes are known to be usable with this prefix:
Locative applicatives sometimes have idiomatic meanings. For instance:
- ilagon "to lie" → bē·v·ilagon literally "to lie upon," but more broadly "to be incumbent on," "to be necessary for."
- māzigon "to come" → u·māzigon literally "to come from/onto/under/from/out of," but probably has the idiomatic meaning of "to come upon, to find" if AV umazigho is any indication.
Type III: Adpositional prefixes
Adpositions may be incorporated into an applicative. If this happens the adposition should not be prepeated independantly. In total, four adpositions are attested incorporated into verbs as prefixes: bē-, gō-, hen-, and syt-. It is not known if any other adpositions can be used this way (va seems like a good candidate), but it is certain that naejot cannot.
This group may be subdivided into prefixes that go with an oblique applicative, and prefixes that go with a locative applicative.
Oblique adpositional prefixes
Locative adpositional prefixes
- ↑ http://www.dothraki.com/2013/05/hepnon/#comment-1540
- ↑ Additionally, in personal communication, DJP has confirmed that the fictitious verb *sprintagon would be the correct instrumentive of the equally fictitious verb *printagon.)
- ↑ http://dedalvs.tumblr.com/post/82008642642/hey-im-a-huge-fan-of-the-work-you-do-on-game-of#comment-1323412628
- That's just variation. The prefix is generally only j- when it can't be i-, and it can be i- before e. These are coined on the fly, though; they're not separate words. For example, there is no word jepagon in my lexicon. I probably did j- for him since he's a modern speaker (kind of the way most Latin words beginning with iV- get turned into jV-).
- DJP, email, 2/1/14
- ↑ For instance, if you wanted to say "The woman's daughter went out," you could say *Ābro tala hembistas, but *Tala ābro hembistas could only mean "The daughter went out of the woman."
- DavidJPeterson: It would be qrimvilagon. To mislie, or to lie badly, or something.
- DavidJPeterson: To be misplaced.
- Mad_Latinist: OK, but I take it I shouldn't count that.
- DavidJPeterson: It's a licit construction.
- —IRC 11-20-13
- ↑ 7.0 7.1 http://www.dothraki.com/2014/01/asshekhqoyi-anni-save-save/#comment-7012