Editing Umbrian language (section)

==== Perfect formation ====

===== Perfect stem =====
Like other Italic languages, the Umbrian language merged the aorist and perfect tense found in [[Proto-Italic language|Proto-Italic]] and [[Proto-Indo-European language|Proto-Indo-European]], although the Sabellic languages, a language family of which Umbrian is a member, preserved the forms of the Proto-Indo-European athematic second aorist while Latin preserved the perfect forms of Proto-Indo-European. These etymological differences created numerous morphological discrepancies between the Sabellic languages and the Latino-Faliscan languages, the subgroup of Italic languages containing Latin. In Umbrian, the perfect subjunctive was marked by the addition of the vowel {{lang|xum|-ē-}} to the ending while in Latin, it was marked by the vowel {{Lang|la|-ī-}}.{{Sfn|Piwowarczyk|2011|p=107}} Umbrian perfect stems likely could be formed by 5 distinct types of modification applied to the present stem of the verb: reduplication, the simple perfect, k-perfect, f-perfect, and—a form exclusive to Umbrian—the nky-perfect.{{Sfn|Piwowarczyk|2011|p=107}} Reduplication was the most common method of forming the perfect in the original Proto-Indo-European language and it typically involved the addition of the vowel {{Lang|xum|-e-}} following the reduplicated syllable. Remnants of this technique appear in Umbrian verbs such as {{Lang|xum|peperscust}}, in which the initial consonant of {{Lang|xum|p-}} is reduplicated with an {{Lang|xum|-e-}} vowel added between the two letters. However, perfect forms that are—in origin—reduplicated perfects may not follow this pattern.{{Sfn|Piwowarczyk|2011|p=109}} This category, referred to as the "simple perfects," comprise verbs such as {{Lang|xum|dersicust}}, which likely derives from {{Lang|xum|*dedik-}}, the perfect stem of the Proto-Italic verb {{Lang|itc|*deikō}}. Although the original form was reduplicated according to the aforementioned pattern, the {{Lang|itc|-d-}} changed to {{Lang|xum|-ř-}} during the transition from Proto-Italic to Umbrian.{{Sfn|Piwowarczyk|2011|p=110}}

Umbrian perfect forms such as {{Lang|xum|andirsafust}} demonstrate the f-perfect, a type of modification that forms perfect stems through the addition of the consonant {{Lang|xum|-f-}}. The origin of this type of augment is unclear, although it may have derived from the [[univerbation]] of older terms. For instance, in the case of {{Lang|xum|andirsafust}}, the term may have originated from the {{Lang|itc|am-di-da-nt-s fust}}, although this etymology specifically is disputed.{{Sfn|Piwowarczyk|2011|p=113}} In Umbrian, perfect stems possibly could be marked through the addition of the consonant {{Lang|xum|-s-}}, a modification that likely originates from the original [[Sigma|sigmatic]] aorist of the Proto-Indo-European language. This form, the s-perfect, is entirely unattested in Umbrian with the possible exception of one form: {{Lang|xum|sesust}}. However, this form is more often interpreted as a reduplicated perfect, leaving no evidence of the existence of s-perfects in Umbrian and thus compelling some linguists to reject the existence of such forms in the language.{{Sfn|Piwowarczyk|2011|p=111}}  

The perfect marker {{Lang|la|-nsi-}}, {{Lang|xum|-ns-}}, or {{Lang|xum|-nç-}} appears in Umbrian terms such as {{Lang|xum|'''purtinçus'''}} and {{Lang|la|purdinsiust}}.{{Sfn|Markey|1985|pp=261-262}} The etymological origins of this root are unclear, it may have emerged from a reconstructed Proto-Italic form {{Lang|la|*-nki-}}, itself possibly related to Proto-Indo-European {{Lang|pie|*h₁neḱ-}} ("to bear, to bring.").{{Sfn|Markey|1985|p=263}} This form was preserved into the Ancient Greek term {{Lang|grc|ἤνεγκα}} ({{Lang|grc|ḗnenka}}), the aorist form of {{Lang|grc|φέρω}} ("{{Lang|grc|phérō}}," "to bear," to "bring") and the [[Old Irish]] suffix {{Lang|sga|-icc}}, found as a marker of [[perfective aspect]] in forms such {{Lang|sga|do·uic}}, the perfect stem of {{Lang|sga|do·beir}} ("to give," "to bring").{{Sfn|Markey|1985|pp=263-264}} The linguist Kenneth Shields, Jr. argued that this perfect ending originated from the combination of third-person singular forms ending in {{Lang|pie|*-Ø-}}  with the [[Deixis|deictic particle]] {{Lang|pie|*-N}}, creating {{Lang|pie|*-Ø-N}}. According to Shields, this form was later reanalyzed to produce  {{Lang|pie|*-N-Ø-}} and was then suffixed with {{Lang|pie|*-ki}}, culminating in the form {{Lang|pie|*-N-Ø-ki}}. This form may have then been reanalyzed as {{Lang|pie|*-nky-Ø}}, concluding the process of evolution and creating the Umbrian perfect morpheme.{{Sfn|Shields|1989|p=79}} Shields proposes that the term may be cognate with the Lithuanian imperative suffix {{Lang|lt|-ki}} and that the deictic particle {{Lang|pie|*-k}} can be observed in terms such as Latin {{Lang|la|cis}} ("on," "to this," "on this side") or [[Ancient Greek]] {{Lang|grc|τῆτες}} ({{Lang|grc|têtes}}, "this year").{{Sfn|Shields|1989|pp=79-80}} The linguist David Jerrett, noting that perfect marker exclusively appears in [[Denominal verb|denominal verbs]], argued that the perfect stem originated from nouns combined with the perfect forms of an unattested Umbrian verb deriving from Proto-Indo-European {{Lang|pie|ḱey-}} (meaning, "to lie down, to settle"), which may have developed a new meaning akin to "to set in motion, be in motion." Such a semantic shift occurred in other Indo-European languages: the Ancient Greek verbs "{{Lang|grc|κινέω}}" ("{{Lang|grc|kīnéō}}," "to set in motion, stir, meddle") or "{{Lang|grc|κῐ́ω}}" ("{{Lang|grc|kĭ́ō}}," "to go") and the Latin verb {{Lang|la|cieo}} ("to set in motion, move, stir") all demonstrate this transformation.{{Sfn|Jerrett|1974|pp=174-175}} When certain nouns were used alongside this unattested verb in [[Periphrasis|periphrastic]] phrases, they may have merged together to create new verbs. Jerrett cites one possible example of such a development in the verb {{Lang|xum|combifiansiust}}, which may have originated from the reconstructed noun {{Lang|xum|*combifiam}} combined with the verb form 3rd person singular future perfect active form {{Lang|xum|siust}}.{{Sfn|Jerrett|1974|p=176}} Thus, Jerret proposes a semantic shift from "{{Lang|xum|combifiam}} {{Lang|xum|siust}}," meaning "has made an announcement," to "{{Lang|xum|combifiansiust}}," meaning "has announced."{{Sfn|Jerrett|1974|p=178}}

===== Future perfect formation =====
Uniquely, Sabellic future perfects are marked with the ending {{Lang|xum|-us-}} and, in some cases, {{Lang|xum|-ur-}}. The "{{Lang|xum|-ur-}}" form appeared as, in Umbrian, intervocalic -{{Lang|xum|s}}- became {{Lang|xum|-r-}}.{{Sfn|Zair|2014|p=368}} Both forms are of disputed etymology: it is possible that it relates to the Proto-Italic form {{Lang|itc|fuiō}}, from the Proto-Indo-European form {{Lang|pie|bʰuH-}}. These verbs, both meaning "to be," evolved into the Umbrian form {{Lang|la|'''fust'''}}, which possibly predicated the development of the similar future perfect endings. However, the linguist Nicholas Zair suggests that, given the dual meaning of {{Lang|xum|'''fust'''}} as both a future and future perfect term,{{Sfn|Beeler|1980|p=3}} it is unlikely that it would evolve into an exclusively future perfect suffix. Furthermore, Zair considers it unlikely that the term would be reanalyzed into a unique suffix as it already consists of {{Lang|xum|*-fu-}} combined with the future marker {{Lang|xum|*-s}}.{{Sfn|Zair|2014|p=373}} One proposal to rectify these concerns suggests that the suffix may originate from a reduplicated future perfect stem {{Lang|xum|*fefus-}} , which, although formed from {{Lang|xum|*fe-fu-s}}, came to be reanalyzed as {{Lang|xum|*fe-f-us}}.{{Sfn|Zair|2014|p=373}} Alternatively, it may have emerged due to the generalization of the [[Zero grade|zero-grade]] [[Proto-Indo-European language|Proto-Indo-European]] perfect active participle root {{Lang|pie|*-us-}} or the lengthened grade {{Lang|pie|*-uōs-}} , itself possibly from {{Lang|pie|*-uūs-}}. In either scenario, the forms would yield to {{Lang|xum|-us-}} in Umbrian due to inevitable loss of initial {{Lang|pie|*-u-}} after most consonants and the loss of long {{Lang|pie|*-ū-}} in Oscan-Umbrian in non-initial syllables.{{Sfn|Zair|2014|p=369}} However, the linguist [[Madison Beeler]] critiqued this theory, arguing that there is insufficient evidence for the existence of a perfect active participle in any Italic language related to the Proto-Indo-European perfect active participle, and consequently no evidence for the existence of such a form of Proto-Italic.{{Sfn|Beeler|1980|p=4}} 

Another possibility is that this form is related the u-perfect in Latin, as seen in verbs such as {{Lang|la|habui}} or {{Lang|la|tenui}}.{{Sfn|Piwowarczyk|2011|p=115}} This theory holds that the original Sabellic future marker, {{Lang|xum|*-s-}}, likely combined with a perfect marker in {{Lang|xum|*-u-}} to form the Umbrian future perfect form {{Lang|xum|-us-}}.{{Sfn|Zair|2014|p=370}} Zair suggests that, although the Umbrian future perfect form was based on an original Sabellic perfect ending, it is entirely unrelated to the Latin {{Lang|la|-u}} perfect.{{Sfn|Zair|2014|p=370}} Instead, Zair argues that it was likely related to the possible [[South Picene language|South Picene]] {{Lang|xum|-ō-}} perfect formation, which is represented orthographically by {{Lang|spx|-ú-}} and may appear in terms such as {{Lang|spx|adstaíúh}} (meaning, "they set up").{{Sfn|Zair|2014|p=378}} According to Zair, the original Proto-Indo-European language formulated perfect terms through the reduplication of the initial consonant and the shift of the root into the o-grade, leading the creation of a perfect stem {{Lang|pie|*-ō-}} that was transformed into a future perfect stem in Proto-Sabellic through the addition of the morpheme {{Lang|xum|*-s-}}. Zair continues, proposing that the Proto-Sabellic language likely utilized the {{Lang|pie|*-ō-}} morpheme in its perfect and aorist tenses, although these were largely lost during the generalization of the perfect stems following the loss of the aorist tense, leaving the future perfect form as the only remnant of the original {{Lang|pie|*-ō-}} stem as there were no aorist parallels.{{Sfn|Zair|2014|pp=380-381}}