Verb-initial
# grammars: A multilingual/parallel perspective
# ESRC Project RES-000-23-0505
# Oxford University
# Charles Randriamasimanana
# Malagasy syntax & semantics:
# Malagasy simple sentences Case-marking & NPs: ESRC-OX-05-CR101/201.
# My purpose here is to show how case-marking works in Malagasy, in the process distinguishing
# between three different types of nouns depending on the nature of the phrasal head, i.e. a D
# accompanying a proper name(PN), a Det accompanying a common noun (CN) or a DX, which can go
# with either PN or CN. In the second half of this document, it will be seen that a verb that
# assigns accusative case does not necessarily undergo passive.
# Three main types of noun:
# A. Proper name (PN) = bare NP headed by a D; pattern A;
# B. Common noun (CN) = optionally headed by a Det thus forming a DetP,
# i.e Definite phrase; pattern B;
# C. Either PN or CN headed by DX; pattern C.
# Case-marking pattern:
# Pattern A. With bare NPs with D, i.e. PN
# Nominative = Accusative = an.
# Pattern B. With DetP no accusative marker an at all possible.
# Pattern C. With DX previous mention ilay, accusative marker an
# is optional.
# Three types of determiner:
# Type D = { i 'intimate + proper name', ry 'proper name & associates',
# Ra... 'familiar + proper name', Ikala 'intimate + female proper name',
# Ingahy 'familiar + male proper name', Ramatoa 'Mrs.', Ramatoakely 'Miss',
# Andriamatoa 'The honorable Mr...', Itompokovavy 'the late + intimate + female
# proper name', Itompokolahy 'the late + intimate + male proper name', ... }.
# Type Det = ny 'the entire set (of entities)'.
# Type DX = { ilay 'the previous mention', ity 'this.singular.near.speaker',
# ireto 'these.plural.near.speaker', iny 'this.singular.near.hearer', ireny 'these.
# plural.near.hearer', iry 'this.visible.far', irery 'these.plural.visible.far',
# itsy 'this.singular.to.the.side', iretsy 'these.plural.to.the.side',
# iro ''this.singular.visible', irero 'these.plural.visible', ...}.
# Typical traditional proper names = {i Koto, Ikoto, Rakoto; i Naivo,
# Ranaivo, i Soa, Rasoa; IKala, Rakala; i Ketaka, Iketaka, Raketaka;
# i Tefy, Ratefy; i Vololona, Ravololona}.
# Apparent exceptions behave as expected, i.e. as PN following pattern A
# above: { NyAvana, NyOny, NyAina,}
# F 1: Testfile for Case in simple Malagasy sentences.
# However see File 3 below for Case in sentences involving complex predicates as many verbs
# appear to be superficially simple and yet involve rather complex morphology.
# General pattern A: Nominative case-marker but D is obligatory.
# Ungrammatical
Bakoly ve no manapaka bozaka? (0! 0 0 0)
# Bakoly ve no m- an-(t)apaka bozaka?
# Bakoly Q focus pres-prf.an-root.tapaka grass
# "Is it Bakoly who is cutting grass?"
# This sequence is adapted from Ileana Paul (2000: 177, ex. 48).
# Note the absence of D.sg 'i' in front of 'Bakoly'.
I Bakoly ve no manapaka bozaka?
# I Bakoly ve no m- an-(t)apaka bozaka?
# D.sg Bakoly Q focus pres-prf.an-root.tapaka grass
# "Is it Bakoly who is cutting grass?"
# Note the presence of D.sg 'i' in front of 'Bakoly'.
# Ungrammatical
Omaly ve Bakoly no manapaka bozaka? (0! 0 0 0)
# Omaly ve Bakoly no m- an-(t)apaka bozaka?
# yesterday Q Bakoly focus pres-prf.an-root.tapaka grass
# "Was it yesterday that Bakoly was cutting grass?"
# This sequence is adapted from Ileana Paul (2000: 178, ex. 49).
# Note the absence of D.sg 'i' in front of 'Bakoly'.
Omaly ve i Bakoly no nanapaka bozaka?
# Omaly ve i Bakoly no n-an- (t)apaka bozaka?
# yesterday Q D.sg Bakoly focus past-prf.an-root.tapaka grass
# "Was it yesterday that Bakoly was cutting grass?"
# Note the presence of D.sg 'i' in front of 'Bakoly'.
# Ungrammatical
Bakoly ve dia manapaka bozaka? (0! 0 0 0)
# Bakoly ve dia m- an-(t)apaka bozaka?
# Bakoly Q topic pres-prf.an-root.tapaka grass
# "Bakoly, is she cutting grass?"
# This sequence is adapted from Ileana Paul (2000: 178, ex. 50).
# Note the absence of D.sg 'i' in front of 'Bakoly'.
I Bakoly ve dia manapaka bozaka?
# I Bakoly ve dia m- an-(t)apaka bozaka?
# D.sg Bakoly Q topic pres-prf.an-root.tapaka grass
# "Bakoly, is she cutting grass?"
# This sequence was adapted from Ileana Paul (2000: 178, ex. 50).
# Note the presence of D.sg 'i' in front of 'Bakoly'.
# General pattern A: Accusative-marker 'an' obligatory on D.
Nahita ani Koto i Soa.
# Nahita an i Koto i Soa.
# N-a-hita an' i Koto i Soa.
# past-pref.a-root.hita acc D.sg Koto D.sg Soa
# "Soa (intimate) saw Koto (intimate)."
# Ungrammatical
Nahita i Koto i Soa.(0! 0 0 0)
# N-a-hita i Koto i Soa.
# past-pref.a-root.hita D.sg Koto D.sg Soa.
# "Soa (intimate) saw Koto (intimate)."
Nahita an-dry Koto ry Soa.
# N-a-hita an- d-ry Koto ry Soa.
# past-pref.a-root.hita acc-d-D.pl Koto D.pl Soa
# "Soa & associates saw Koto & associates."
# Note the presence of the accusative particle 'an'.
# Note the epenthetic d between 'an' and D.plur 'ry.'
# Ungrammatical
Nahita ry Koto ry Soa.(0! 0 0 0)
# N-a-hita ry Koto ry Soa.
# past-pref.a-root.hita D.pl Koto D.pl Soa
# "Soa & associates saw Koto & associates."
# Note the absence of the accusative particle 'an'.
Nahita an-dRaKoto Rasoa.
# N-a-hita an- d-Ra- koto Ra-soa.
# past-pref.a-root.hita acc-d-D.sg koto D.sg-soa
# "Rasoa (familiar) saw Rakoto (familiar)."
# Note the presence of the accusative particle 'an'.
# Note the epenthetic 'd' between 'an' & 'Rakoto'.
# Ungrammatical
Nahita RaKoto Rasoa.(0! 0 0 0)
# N-a-hita Ra-koto Ra-soa.
# past-pref.a-root.hita D.sg-koto D.sg-soa
# "Rasoa (familiar) saw Rakoto (familiar)."
# Note the absence of the accusative particle 'an'.
Nahita an'Ikala Soa i Koto.
# N-a-hita an Ikala Soa i Koto.
# past-pref.a-root.hita acc D.sg Soa D.sg Soa
# "Koto (intimate) saw Soa (very intimate, only between females)."
# Note the presence of the accusative particle 'an' in front of
# D.sg 'Ikala".
# Ungrammatical
Nahita Ikala Soa i Koto.(0! 0 0 0)
# N-a-hita Ikala Soa i Koto
# past-pref.a-root.hita D.sg Soa D.sg Koto
# "Koto (intimate) saw Soa (very intimate, only between females)."
# Note the absence of the accusative particle 'an' in front of
# D.sg 'Ikala'.
Nahita an-dRamatoa Rakala i Koto.
# N-a-hita an- d-Ramatoa Rakala i Koto.
# past-pref.a-root.hita acc-d-D.sg Rakala D.sg Koto
# "Koto (intimate) saw Mrs. Rakala (deferential)."
# Note the presence of the accusative particle 'an' in front of
# D.sg 'Ramatoa'.
# Note the epenthetic 'd' between the accusative particle 'an' and
# D.sg 'Ramatoa'.
# Ungrammatical
Nahita Ramatoa Rakala i Koto.(0! 0 0 0)
# N-a-hita Ramatoa Rakala i Koto
# past-pref.a-root.hita D.sg Rakala D.sg Koto
# "Koto (intimate) saw Mrs. Rakala (deferential)."
# Note the absence of the accusative particle 'an' in front of
# D.sg 'Ramatoa'.
Nahita anIngahy Rakoto i Soa.
# N-a-hita an' Ingahy Rakoto i Soa.
# past-pref.a-root.hita acc D.sg Rakoto D.sg Soa
# "Soa (intimate) saw Mr. Rakoto (somewhat familiar)."
# Note the presence of the accusative particle 'an' in front of
# D.sg 'Ingahy'.
# Ungrammatical
Nahita Ingahy Koto i Soa.(0! 0 0 0)
# N-a-hita Ingahy Rakoto i Soa.
# past-pref.a-root.hita D.sg Rakoto D.sg Soa
# "Soa (intimate) saw Mr. Rakoto (somewhat familiar)."
# Note the absence of the accusative particle 'an' in front of
# D.sg 'Ingahy'.
Nahita anAndriamatoa Rakoto i Soa.
# N-a-hita an' Andriamatoa Rakoto i Soa.
# past-pref.a-root.hita acc D.sg Rakoto D.sg Soa
# "Soa (intimate) saw Mr. Rakoto (deferential)."
# Note the presence of the accusative particle 'an' in front of
# D.sg 'Andriamatoa'.
# Ungrammatical
Nahita Andriamatoa Rakoto i Soa.(0! 0 0 0)
# N-a-hita Andriamatoa Rakoto i Soa.
# past-pref.a-root.hita D.sg Rakoto D.sg Soa
# "Soa (intimate) saw Mr. Rakoto (deferential)."
# Note the absence of the accusative particle 'an' in front of
# D.sg 'Andriamatoa'.
Nahita ani NyAina i NyAvana.
# N-a-hita an' i Nyaina i NyAvana.
# past-pref.a-root.hita acc D.sg Nyaina D.sg NyAvana
# "NyAvana (intimate) saw NyAina (intimate)."
# Note the presence of the accusative particle 'an' in front of
# D.sg 'i'.
# Note that 'NyAina' is a proper name, not analyzed as a combination
# of 'Ny + Aina'.
# Ungrammatical
Nahita NyAina i NyAvana.(0! 0 0 0)
# N-a-hita Nyaina i NyAvana.
# past-pref.a-root.hita Nyaina D.sg NyAvana
# "NyAvana (intimate) saw NyAina (intimate)."
# Note the absence of the accusative particle and that of
# D.sg 'i' in front of the proper name 'NyAina'.
# Ungrammatical
Nahita an'NyAina i NyAvana.(0! 0 0 0)
# N-a-hita an' Nyaina i NyAvana.
# past-pref.a-root.hita acc Nyaina D.sg NyAvana
# "NyAvana (intimate) saw NyAina (intimate)."
# Note the absence of D.sg 'i' in front of the proper name 'NyAina'.
# General pattern B: accusative-marker 'an' incompatible with Det 'ny'.
Nahita ny ankizy i Soa.
# N-a-hita ny ankizy i Soa.
# past-pref.a-root.hita Det.unspec child(ren) D.sg Soa
# "Soa (intimate) saw the (previous mention) children."
# Note the absence of the accusative case particle 'an' in front
# of Det.unspec 'ny'.
# Ungrammatical
Nahita anny ankizy i Soa.(0! 0 0 0)
# N-a-hita an ny ankizy i Soa.(0! 0 0 0)
# past-pref.a-root.hita acc Det.unspec child(ren) D.sg Soa
# "Soa (intimate) saw the (previous mention) children."
# Note the presence of the accusative case particle 'an' in front
# of Det.unspec 'ny'.
# Note that Det 'ny' is unspecified in that it refers to a set, which
# can comprise a whole set of entities (plural) or only one single
# entity (singular) in case of familiarity.
# Thus the following sequence
# Nahita ny lehilahy ny vehivavy.
# N-a-hita ny lehilahy ny vehivavy.
# past-pref.a-root.hita Det.unspec man/men Det.unspec woman/women
# "The woman saw the man." (original translation)
# This sentence is adapted from Ileana Paul (2000: 15, ex 23a.).
# It should be translated as 'The (previous mention) women saw
# the (previous mention) men."
Nahita ny lehilahy ny vehivavy.
# N-a-hita ny lehilahy ny vehivavy.
# past-pref.a-root.hita Det.unspec man/men Det.unspec woman/women
# "The woman saw the man." (original translation)
# This sentence is from Ileana Paul (2000: 15, ex 23a.).
# It should be translated as 'The (previous mention) women saw
# the (previous mention) men."
# The above sequence cannot appear discourse-initially; first
# one needs (a) to introduce all the entities comprised in
# the set in a previously explicit discourse and (b) refer to
# all members of the set.
Nahita ny alika ny zazavavy.
# N-a-hita ny alika ny zazavavy.
# past-pref.a-root.hita Det.unspec a/dog(s) Det.unspec a/girl(s)
# "The girl sees the dog." (original translation)
# This sentence is from Matthew Pearson (2001: 28, ex 19a.).
# It should be translated as 'The (previous mention) girls saw
# the (previous mention) dogs."
# The above sequence cannot appear discourse-initially; first
# one needs (a) to introduce all the entities comprised in
# the set in a previously explicit discourse and (b) refer to
# all members of the set.
# General pattern C: accusative-marker an optional on DX.
Nahita ilay ankizy i Soa.
# N-a-hita ilay ankizy i Soa.
# past-pref.a-root.hita DX.sg child(ren) D.sg Soa
# "Soa (intimate) saw the (previous mention) child."
# Note the absence of the accusative particle 'an' in front of
# DX.sg 'ilay''the previously mentioned'
Nahita anilay ankizy i Soa.
# N-a-hita an' ilay ankizy i Soa.
# past-pref.a-root.hita acc DX.sg child(ren) D.sg Soa
# "Soa (intimate) saw the (previous mention) child."
# Note the presence of the accusative particle 'an' in front of
# DX.sg 'ilay''the previously mentioned'
Nahita ity ankizy i Soa.
# N-a-hita ity ankizy i Soa.
# past-pref.a-root.hita DX.sg child(ren) D.sg Soa
# "Soa (intimate) saw this (previous mention) child."
# Note the absence of the accusative particle 'an' in front of
# DX.sg 'ity''this.near.the.speaker'
Nahita anity ankizy i Soa.
# N-a-hita an' ity ankizy i Soa.
# past-pref.a-root.hita acc DX.sg child(ren) D.sg Soa
# "Soa (intimate) saw the (previous mention) child."
# Note the presence of the accusative particle 'an' in front of
# DX.sg 'ity''this.near.the.speaker'
Nahita ireto ankizy i Soa.
# N-a-hita ireto ankizy i Soa.
# past-pref.a-root.hita DX.pl child(ren) D.sg Soa
# "Soa (intimate) saw these (previous mention) children."
# Note the absence of the accusative particle 'an' in front of
# DX.pl 'ireto''these.near.the.speaker'.
Nahita anireto ankizy i Soa.
# N-a-hita an' ireto ankizy i Soa.
# past-pref.a-root.hita acc DX.pl child(ren) D.sg Soa
# "Soa (intimate) saw these (previous mention) children."
# Note the presence of the accusative particle 'an' in front of
# DX.pl 'ireto''these.near.the.speaker'.
Nahita iro ankizy i Soa.
# N-a-hita iro ankizy i Soa.
# past-pref.a-root.hita DX.sg child(ren) D.sg Soa
# "Soa (intimate) saw this (previous mention) child."
# Note the absence of the accusative particle 'an' in front of
# DX.sg 'iro''this.near.the.speaker/hearer'.
Nahita aniro ankizy i Soa.
# N-a-hita an' iro ankizy i Soa.
# past-pref.a-root.hita acc DX.sg child(ren) D.sg Soa
# "Soa (intimate) saw this (previous mention) child."
# Note the absence of the accusative particle 'an' in front of
# DX.sg 'iro''this.near.the.speaker/hearer'.
Nahita irero ankizy i Soa.
# N-a-hita irero ankizy i Soa.
# past-pref.a-root.hita DX.pl child(ren) D.sg Soa
# "Soa (intimate) saw those (previous mention) children."
# Note the absence of the accusative particle 'an' in front of
# DX.pl 'irero''these.near.the.speaker/hearer'.
Nahita anirero ankizy i Soa.
# N-a-hita an' irero ankizy i Soa.
# past-pref.a-root.hita acc DX.pl child(ren) D.sg Soa
# "Soa (intimate) saw those (previous mention) children."
# Note the presence of the accusative particle 'an' in front of
# DX.pl 'irero''these.near.the.speaker/hearer'.
# Remarks on pseudo-transitive verbs in Malagasy:
# In Malagasy, many verbs look like could they be simple transitive verbs; in reality, they are
# not as they do not undergo passive. A typical example is 'mahita''to see'. This phenomenon seems
# to be generally related to the morphological complexity of the verbal predicate under consideration.
# At first, we will consider apparently simple predicates before delving into complex ones.
# Part A: Morpheme aha > a & lack of morphological Fusion, i.e. a complex predicate
# comprising this morpheme remains complex in that there is no fusion between the
# higher and the lower predicates. See Randriamasimanana (1986:129-177) for further
# details. This contrasts with the situation to be dealt with in Part B.
hita
# hita root 'be.seen';
mahita
# mahita m+a+hita pseudo-transitive verb 'to see' or 'to be able to see';
# at first sight, the form mahita looks like a straightforward and simple predicate;
# in reality, it is a complex predicate, which will assign accusative case to its
# complement, but not be able to undergo passive as a morphological unit; we will
# label this type of verb as 'pseudo-transitive';
# mahita < m+a+hita < m+aha+hita 'active.voice.pres-aha-root.hita'; standard Malagasy being
# an h-dropping variety, aha reduces to aa > a, hence the apparent prefix a; evidence for
# deriving a from aha, mahita = 'see' as well as 'can see'(ABILITATIVE meaning); the
# above solution was first suggested in Rabenilaina (1996:133, Table 11).
# Evidence that mahita is a pseudo-transitive V:
# (A) It is true that 'mahita' requires a complement, which appears to be a Direct Object,i.e.
# DO at least, as far as case-assignment is concerned; consider
Nahita an'Ikala Soa i Koto.
# N-a-hita an Ikala Soa i Koto.
# past-pref.a-root.hita acc D.sg Soa D.sg Soa
# "Koto (intimate) saw Soa (very intimate, only between females)."
# Ungrammatical:
Nahita Ikala Soa i Koto.(0! 0 0 0)
# N-a-hita Ikala Soa i Koto
# past-pref.a-root.hita D.sg Soa D.sg Koto
# "Koto (intimate) saw Soa (very intimate, only between females)."
# (B) However, if mahita was really a transitive V it should be possible to 'promote' its DO
# to grammatical subject. Now this is totally impossible! The reason for this kind of
# phenomenon can be related to the general behavior of ABILITATIVE aha when combining with other
# roots, i.e. passive roots (exactly the case of mahita) as well as nonpassive roots.
# In the sequence ABILITATIVE aha + root, we have two separate predicates, i.e.
# the higher predicate aha and the lower root predicate. The complex aha + root
# can assign accusative case to its complement; but because we do not have one single
# predicate but two different ones, passive of the complex sequence is totally impossible.
# The above phenomenon is not restricted to the pseudo-transitive V mahita. It applies
# to all sequences comprising ABILITATIVE aha + passive root (first section). Consider
# the examples below
Nahaloaka an'ilay varavarana i Koto.
# N-aha-loaka an' ilay varavarana i Koto.
# past-aha-root.loaka acc DX.sg window D.sg Koto
# 'Koto managed to perforate the window.'
# Ungrammatical:
Nahaloakan'i Koto ilay varavarana.(0! 0 0 0)
# N-aha-loaka-n(a) i Koto ilay varavarana .
# past-aha-root.loaka-suff.ana D.sg Koto DX.sg window
# 'The window was managed to be perforated by Koto.'
# The above sentence simply does NOT exist in Malagasy and the initial sequence does not
# have a passive voice counterpart comprising the verbal prefix aha: An attempt was made
# to passivise on the complex verbal predicate aha + loaka, with the AGENT encoded as a
# genitive introduced by -n(a). In fact, the nearest equivalent to a passive voice sequence
# is the following
Loakan'i Koto ilay varavarana.
# Loaka-n' i Koto ilay varavarana.
# root.loaka-clitic.na D.sg Koto DX.sg window
# 'The (previous mention) window has been successfully perforated by Koto.'
# where the ABILITATIVE prefix aha has been taken off the passive root and the AGENT
# is now encoded as a genitive via the clitic -n(a) 'by'. Note that the two sequences
# ' Nahaloaka an'ilay varavarana i Koto' in the active voice with aha and the perfective
# aspect sequence 'Loakan'i Koto ilay varavarana' are cognitively equivalent.
# The above phenomenon is not restricted to the pseudo-transitive V mahita. It also applies
# to all sequences comprising ABILITATIVE aha + nonpassive root (second section); recall that
# this is an open-ended process. Consider the following examples
Nahasasa an'ilay lamba i Soa.
# N-aha-sasa an' ilay lamba i Soa.
# past-aha-root.sasa acc DX.sg linen D.sg Soa
# 'Soa managed to wash the (previous mention) linen.'
# Ungrammatical:
Nahasasan' i Soa ilay lamba.(0! 0 0 0)
# N-aha-sasa-n(a) i Soa ilay lamba.
# past-aha-root.sasa-suff.(a)na D.sg Soa DX.sg linen
# 'The (previous mention) linen was managed to be washed by Soa.'
# The above sentence simply does NOT exist in Malagasy and the initial sequence does not
# have a passive voice counterpart comprising the verbal prefix aha: An attempt was made
# to passivise on the complex verbal predicate aha + sasa, with the AGENT encoded as a
# genitive introduced by -n(a). In fact, the nearest equivalent to a passive voice sequence
# is the following
Voasasan' i Soa ilay lamba.
# Voa-sasa-n(a) i Soa ilay lamba.
# Perf-root.sasa-clitic.na D.sg Soa DX.sg linen
# 'The (previous mention) linen has been successfully washed by Soa.'
# where the ABILITATIVE prefix aha has been taken off the nonpassive root and the AGENT is now
# encoded as a genitive via the clitic -n(a). The two sequences 'Nahasasa an'ilay lamba i Soa'
# in the active voice with aha and the perfective aspect sequence 'Voasasan' i Soa ilay lamba'
# are cognitively equivalent.
# Part B: Absence of morpheme aha > a & morphological Fusion, i.e. a complex predicate
# from which this morpheme aha > a is absent indicates that there may be a fusion between the
# higher and the lower predicates. This contrasts with the situation dealt with in
# Part A above.
tapaka be.cut
manapaka m+an+(t)apaka > manapaka genuinely transitive verb 'to cut'
# In the active voice, Fusion of the lower predicate, i.e. the root 'tapaka''be.cut' into the
# the higher predicate 'an' is quite apparent in the disappearance of the initial consonant of
# the root and in the homorganic nasal assimilation usually affecting the nasal of prefix an.
# Evidence that manapaka 'to cut' is a genuinely transitive verb. Its complement can not only
# take the accusative case-marker 'an'; its direct object can also become a subject, as in
Nanapaka an'ilay tady i Paoly.
# N-an-(t)apaka an' ilay tady i Paoly.
# past-pref.an-root.tapaka acc DX.sg rope D.sg Paul
# "Paul cut the (previous mentioned) rope.'
Notapahin'i Paoly ilay tady.
# No-tapah-in' i Paoly ilay tady.
# past.pass-stem.tapah-suff.ina D.sg Paul DX.sg rope
# 'The (p.m.) rope was cut by Paul.'
# This last sequence shows that the initially complex sequence n+an+tapaka has become one
# single unit and that passive on the entire sequence is possible with passive 1 affixes
# no...ina. This is characteristic of a typically transitive verb, which is explicitly
# indicated in the lexicon by a label like MANIPULATIVE an in the active voice sequence
# m+an 'active.voice.present.tense-prf.an' along with its passive 1 counterpart involving
# the affixes 'no...ina'.
# As expected, with manapaka m+an+(t)apaka the prefix an is MANIPULATIVE an since the embedded
# predicate tapaka 'be.cut' is not a psychological predicate. See Randriamasimanana (1986:
# 20-26), for example, for further illustrations. If, on the other hand, the prefix an was
# CAUSAL, then the embedded predicate is more than likely to be a psychological predicate.
# Part C: Pseudo-transitive complex predicates.
# It is not just the sequence ABILITATIVE aha+root, which gives rise to a pseudo-transitive
# complex predicate, as was shown to be the case in Part A above. It is also the case with
# most other causative constructions including the following, which all clearly involve
# an embedded psychological predicate:
# CAUSAL aha+root
mahafaty
# root faty N 'corpse';
# m+aha+faty 'pres-aha-faty''dangerous'
Io no nahafaty an'i Paoly.
# Io no nahafaty an' i Paoly
# this focus past-aha-root.faty acc D.sg Paul
# 'It is this which caused Paul's death.'
# Ungrammatical:
Io no nahafaty i Paoly.(0! 0 0 0)
# Io no nahafaty i Paoly
# this focus past-aha-root.faty D.sg Paul
# 'It is this which caused Paul's death.'
# The accusative case-marker an has been left out and the second sequence is irretrievably
# ungrammatical.
# Ungrammatical:
Nahafatesan'i Paoly io.(0! 0 0 0)
# N-aha-fate-s-an' i Paoly io.
# past-aha-stem.fat(i>e)-s-suffix.ana D.sg Paul this
# 'It is this which caused Paul's death.'
# Putative passive voice for 'Io no nahafaty an'i Paoly'.
# An attempt was made to passivise on the complex predicate aha+faty and the output is
# total...