Vadi

Introduction

Vadi is an extinct language once spoken in Minhay. A small parchment fragment was discovered in April 1963 in a cave outside of Peħħat, a small township in Sakkeb Prefecture. More fragments were discovered in subsequent years in the vicinity of the cave, named Aħħur, where the first texts were discovered. Finally, in the late 90's, the wonderfully preserved Kalapái Scriptum were discovered in an isolated hut approximately ten kilometres north of Peħħat. These texts were dated to the mid to late 1800's CE. The Kalapái Scriptum is a collection of letters between two farmers, Éro and Sorvin, who were embroiled in an ongoing feud regarding the property lines between their lands. The letters were written in Vadi intermixed with words from the unrelated Peshpeg and Minhast languages. A few letters were written entirely in the the extinct Minhast Knife Speaker dialect. Also found among the letters are legal papers drawn from the Prefect of Dog Speaker Country. The farmers' letters contained several texts clearly indicating code-switching between Vadi and the Knife Speaker dialect. The portions containing the intermixed Knife Speaker and Peshpeg words were used to decipher the Vadi texts. The Dog Speaker papers did not contribute directly to the decipherment of the language, but as an external source it provided a great deal of context of the nature of the feud between the litigants. This external contextual source clarified the translation of otherwise ambiguous passages. The Kalapái Scriptum is thus popularly referred to as the "Minhast Rosetta Stone".

Today, two linguists, Dr. Iyyaħmi min Ruššur of the University of Aškuan, and Dr. Napayshni Tashunka of the University of the Lakota Nation at Three Pipes, are the two leading Vadists of the the Šibbūru¹ school. A third Vadist, Dr. J.F. Schumann of Universidad Autónoma de Madrid¹, is the pre-eminent leader of the traditionalist school.

This article presents the findings and analyses of both schools, the Šibbūru school as represented by Iyyaħmi and Tashunka, as well as the opposing viewpoint of the traditionalists, as represented by Schumann's.

Phonology

The excellent condition of the texts from the Scriptum were a veritable gold mine for understanding Vadi phonology and phonotactics and led to the conclusion that the phonemic inventory was underestimated. The texts from the Scriptum also eventually dispelled earlier views that Vadi phonotactics were simple. Rather, Vadi phonotactics were far more complex than earlier thought.

The texts were written entirely in the indigenous Minhast Širkattarnaft. While the Širkattarnaft, an abugida, works quite well for the Minhast language, it has presented a major challenge to Vadists in determining the phonology and phonotactics of Vadi. It became immediately apparent that there were several spelling variants and inconsistencies, and numerous occurrences of digraphs appear in both litigants' texts. Many of the digraphs also utilized some of the few surviving ideograms in the Širkattarnaft (e.g. the <MIN> ideogram).

Tashunka, who was working on Minhast dialectology at the time of the discovery of the Scriptum, was among the first scholars to examine the texts. He immediately realized, based on the aforementioned anomalies, that there were phonemes not accounted for in previous Vadists' works. Based on these anomalies, he posited additional phonemes in Vadi. From his analyses, he proposed that either the phoneme /β/ or /v/ was part of the phonemic inventory of Vadi, based on the digraph <bi><f> found in the Scriptum's texts. These digraphs coincided with the alternation of <ba> and <wa> in the Aħħur texts. So while the Aħħur texts' spelling for "Vadi" alternated erratically between <ba><di> or <wa><di>, the texts from the Scriptum consistently spelled "Vadi" as <bi><f><'a><di>. Similar correlations of the anomalies the Scriptum with the Aħħur materials led him to conclude there was also a /ð/, and another albeit indeterminate sibilant, perhaps /ɕ/. Several years after Tashunka returned to his work on Minhast dialectology, Iyyaħmi took up where Tashunka left off and determined that the sibilant was the apico-alveolar fricative /s̺/.

Iyyaħmi, inspired by Tashunka's earlier work, later discovered previously unknown mutations and other sandhi processes in the language. He developed a new romanization scheme show the complex sandhi processes that were represented by the litigants' unorthodox use of the indigenous Minhast Širkattarnaft script.

Before turning to the actual phonemic inventory and phonotactics of Vadi, a knowledge of the three principal orthographies is essential, as these orthographies illustrate how the Vadi mutation system was eventually reconstructed by the Šibbūru school. Their findings remain a hotly contested issue beyond Vadi's phonological system, as it poses major implications in other areas of the Vadi language, especially morphology and morphosyntax.

Consonants

Vadi Consonantal Inventory

	Bilabial	Alveolar	Palatal	Velar	Uvular	Glottal
Nasal	m	n
Plosive	p b	t d		k g		ʔ
Fricative	v	θ* ð* s s̺*				h
Affricates			d͡ʒ
Approximants			j
Flap		ɾ
Lateral		l ɬ*

* Not reflected in Schumann's phonemic inventory.

Vowels

	Front	Near- front	Central	Near- back	Back
Close	i u o ɛ a
Near-close
Close-mid
Mid
Open-mid
Near-open
Open

Prosody

Stress

The issue of stress, as in most areas of Vadi phonology and grammar, is a continuous are of contentious dispute between the Traditionalist and Sibbūru school. The texts discovered at Aħħur showed various marks adjacent to the Širkattarnaft characters, which early Vadists interpreted as diacritics indicating stress. This conclusion was based on the observation that a vertical slash would appear either on the ultimate or penultimate syllables, with more than half appearing on the final syllable. On about five percent of words in the entire Aħħur corpora, both the ultimate and penultimate syllables bore the slash mark.

The corpora of the Kalapái Scriptum, provided evidence that vowel length was also a significant feature of the Vadi phonological system and affected word stress. Although the writers of the material from the Scriptum used the Širkattarnaft very differently from those found in Aħħur, the Scriptum provided a richer dataset for researchers to demonstrate that word stress in Vadi was more complex than earlier thought. Rather than being simply a fixed lexical feature, word stress could shift from various factors, especially when morphophonatactic interactions triggered mutations and word-internal sandhi processes.

Phonotactics

Before the Kalapái Scriptum was discovered, little was known of Vadi phonotactics. The spelling in the Aħħur texts was consistently CV, with CVn allowed in final syllables. The spelling yielded no discernible evidence of sandhi processes, if anything it suggested that Vadi phonotactics were quite uncomplicated. Once the Kalapái Scriptum was discovered, this viewpoint shifted significantly, creating a rift in the Vadist community which led to its split into the Traditionalist and Šibbūru schools.

For the Šibbūru school, the excellent condition of the texts from the Scriptum were a veritable gold mine. It led to a revolutionary reevaluation of Vadi phonology and phonotactics and revealed that Vadi phonotactics were anything but simple. If anything, the Scriptum revealed that just like the knowledge of Vadi phonology was incomplete, its phonotactics were also greatly underestimated.

Iyyaħmi's work led to the discovery of consonantal mutations reminiscent of the Celtic languages, which initially sparked much controversy in the Vadist community. To support his theory, he grouped several of the spelling anomalies shared by both litigants into different categories, and via various statistical analyses, he was able to establish frequencies of these anomaly classes which demonstrated that the litigants' spellings were anything but random, but followed discernible patterns that were statistically significant. These patterns led Iyyaħmi to conclude that only underlying sandhi processes could account for these spelling patterns. Indeed, his statistical evidence was so compelling (Iyyaħmi's analyses were significant with a p-value of less than .05) that the present-day consensus among Vadists is that mutations and other sandhi processes are indeed a significant feature of the language.

Regardless, the Traditionalists have been vocal about their criticisms towards Iyyaħmi's analyses. In particular, Schumann has argued, and continues to maintain, that the spelling anomalies in the texts simply indicate dialectal differences between the litigants.² Iyyaħmi does agree that dialectal differences may account for some of the spelling anomalies. However, this argument alone is too simplistic and is insufficient to explain away the frequencies.

Mutations

According to the Šibbūru school, Vadi mutations fall two broad mutation processes, further divided into several subtypes:

1. Fortition
2. Lenition

Mutation Table

Mutation Type	Subtype	Realization	Trigger
Fortition
	Voicing	p -> b t -> d k -> g
	Defricativization	ð -> d θ -> t s -> θ f -> p
	Gemination	p -> pp k -> kk t -> tt l -> ll n, m -> mm m, n -> nn	Etymologically resulted from a preceding nasal consonant /m,n/
Lenition	Spirantization	d -> ð t -> θ θ -> s ɫ -> s̺ p -> f
	Palatalization	d -> d͡ʒ t -> t͡ʃ s ->ʃ
	Debuccalization	s -> h

For example, the combination of the past tense marker plus the genitive triggers voiced fortition and syncope with kilay "heart, interior" in the phrase (Tashunka's transcription) Junyé:ne glai haglev man ukanyen "I will enter your house and ruin you", i.e. junyé:ne glái << jula-nai-na kilay. If the future marker -nai- did not appear, fortition would not be triggered. Instead, the result will be Julana kilai haglev man ukanyen "I am entering your house to ruin you."

Gemination² occurs among a restricted set of consonants, particularly /k/, /n/, /l/ and /t/. Consonantal mutations occur, with evidence of both lenition and fortition. Metathesis also occurs with certain consonantal combinations, traceable to Minhast influence.

Orthography

As mentioned earlier, the Scriptum was entirely written in the Širkattarnaft, an Indic-derived script containing graphemes of an even older ideographic system:

The Širkattarnaft

• 

The Širkattarnaft found in the Scriptum, unlike that of the Aħħur materials, was replete by numerous digraphs, spelling anomalies and other irregularities throughout the texts. The older Romanized orthography, used by Schumann and favoured by traditionalist schools, comes from earlier material developed during early 1960's, before Iyyaħmi's and Tashunka's later systems were developed.

Traditionalist Romanization System
a, á, e, é, i, í, o, ó, u, ú, b, p, f, v, d, t, g, k, j, n, m, l, r, z, s, h, w, y

Iyyaħmi uses devised a new Romanized system for glossing purposes, developed independently of Tashunka. This system was a compromise system of providing a phonetic representation of the Širkattarnaft where the litigants used it in the orthodox Minhast manner, while at the same time including the digraphs, spelling anomalies and irregularities the litigants used to represent Vadi phonemes not found in the Širkattarnaft, or to represent cues that a mutation has taken place. Iyyaħmi's system was inspired by Assyriologists who use a combination of hybrid system employing characters to provide a phonemic representation of Akkadian glyphs for transcribing phonemes, intermixed with a convention of representing Sumerian ideograms, also using Latin characters.

Within Minhay, Iyyaħmi's publications use the original Širkattarnaft to maintain Vadi's historical link with the actual system in the Scriptum; the litigants' idiosyncratic use of the Širkattarnaft is also well known among native Minhast linguists. However, for international publications, Iyyaħmi uses his Romanized system for transcribing Vadi. This transcription system aims to combine the phonemic goals of the Ammerkast system with that of the Širkattarnaft as used by the litigants. For these reasons, Iyyaħmi calls this the Illum (bridge) system.

Iyyaħmi's Illum Transcription System
Type	Phonemes (Ammerkast-Derived)	Romanized Širkattarnaft Digraphs
Vowels	a, ā e, ē i, ī u, ū	(none)
Vowel Digraphs	á, â é, ê í, î o, ō ó, ô, ú, ū	<MIN><'a>, <MIN><'ā> <MIN><'e>, <MIN><'ē> <MIN><'i>, <MIN><'ī> <SUT><'o>, < SUT ><'ō> <MIN><'o>, <MIN><'ō> <MIN><'u>, <MIN><'ū>
Consonants	b, p, f d, t g, k n, m l, r z, s, š h, w, y	(none)
Consonants Digraphs	[v] [d͡ʒ] [ð] [ɲ] [ɬ] [s̺]	<bif> <idy>, <diy> <difš>, <dift> <nh> <fisl> <isš>

Iyyaħmi's system also contains stand-alone graphemes that are used to signify where mutations have taken place:

Mutation Table

Mutation Type	Subtype	Grapheme	Affected phonemes
Fortition
	Voicing	<gu>	/p/: <gu> ~ <p> -> /b/ /t/: <gu> ~ <t> -> /d/ /k/: <gu> ~ <k> -> /g/
	Defricativization	<du>	/ð/: <du> ~ <difš>, <dift> -> /d/ θ -> t s -> θ f -> p
	Gemination	<gu>	p -> pp k -> kk t -> tt l -> ll n, m -> mm m, n -> nn
Lenition	Spirantization	<fu>	d -> ð t -> θ θ -> s ɫ -> s̺ p -> f
	Palatalization	<yu>	d -> d͡ʒ t -> t͡ʃ s ->ʃ
	Debuccalization	(none)

Tashunka incorporated Iyyaħmi new findings on Vadi phonology and phonotactics to an alternate orthography focused on the actual phonetic realization after sandhi processes were applied. Tashunka's system has become very popular among Vadists, and now the prevailing practice is to present Tashunka's and Iyyaħmi's transcription side by side, the first for its simplicity and focus on the final phonetic realization in a format familiar to many linguists who utilize the Americanist system, and the latter for illustrating underlying mutation and other sandhi processes. Where mutations occurred, Tashunka makes note of that in his glosses.

The Tashunka system indicates stress with an acute accent, and vowel length by a colon, e.g. ta:dévi "encroachment". The complete Tashunka system is represented in the following table:

Tashunka's Romanization
a, ã, á, e, é, i, í, í, o, ó, u, ú, b, p, f, v, d, ð, t, j [d͡ʒ], g, k, n, m, l, ɬ, r, z, s, s̺, š, h, w, y

One phoneme represented in Tashunka's system that is absent in Iyyaħmi's system is /ã/. In some texts, a certain ideograph, the <MIN> ideogram appears twice in a row followed by the Širkattarnaft character <'ā> representing the glottal stop followed by a long vowel. In other texts this ideograph is absent, even though the words have the same meaning. So whether the phoneme /ã/ exists in Vadi phonology remains indeterminate.

All three Vadists do employ IPA to represent the actual phonology of the textual material, but even here significant divergences often emerge. The IPA of Iyyaħmi reflects his more recent work on how the Širkattarnaft was used to represent Vadi phoħħnology and phonotactics, and for the most part Tashunka's IPA is concordant with that of Iyyaħmi. The traditionalist system Schumann uses is the main source of discordant IPA renditions of the same text. These divergences can impact other areas of reconstructing the Vadi language, as the IPA rendition in one system may yield separate words or morphemes that are lacking in the other system.

These different orthographies are emblematic of the split in the Vadist community. The existence of the Aħħur texts have only served to add confusion, as the spelling in the fragments of the textual material is consistent and regular, suggesting the Vadi phonemic inventory is even smaller than that of Minhast. A compromise solution has been advanced by a minority of Vadists, that the Aħħur texts represent an even later form of Vadi increasingly influenced by the dominant Minhast language. Thus far, no conclusive evidence has been found to support this hypothesis. Alternative tests, non-linguistic tests for this hypothesis have been proposed, the most recent one being carbon dating of the Aħħur and Kalapái parchments. Unfortunately, carbon dating analysis of the texts cannot provide support for that hypothesis as both corpora have been already contaminated by the number of hands that have handled the material.

Morphology

Vadi is a mildly fusional language with some agglutinative characteristics. Cliticization is a prominent feature of the language, such that considerable debate has arisen as to whether some affixes should be reclassified as clitics. Further adding to this debate is the ambiguities arising from the orthography in the texts: a given morpheme may appear as part of a word, and other times may be written as separate and distinct from its target word. Where the morpheme is written separately from its host, the initial or internal phoneme of the separated morpheme is often represented by a different character from the one usually used when it is attached directly to its host. This most likely indicates some sort of sandhi process occurred, suggestive of an affixal status, despite being written separately. Recognizing the clitic-vs-affix debate remains a contested area, this article generally follows Schumann's classification system, although some conventions based on Iyyaħmi's more recent work will be used, as many of his analyses have been have become the prevailing view among Vadists. Iyyaħmi's conventions, wherever they appear in this article, will be noted.

Nouns

Gender

Vadi nouns for the most part do not mark for gender, but there appear to be remnants of gender marking on special singulative forms, interestingly confined to body parts which also demonstrate an interesting split on the number system:

Gender-Number Marking on Body Part Nouns

Body Part	Singulative	Unmarked	Plural
Head	--	jidár	jidárha
Eye	kulúri	kulun	--
Ear	támika	taka	--
Nose	--	uvaz	uváza
Heart	--	rilai	rilaia
Hair	varláka	varlat	--
Lips	túnturi	tuntu	--
Tongue	--	nipáz	nipáza
Finger	patáka	pata	--
Hand	uzáka	uzap	--
Testicle	vúlari/vúluka	vula	--
Penis	niátari/niátak(a)	niat	niata

The singulative forms seem to roughly correspond with animacy. The -(r)i forms seem associated with animate nouns, whereas the -ka forms correspond largely with inananimates. However, exceptions do appear, as in uzáka instead of expected uzap(r)i "hand". The plural forms, consisting of the sole affix -a, shows no animacy distinction.

Unfortunately, these body parts are the only ones attested from the Scriptum. The last two body parts listed were extracted from letters that became especially laced with various vulgar ad hominems between the two litigants. Note also that the two last terms have both the presumptive animate and inanimate forms. The inanimate forms appear to be used as proxy pronouns for the addressee (implying impotence), while the animate forms appear to be proxy pronouns for the addresser, especially in passages threatening physical violence.

Number

Number exhibits a two-way distinction in nouns: singular and non-singular. Singular number, or more accurately the Singulative, is explicitly marked with the determiner han/ha (the latter considered by Iyyaħmi as either a prefixe or clitic), while plural and collectives are unmarked. Nouns beginning with a vowel are preceded by the allomorph han, while ha appears before nouns beginning with a consonant.

Sometimes the singulative is represented by inflection, particularly body parts. Number distinction in body part nouns displays an unusual split, one being the typical singulative-unmarked distinction, and the other being a plural-unmarked distinction. Body parts that usually come in pairs (e.g. hands), mass/collective (e.g. hair), or multiple instances (e.g. fingers) follow the singulative/non-singulative distinction. Those that occur (usually) as a single instance on the human body, such as the head, tongue, and nose follow a singular/plural distinction, with the singular being the unmarked condition. The last body part in the table follows a singulative-plural-unmarked distinction, a highly divergent pattern combined with possessing both an animate/inanimate distinction in the singulative form. How these forms arose remains speculative, but if the texts between the litigants is any guide, the form chosen was employed to intensify some of their more vitriolic correspondence, the animate singulative or plural form serving as the proxy noun of the speech transmitter, and the inanimate singulative or simple unmarked form serving as the proxy noun of the speech recipient. Regardless, the usual singulative determiner han/ha does not co-occur with singular body part nouns in any of the texts of the Scriptum, suggesting that double-marking with the determiner is ungrammatical.

Case

Six postpositional morphemes have been identified. They were originally classified by Schumann as clitics, whereas Iyyaħmi has convincingly argued for their status as affixes, based on statistical frequencies of certain anomalies in the litigants' use of the Minhast orthography. This article follows Iyyaħmi's classification, as this has now become the prevailing position of Vadists today:

Postpositional Affixes

Case	Postposition
Genitive	-na
Dative-Benefactive	-li
Ablative	-ta
Locative	-éva
Instrumental	-eta
Comitative	-kalí

Body part nouns may precede a head noun already marked with one of the case clitics to add more locational, directional, or positional precision. The singulative determiner does not surface before a singular body part noun, as would be expected. The following example illustrates this type of structure. Here, the noun kilái "heart", often combines with a noun marked with the locative clitic -eva to convey an inessive sense.

Junyé:ne gláy hagleva mana, ukanyen!/Dyulnheyne gu~kilay ha~gu~klebfe man ukan~ya~hen!/Julanáina gu kilái ha kileva mana, ukan hen! (Schumann)
/d͡zunɲeɪnɛ glaɪ 'haglɛv man u'kãɲɛn/ (Tashunka, Iyyaħmi), /d͡zula'naɪna gu kɪ'laɪ ha kɪlɛva mana 'ukan hɛn/ (Schumann)
ji-ula-nai-na kilái ha=kil-ev man ukan=hen (Tashunka, Iyyaħmi), ji-úla-nai-na gu kilái ha kil-eva mana ukan hen (Schumann)
1S.NOM-2S.ACC-FUT-GEN heart SGV=house-LOC go-EMPH (Tashunka, Iyyaħmi), 1S.NOM-2S.ACC-FUT-GEN [PART] heart SGV house-LOC still go wreak.havok EMPH (Schumann)
I will enter your house and still ruin you! (Tashunka, Iyyaħmi), I will enter your [?] house and still ruin you! (Schumann)

With the exception of the portmanteau pronouns, core arguments take no marking in transitive clauses. When an explicit noun appears in transitive clauses, strict SOV order is followed:

1) Jiné: gzare yafts̺ / Dyiney gu~kisare~a yabfiti~sš / Jinái gu kusare a yaviti (Schumann)³
/d͡zɪ'neɪ: gzarʌ jafts̺/ (Tashunka, Iyyaħmi), /d͡zɪ'naɪ: gu kusare a 'javɪti/ (Schumann)
ji-nai gzare yafts̺ (Tashunka, Iyyaħmi), ji-nai gu kusare a yaviti (Schumann)
1S-FUT meat buy (Tashunka, Iyyaħmi), 1S-FUT [PART] meat [PART] buy (Schumann)
I shall buy meat. (Tashunka, Iyyaħmi), I shall [?] buy [?] meat. (Schumann)

Tense markers, when they cliticize, do so only with pronominal patient arguments. As this explicitly marks the direct object, word order can vary without ambiguity. In sentences 2b and 2c, the past tense marker hai has cliticized to the first person singular ji. As the patient is now explicitly marked, it can now appear after the subject (its default position), or it may precede it. Compare these to 2a where no cliticization has taken place:

2a) Ávaron jiɬe / Ávaron dyi bifslei / Ávaron ji vasáli hai
/'avaɾon 'd͡ʒɪɬɛɪ/ (Tashunka, Iyyaħmi) /'avaɾon d͡ʒi va'sali haɪ/ (Schumann)
ávaron ji-vasáli=hai / ávaron ji vasáli hai
PN 1S-visit=PST / PN 1S visit PST /
Ávaron has visited me.

2b) Ávaron dyai bfasáli / Ávaron jai vasáli
'avaron d͡ʒaɪ va'saɬ / 'avaron d͡ʒaɪ va'sali
ávaron ji-hai vasáli
PN 1S.DO-PST visit
Ávaron has visited me.

2c) Jai Ávaron vasáli.
d͡ʒaɪ 'avaron va'sali
ji-hai ávaron vasáli
1S.DO-PST PN visit
Ávaron has visited me.

An object may also come before the subject if the object is topicalized by the particle ipan, which is also followed by an audible pause:

Kusar fisl~pan, yabfiti~sš nai (Iyyaħmi) / Kusara visil ipán, ji yaviti nai (Schumann)
kusar ɪɬ'pan jafts̺ naɪ (Iyyaħmi) / kusara visil i:'pan d͡ʒi ja'vi:ti naɪ (Schumann)
kusar iɬpan yafts̺ nai (Iyyaħmi) / kusara visil ipán ji yaviti nai (Schumann)
meat TOP 1S buy FUT (Iyyaħmi) / meat PART TOP 1S buy FUT (Schumann)
Meat, I shall buy (some) / Meat [?], I will buy (some)

Case Chaining

A specific form of agglugination seen in Vadi nouns is case chaining, whereby two or more markers become conjoined to each other. This process is used to convey more granular theta-role information of the host noun. This is seen most often in possessive phrases, whereby the possessor noun undergoes clitic/affix chaining to specify the semantic role of its head, and oftentimes to indicate tense, as in the third sentence:

Jinne val ippela ujãdé, jutavárana skuvan (Tashunka) / Jilina valin pela udahadi hai, jutavárana iskuvan (Schumann)
ji-li-na valin pela ujãdi=hai jutavara=na skuvan / ji-li-na valin pela udahadi hai jutavara-na iskuvan
1S-DAT-GEN family cloth buy PST winter.clothing-GEN PURP
For my family I bought cloth to make winter clothing.

Dye<bif>dan gu~kilái handorim ésai, kunniri ajul skuvan (Iyyaħmi) / Jevdan glai handorim essai, kundiri-ahalin ajul skuvan (Tashunka) / Jinaevata kilái han odorim hesa hai, kuniri-ahalin ajul iskuvan (Schumann)
ji-eva-ta-na kilái han=odorim hesa=hai kuniri-ahalin ajul iskuvan
1S-ABL-LOC-GEN heart SGV house coin.PL bring=PST disk-payment iron PURP
From inside my house I brought out money to pay for the iron ingot.

Dyulnheyne gu~kilay ha~gu~klebfe man ukan~hen! (Iyyaħmi) / Julné:na glay haglev man ukɑ̃nyen! (Tashunka) / Julanáina gu kilái ha kileva mana, ukan hen! (Schumann)
/d͡zunɲeɪnɛ glaɪ 'haglɛv man u'kãɲɛn/ (Iyyaħmi, Tashunka) /d͡zula'naɪna gu kɪ'laɪ ha kɪlɛva mana 'ukan hɛn/
ji-ula-nai-na guklái ha=gukl-ebfe man ukan-hen / ji-úla-nai-na gu kilái ha kil-eva mana ukan hen
1S.NOM-2S.ACC-FUT-GEN house SGV=house-LOC go-EMPH / 1S.NOM-2S.ACC-FUT-GEN [PART] heart SGV house-LOC still go wreak.havok EMPH
I will enter your house and still ruin you! / I will enter your [?] house and still ruin you!

Pronouns

Vadi pronouns do not mark for gender. Case marking on core arguments for transitive clauses do exhibit a nominative-accusative distinction, but only when both arguments are singular in number.

Plurality in the independent pronouns are indicated by the suffix -ta. This contrasts with number marking on nouns, where plurals are unmarked but singulars are preceded by the determiner han/ha.

	Singular	Plural
First	ji	jita
Second	úla	ulta
Third	ka	kata

The nominative-accusative distinction occurs in the portmanteau forms, which are etymologically transparent. The first segment of the portmanteau affix marks the nominative argument, while the second marks the accusative argument:

Agent			Patient
	Singular
		Person	Singular			Plural
			First	Second	Third	First	Second	Third
		First	--	júla	jíka	--	julta	jikata
		Second	uláji	--	uláka	ulajta	--	ulkata
		Third	kaji	kúla	káka	katáji	kulta	katá
	Plural
		First	--	jitúla	jitakta	--	jitulta	jitkata
		Second	ultaji	--	ultakta	ultajta	--	ultakta
		Third	kataji	katúla	katáta	katajta	katulta	kakta

Affix/clitic chaining can occur with all pronominal forms, as in one of the sample sentences in the previous section, demonstrated again here:

Julanáina gu kilái ha kileva mana, ukan hen!/Dyulnheyne gu~kilay ha~gu~klebfe man ukan~hen!/Julné:na glay haglev man ukɑ̃nyen
/d͡zula'naɪna gu kɪ'laɪ ha kɪlɛva mana 'ukan hɛn/ -- /d͡zunɲeɪnɛ glaɪ 'haglɛv man u'kãɲɛn/
ji-úla-nai-na gu kilái ha kil-eva mana ukan hen / ji-ula-nai-na guklái ha=gukl-ebfe man ukan-hen
1S.NOM-2S.ACC-FUT-GEN [PART] heart SGV house-LOC still go wreak.havok EMPH/1S.NOM-2S.ACC-FUT-GEN house SGV=house-LOC go-EMPH
I will enter your [?] house and still ruin you!/I will enter your house and still ruin you!

Adjectives

Adjectives in Vadi precede their heads. In terms of morphology, there are no special affixes that distinguish them from either nouns or verbs. Iyyaħmi believes they may trigger mutation, but admits that further analysis is needed. Interestingly, and unfortunately, the Scriptum contain far fewer adjectives than one would expect in such a significant corpora, and the ones that do occur are mostly in the forms of insults found in the Waškixrapmakirimērumbustikmaban. Unfortunately, even the texts in the Waškixrapmakirimērumbustikmaban contain relatively few adjectives, not enough for Iyyaħmi to definitively say that a mutation has occurred or not. Moreover, the adjectives that are found occur in phonemic environments that are unlikely to trigger sound changes. One such example from the corpora, cited by Iyyaħmi, follows:

Ûla erva vûluka ran
/'u:la erva 'vu:luka ran/
úla erva vûlu-ka ran
2S small testicle-INAN.SGV man
You have no ba**s! (lit. "You are a small one-testicled man")

The adjective erva "small" both begins and ends with a vowel, as seen from the Ammerkast and the original Širkattarnaft text. No mutation, vowel syncope, or other sandhi of úla "you" is indicated. While vowels usually lenite a stop to a fricative, they usually have no effect on a pre-existing fricative in the following word, unless the word has the determiner ha= cliticized to it; such a case would trigger either palatalization and possibly syncope as in *erva yulka⁴ << erva ha=vûluka.

Vadi has no copula. To express the copula, the adjective is placed at the end of the clause, preceding any tense markers that may be present. The next example, also from Iyyaħmi, demonstrates this:

Danar kúda hai imu Sorvin ga~kamatkamat.
/'danar 'ku:da haɪ 'imu 'sorvin 'gamatkamat/
danar kúda hai, imu sorvin gamat-kamat
place.name beautiful PST until PN burn-burn
Danar was a beautiful place until Sorvin utterly torched it.

Again, neither the Širkattarnaft or the Ammerkast transcriptions indicate that the adjective kuda "beautiful" has undergone or triggered any phonological changes. The phonemes /ɾ/ + /k/ do not undergo any changes when adjacent to each other, and kuda does not lenite to /'kud͡ʒɛɪ/ even though the past tense clitic hai appears after it, since a final /a/ prevents lenition if it was etymologically derived from /aʔ/ or /aɦ/, as it appears to be the case with kúda << */'ku:dðaʔ/

Verbs

Vadi verbs for the most part contain no grammatical inflection. They do not mark for person, number, tense, or aspect. Few, if any clitics attach to them. Any affixes that appear on the verb are predominately derivational, most of which are deverbals to derive instrumental, locational, and occupational nouns. Examples with the verb kora "to cut":

• -vir: instrumental deverbal, e.g. kora-vir "knife"
• -ðai: locational deverbal, e.g. kor-ðai "butcher shop"
• -ji: occupational deverbal, e.g. kor-ji "butcher"

The language has a variety of converbs which appear after their head. These converbs encode modality, aspect, manner, and various adverbial meanings. The converbs may be separated from their head by other elements, such as illocutionary particles, discourse markers, etc. Common converbs include:

	Meaning	Undergoes Mutation				Triggers Mutation		Header #3	Header #4
Independent Form		Undergoes Mutation	Mutation Type	Triggers	Result	Colspan + Rowspan
sútu	"and; additionally; to add"; a direct loan from Minhast suttu "and"	X	Voicing Palatalization	hen cliticized to noun in the Past-Genitive	-zut -šut	rowspan-	d
Row #2:	s-	v-	m-	jor-

• sutu, sútu: "and; additionally; to add"; a direct loan from Minhast suttu "and"
• ura: "again; to repeat"
• koni: "to start, begin"
• kila: "to stop, cease"
• helas: "to attempt, to try to do something"
• ila: "immediately; to do quickly"
• ókapi: "suspiciously; to do with malfeasence; to be deceptive"
• kitál: assertive particle, usually translated as "You did do it!"
• yoawíka/yawikâ /jo'ka:ʔ/ (Iyyaħmi): presentational particle, usually translated as "Look here (at the evidence I present you)"; possibly a loan from Minhast wahēk "behold; thus; and then"
• oan: evidential, lit. "It is said", but what the speaker really means is "He (the other litigant) falsely claims"
• peliár: "to be wrong"; used for repudiation, disputation, e.g. "What he (the other litigant) says is a lie/falsehood"
• oni: "to be empty", in the context of the lawsuit, "There is no evidence that..."
• neko: "to continue; still", in the context of the lawsuit, re-establishes a statement the litigant said that has been agreed by the Prefect to be true, setting up support for the next statement the litigant is going to say next, e.g. Neko Ápan jibái Yéso jibáitane!, i.e. "Just as was established (by the Prefect earlier) that Ápan is mine, so is Yéso!"
• penór: privative, "to undo; to wash away", e.g. Penór ójokan "to dismiss, disregard" (lit. "to un-hear"); in the context of the lawsuit, "to dismiss the charges (against me)", e.g. Neko hino jikátahai sonda, Ikúni mek Sorvinna penór mek piláhi!, i.e. "Just as this lowly one has presented thee evidence and thou dost agree, I beseech thee, Lord, that thou dismissest Sorvin's accusations (against me)!" Piláhi means "to accuse", so in this sentence, the litigant Éro is literally saying, "...un-accuse me!"

In most instances, converbs can stand alone as bona-fide verbs, e.g. Úla ókapi hai hen! "You did lie!"

As the language's canonical word order is SOV, the verb phrase occupies the final position of the clause, with the tense markers nai and hai appear at the very end of the clause.

Particles

• nai: past tense particle
• hai: future tense particle
• hen: emphatic particle
• ipan: topic particle
• ani: "yes"
• yal: "no"
• ka: "if"
• ki: "well then, so"
• mi: hortative particle
• bia: "behold"
• kunet: interrogative particle
• vali: conjecture particle, "perhaps"
• puni: resultative particle
• imu: until
• mek : "please" (borrowed from Minhast mek, a particle that softens not just commands, but statements, or simply to convey deference)

Derivation

Derivational affixes occur in greater frequency, but the texts from the Scriptum suggest the overwhelming number of these affixes are no longer productive. For this reason, the general consensus among Vadists is that these affixes have been fossilized. A prefix pesa- occurs among some verbs that tends to give them causative meaning. For this reason some have speculated this is a borrowing from the Minhast causative prefix -šp-. This view is problematic however, because this prefix sometimes appears to intensify the meaning of the root. Moreover, it is also found attached to some nouns, but its addition does not appear to affect the semantics of the noun.

Causative pesa-:

Anu úla pesadíka nai, hen. /Annu úla pesátkanen.
Anu úla pesa-díka nai hen
PN 2P CAUS-run.off FUT EMPH
(Prefect) Annu will drive you out!

Intensive pesa-:

Ji pesakúna, kaman uláta, ji nokan pesakúdi. / Ji pesakúna, kamallúta, jonkan peskúd.
ji pesa-kúna nai, kaman uláta ji nokan pesa-akúdi.
1P CAUS-go FUT head úla=ABL 2S rock CAUS-throw.
I will climb (that mountain) above you (so that) I can rain down rocks (upon you).

Nominal pesa-:

Pesarona hokun.
pesa-rona hokun
CAUS-snake eat.
Snakes will eat (you).

Contrast this with the use of the actual Minhast causative -šp-, by Prefect Annu (Minhast, Dog Speaker Dialect):

Tašpintaknataheknessuš, tašpintaknaknessuš. Marentaħmankilmakš, yattax! Ikšitamaškidustitaħmāš!
ta-šp-nt-ikna-tahek-ness-u=š ta-šp-nt-ikna-k-ness-u=š maren-tahem-an-kilmakš yattax kš-tamašk-dust-tahem-an=š
NEG-CAUS-INT-go-2S.ACC+1S.NOM-FUT-TRNS=IRR NEG-CAUS-go-FUT-TRNS=IRR pest-2P.NOM-MIR.NEG DEPR CESS-stalk.while.hunting-REC.ADVS-2P.NOM-INTR=IRR
I don't plan on throwing you off your land, and I'm not planning on throwing him off his land. You two pests should leave each other alone!

Other derivational affixes, both prefixes and suffixes, are more common, some appear to be highly productive and suggest they can be spontaneously generated:

• -kai: a deverbal for creating place nouns, e.g. iavati-kai "a place for buying", i.e. marketplace.
• -pallái: another deverbal for creating place nouns, e.g. tukin-pallái "a place for sitting", i.e. hearth, dining room.
• -kattá: a deverbal affix deriving agent nouns, e.g. iavati-kattá "one who buys, a buyer"

Some affixes may be chained, as in tukimpallaikattá < tukin-pallai-kattá "a place for sitting ones", i.e. "dinner guests, diners". However, affix chaining is rather uncommon.

Syntax

Vadi's canonical SOV word order is used to indicate the case roles of the core arguments of a clause, to delineate the constituents of a noun phrase, and to indicate the pivot in multiclausal sentences. Although the language is primarily SOV, the verb phrase may occur in different locations in the clause for pragmatic purposes. However, if both the subject and object arguments are expressed as overt nouns, the S-argument invariably precedes the O-argument in transitive clauses. Tense particles typically occur in clause-final position, but as discussed previously, in transitive clauses they can and often move from clause-final position to cliticize to an object pronoun.

Noun phrase

Possessums follow their heads, which serve as host to the genitive affix, and any other affixes that indicate the head noun's theta role and/or tense-marking clitics.

Adjectives typically follow their head nouns. However, in possessive constructions, whenever an adjective modifies the possessor, it must come immediately before the possessor, otherwise its default position means it is modifying the possessum. As for adverbs, they appear after the adjectival phrase they have scope over. These features of adjectives following their head nouns and adverbs following the adjectival phrase are more typical of VSO languages. It is thus hypothesized by some linguists that Vadi went through a stage in its evolution as a VSO language, but this remains speculative among most Vadists.

Example texts

Ad-Hominems and Other Insults

The majority of the materials found in the Scriptum are known as the Waškixrapmakirimērumbustikmaban lit. "All they do is insult each other incessantly", a Minhast term that describes the nature of these letters accurately, i.e. correspondence containing mostly ad-hominems, threats, taunts, and other insults. Many descend to the level of childishness, but this characteristic is probably the most valuable part of the collection, as the style appears to be in vernacular Vadi, with fewer intrusions from the dominant Minhast language found in the legal correspondence.

Ulájinai tábila nikku osar, petta ulátane
/u:'lad͡ʒnaɪ 'ta:bɪla nɪk:u osaɾ pet:a u'la:tane/
úla-ji-na-hai tábila nikku osar petta úla-tane
2S.NOM-1S.ACC-GEN-PST land.PL seize try, thief 2S-EMP
You tried to seize my lands, you are a thief indeed!

Ka mai naha úlajnaina tábila é, júlanai puni mérkeva!
/ka maɪ 'naha 'u:lad͡ʒnaɪna 'ta:bɪla e 'd͡ʒu:lanaɪ 'pu:ni 'mɛɾkɛva/
ka mai naha úla-ji=nai-na tábila nikku é ji-úla-nai puni mérkeva
if come here 2S.NOM-1S.ACC=FUT-GEN land.PL seize SJV, 1S.NOM-2S.ACC-FUT RSLT kill
If you come here to seize my land, I will kill you! (lit. If come here you-me-my-land to seize it, I-you will kill)

Júla nánani.
/d͡ʒ'u:la 'na:nani/
ji-ula na~nani
1S.NOM-2S.ACC RV~disgust
You truly disgust me.

Valí ulaki úla píhala nai.
/va'li u'laki 'u:la 'pihala naɪ/
valí ulaki úla píhala nai
perhaps authorities 2S seize FUT
Perhaps the authorities shall arrest you.

Dávan biri ulájai kunet? Dumúla rani ikkúni siráha ji?
/'da:van biɾi ul'a:d͡ʒaɪ kunet du'mu:la ɾani ik:'u:ni siɾ'a:ha d͡ʒi/
dávan biri ulá-ji-hai kunet dum-úla rani ikkúni siráha ji
why thus 2S.NOM-1S.ACC-PST commit.wrong Q-2S know prefect stand.behind 1S
Why have you wronged me thus? Do you not know the Prefect has given me his support?

Súnagun jibáitane! Ji súharak kanéneka, báti úla kimúdi, bia!
/'su:nagun d͡ʒi'baɪtane d͡ʒi 'su:haɾak ka'ne:neka 'ba:ti 'u:la: ki'm:udi bija/
súnagun ji-bai-tane ji súharak kanéneka báti ulá kimúdi bia
place.name 1S-ownership-EMP 1S papers hold.in.hand PART 2S read behold
Súnagun is mine! I have the (legal) papers, [?] here, take them and read!

Éro, úla peho te. Júlanai kanarídi. Ikkúni jili Súnagun parati nai.
/'e:ɾo 'u:la pɛho te 'd͡ʒu:lanaɪ kana'ɾidi ɪ'k:u:ni 'd͡ʒili 'sunagun pa'rati naɪ/
éro úla peho te júlanai kanarídi ikkúni ji-li Súnagun parati nai
PN 2S fool truth 1S.NOM-2S.ACC win FUT prefect 1S=DAT place.name reward FUT
Éro, you indeed are a fool. I will prevail over you. The Prefect shall award me Súnagun.

Óşi jínai kanárato bihók şilmá.
/'o:s̺i 'd͡ʒɪnaɪ kana:ɾato bɪhok 's̺ɪl'ma/
Ósi jí-nai kanárato bihók şilmá
yes 1S.NOM-2S.ACC succeed indeed today
Of course I shall prevail (over you) today

Manetór ki pida dáşelek mi, mánjate nihálave yal.
/manɛ'toɾ ki 'pida 'das̺ɛlɛk mi 'mand͡ʒate ni'halavɛ yal/
manetór ki pida dáşelek mi mánjate nihálave yal
fire well.then fall.down heaven HORT anyone care NEG
Well then, let fire fall down (upon you), no one will care.

Isáro ka munek ki bára jikalí, júla emtol. Ulánai paho yal.
/ɪs'aɾo ka 'munɛk ki 'baɾa d͡ʒɪka'li 'd͡ʒu:la 'ɛmtol u'la:naɪ 'paho yal/
isáro ka munek ki bára ji-kalí ji-úla emtol úla-nai paho yal
choose COND call.upon well.then fist 1S-INST 1S.NOM-2S.ACC thwart 2S-FUT win NEG
If you choose to call up (the townspeople) against me as I expect you will, I will stop you. You will fail.

Legal Claims

In contrast, the Legal Claims parchments (Upper Minhast: Yukan Mukan, c.f. Gull Speaker Yuk sut Muk, lit. "Complaint and Judgment") use more formal language and lack the colourful language of the Waškixrapmakirimērumbustikmaban. Minhast words, particularly legal terminology, units of measurement, and other terms requiring precision especially found in lawsuits involving land disputes, are found here. Especially prevalent is the formulaic Minhast phrase: Mek wa [proper noun]-behet širkattartahannamā, "Please write to (my lord X) these words".

It is believed that each litigant wrote these letters carefully in order to have them translated to the Prefect, the surveyors, or to ingratiate themselves to other Dog Speakers who could serve as witnesses in the lawsuit. In at least one case one of the litigants even tried to petition the powerful Salmon Speaker Prefect, Nattaxxawan.

• From Éro to Wasiškila for translating to Ruxparan min Ayyurmi, the headman of Niyyurpi Township:

Mek wa Ruxparan-behet širkattartahannamā: Ulá: hano tšmáka jileykandá. Anuna beiólika pisap ey. Hano tšalt han Garbaratana kalaví Íkavalina menehim sentum ohira julór, ja han zitturum Sarpahé:va rwani duxtim sentum julór. Kon, havvá:ta jóri, ha Bikuévana kippi kaħtam min sentum avóro alo ngad. Bari, han Ikúni hajjina pétetaité: hadábla laváké:, hajjina tširosi umukálé:.
{Mek wa Ruxparan-behet širkattartahannamā} úla han-o temáka jilái-kanda Anu-na beiólika pisap hai han-o talta ha Karbara-ta-na kalaví han Íkava-li-na menehim sentum ohira julór, hi-ya han síturum Sarpahei-eva oroani duxtim sentum julór kon ha bihata jóri ha Piku-éva-na kippi kaħtam min sentum alo-k nad bari han Ikkúni han ji-na pétetaiti hai ha tábila lavák hai han jina tirosi umukál hai
{Embedded Minhast text: [deferential.particle CONN=PN=title.of.respect write to him thus]} 2S SGV-PROX map look-IMP PN-GEN surveyor.PL draw PST SGV-PROX line SGV place.name-ABL-GEN rock-GEN river.name-ALL-GEN shore 4 sentum stretch then-EPEN SGV beaver.dam place.name-LOC 3 sentum cross.over stretch from.that.place.there SGV ridge climb 5 sentum again SGV PN-ABL-GEN farmhouse burn-DVB NMZ thus SGV prefect SGV 1S-GEN viewpoint SGV agree PS SGV land recognize SGV 1S-GEN control give PST
Please write to Wise Elder Ruxparan thus: Look at this map. This is what Prefect Annu's surveyors have drawn. This line from Karbaram Rock stretches to the shoreline of River Íkava for 4 sentum (~6 miles), then extends across the old beaver dam at Sarpahey Spring for another 3 sentum (~5 miles). Thence it curves upwards for another 5 sentum (~9 miles) along the ridge where old Piku's farmhouse was before it burnt down. Thus the Prefect has recognized my territorial claim as the sole legitimate one.

The next letter, from Sorvin to Prefect Nattaxxawan, is a rare find, as it contained many archaisms not found in any of the Waškixrapmakirimērumbustikmaban or the majority of the Yukan Mukan for that matter.

• From Sorvin to Ittahampar for translating to Prefect Nattaxxawan, of Salmon Speaker Country:

Mek wa Ikkūne Nattaxxawan-behet širkattartahannamā: Tódimtana galua jinda pattuk.
{Mek wa Ikkūne Nattaxxawan-behet širkattartahannamā}

Please write to Wise Lord Nattaxxawan: This lowly one places himself before the Great Hero's feet and kisses the ground upon which he steps. This lowly one is not worthy, for thou art the most high and esteemed Lord of all of Salmon Speaker Country. Yet this lowly one beseeches thee and ask thee but only give a light word to the most respected Prefect of all of Dog Speaker Country, the great Prefect Ruxparan under which this lowly one places himself under the Great Lord's feet.

OVS -> matrix clause SC -> dependent clause

SVO -> matrix clause, COND SOV -> dependent clause, RSLT

OSV

Footnotes

1) Šibbūru is an Irish Gaelic borrowing into Minhast, derived from Irish séimhiú "lenition", and urú "eclipsis". This loanword was imported by Minhast Vadists to represent their view that mutations were a major feature of Vadi affecting both phonology and morphology. This is in diametrical opposition of the Traditionalist school, which maintains that Vadi phonology and phonotactics do not exhibit "exotic pseudo-Celtic" characteristics.

An alternative name to Šibbūru, also used often, is the indigenous Upper Minhast term, Saxtakyatan Saxtidammafan, lit. "It becomes both hard and soft", i.e. fortition and lenition.

1) Much of Schumann's objections to Iyyaħmi's findings is that the spelling anomaly categories the latter assembled were based on conjecture and that Iyyaħmi fitted his conclusions to the sample. A major counter-criticism to Schumann's objections, notably from Dr. Tashunka, is that although Schumann is "competent" in the Širkattarnaft, he lacks a full understanding of its development and evolution, and how it was historically used by other non-Minhast minority speakers, namely the Peshpeg and Ín Duári, to transcribe their respective languages; the Širkattarnaft in the two speech communities also show the use of digraphs, unorthodox usage of the characters, and other anomalies in spelling. It has been argued that Schumann remains committed to the traditional model of Vadi phonology and phonotactics, and that he has failed to provide an alternate model that would explain the discrepancies more effectively than Iyyaħmi. While it can be argued that Schumann has an interest in maintaining the traditional model because an alternative model would invalidate his previous work, it must also be recognized that Iyyaħmi studied under Tashunka, who would understandably side with his prodigy.

2) Schumann argues that gemination outside of Minhast loanwords does not occur in Vadi, as the authors' Širkattarnaft does not show any gemination. This argument, as Tashunka notes, is problematic in that most Minhast writings, both past and present, rarely indicate gemination. Evidence of gemination in Vadi surfaces in the orthography between morpheme boundaries where the vowel of the syllable preceding the geminate consonant is lengthened, and the following syllable or an inserted "dummy syllable" starts with a voiced consonant to indicate fortition. Schumann argues this is partial reduplication used for derivation, but Iyyaħmi concurs with Tashunka's analysis.

3) This particular gloss is an excellent example of how the divergence between Schumann and Iyyaħmi's orthographic systems can lead to different conclusions of the underlying grammatical structure.

In the original Širkattarnaft of this sample, Iyyaħmi has indicated, per his system employing tildes, the mutations that occurred between the standalone characters and the word in which they trigger sandhi changes. In this case, the word Dyiney triggers kusarʌ to undergo voicing of the initial consonant as indicated by the tilde in gu~ joined to kusarʌ. Syncope also cooccurs with the elision of the /i/ of kilay. Iyyaħmi's gloss yields the actual phonemic realization /gzarʌ/.

In contrast, Schumann has glossed the original Širkattarnaft grapheme <gu> as a word, which has led his final morphemic gloss of <gu> as a word or particle with as-of-yet unknown meaning. Iyyaħmi's representation however conveys the sandhi process kusarʌ undergoes. As a result, Iyyaħmi's transcription yields two fewer words than Schumann's, the very two that Schumann has glossed as independent words. Iyyaħmi's gloss ultimately shows there are no unknown, independent words in the ligigant's text and avoids Schumann's erroneous conclusion of the <gu> grapheme as a separate morpheme.

As expected, the first line of Tashunka's gloss indicates none of the underlying mutations, only the pronunciation of the final textual realization. Mutations and other sandhi changes can be gleaned only via his morphemic breakdowns, which again closely follow Iyyaħmi's analysis.

4) Iyyaħmi's reconstruction of erva yulka is based on a sentence from one of Éro's texts:

Ikúni behet mek, yawikâ ha yi~bfita mek hen
/i'ku:ni bɛhɛt mɛk jo'ka: ha'jita mɛkjɛn/
Ikúni behet mek, yawkâ ha=yita mek hen
lord VOC.DEFR please here SGV paper please EMPH
My lord, I beseech thee, behold the (legal) documents (in my hand).

Here, the singulative clitic ha= triggers palatalization of the noun vita to yita "paper" (from yi bfita). It is from this example that Iyyaħmi extrapolates the mutation of the adjective phrase (AP).