# Verb-initial grammars: A multilingual/parallel perspective # ESRC Project RES-000-23-0505 # Oxford University # Charles Randriamasimanana # Malagasy syntax & semantics: # Malagasy Causatives & CONTROL: ESRC-OX-05-CR202. # Causative prefixes amp(a)1, an(a)1, ank(a), aha1, amp(a)2, an(a)1, an(a)2 # and aha2 depending on whether lower predicate denotes CONTROL or not. # As explained in Randriamasimanana (1986) the notion of CONTROL plays a crucial role # in the grammar of Malagasy. In this document, an approximation of such a notion # will be provided; this will be followed by examples involving 'subject control' and # 'object control' structures. # This notion of CONTROL will among other things help in determining whether a given # preposition (PP) is an argument of the verb or an adjunct. By and large, a PP with # [ + CONTROL ] feature verb is an argument, whereas one with [ - CONTROL ] feature verb # is an adjunct. Obviously word order can also help in the determination of the status # of a given constituent, as will be illustrated below. But there exist Malagasy verbs # which can contain either [ + CONTROL ] or [ - CONTROL ] feature depending on their use, # with consequences in the grammar. # In addition to the above, certain PPs, i.e. for instance TIME-Oblique, which are # clear-cut adjuncts can also become arguments via incorporation. # Last but not least, towards the end of this document an assessment will be offered # of a number of recent 'Malagasy' Control structures such as the one proposed in # (Law (1995), (8)) as well as a number of similar sequences in the following papers: # (a) 'Some Remarks on Austronesian Clause Structure: Implications from Malagasy Control # Complements. Maria Polinsky, University of California, San Diego & Eric Potsdam, University # of Florida.Handout given at Stanford University, Workshop on Austronesian syntax # and semantics, February 5, 2005. # (b) Abstract entitled 'Backward Object Control: Against an Empty Category Analysis' # submitted by Eric Potsdam & accepted by West Coast Conference on Formal Linguistics 25, # scheduled for April 28, 2006. # CONTROL = DELIBERATE ACTIVITY as shown in Randriamasimanana (1986: 30 & # subsequent pages) and as a result is totally incompatible with # PERCEPTION verbs: Hitan'i Paoly ilay trano. # Hita-n' i Paoly ilay trano. # root.be.seen-by D.sg Paul DX.sg house(s) # Lit. "The (previous mention) house was seen by Paul." # "Paul saw the (previous mention) house." # Hita is a root verb, a PERCEPTION verb and as such canNOT # combine with a CONTROL affix (for example, passive no...ina # corresponding to the causative prefix an(a)1 mentioned above). # Note that the 'n(a)' on root 'hita' is epenthetic 'n(a)' # and as such, there is no Stress Shift, as would be required # it the morpheme came from suffix 'ina'. # Ungrammatical Nohitan'ny lehilahy ny trano.(0! 0 0 0) # No-hita-n' ny lehilahy ny trano. # past.pass-root.be.seen-pass Det.unspec man/men Det.unspec house(s) # "The house was seen by Rabe." (original translation) # From Edward Louis Keenan (2000) in Formal Issues in Austronesian Linguistics, # eds. Ileana Paul, Vivianne Philips and Lisa de Mena Travis, 27-47. Dordrecht, the # Netherlands: Kluwer Academic Press. # no...in(a) denotes CONTROL while the root verb hita is a PERCEPTION predicate. # Contradiction between CONTROL and PERCEPTION = NO CONTROL. # This sentence simply does NOT exist in Malagasy! # Selection of the appropriate verbal prefix depends on CONTROL. # As a rule of thumb, DIRECTIVE causative amp(a)1 is appropriate when # the lower predicate denotes CONTROL; whereas the MANIPULATIVE causative # an(a)1 is appropriate when the lower predicate has NO CONTROL whatsoever. # Ungrammatical Manosika anay hiditra ianareo.(0! 0 0 0) # M-an-(t)osika anay h-iditra ianareo.(0! 0 0 0) # pres-prf-root.push acc.us fut-enter you.pl # "You urge us to go." (original translation) # This sequence is similar to the one in Matthew Pearson (2001:119, # ex. 89a), # Note the prefix 'an(a)1 on the root tosika. It is the MANIPULATIVE # causative prefix as opposed to the DIRECTIVE causative prefix amp(a)1. # In English, it is obvious that the intended meaning is the DIRECTIVE and # not the MANIPULATIVE interpretation, given the meaning of English 'urge'. # This means that the appropriate causative prefix is amp(a)1 in Malagasy: # Mamporisika anay hiditra ianareo. # M-amp-orisika anay h-iditra ianareo. # pres-prf-urge acc.us fut-enter you.pl # "You urge us to go." Mamporisika anay hiditra ianareo. # M-amp-orisika anay h-iditra ianareo. # pres-prf-urge acc.us fut-enter you.pl # "You urge us to go." # Whether we go or not depends on us as the lower predicate 'hiditra' # 'will go' denotes CONTROL. Hence the selection of the DIRECTIVE # causative prefix amp(a)1. Manosika fiara ianareo. # M-an-(t)osika fiara ianareo. # pres-prf-root.push (a) car(s) you.pl # "You are (physically) pushing a car." # This event involves a MANIPULATIVE event. Hence the selection of the # MANIPULATIVE causative prefix an(a)1. # Control structures # Subject Control Nikasa hividy boky i Jeanne. # N-i-kasa h-i-vidy boky i Jeanne. # past-prf-root.intend fut-prf-root.buy a book(s) D.sg Jeanne #'Jeanne intended to buy a book/books.' # The subject Control predicate nikasa constituent-selects the future tense # on the embedded verb hividy 'will buy'. # Nikasa & hividy have DURATIVE, i.e. not PUNCTUAL aspect. # The distinction between DURATIVE and PUNCTUAL aspects was established in # Simeon Rajaona Structure du malgache. PhD dissertation, Sorbonne 1969 & # later Editions Ambozontany, Fianarantsoa, Madagascar. # The distinction is crucial since it helps avoid incongruous combinations # like the following: # Incompatibility type 1 - Ileana Paul (2000:1) example (1)a # Nanapaka ity hazo ity tamin'ny antsy i Sahondra. (0! 0 0 0) # past.AT.cut this tree this pst.P.gen.det knife Sahondra # "Sahondra cut this tree with the knife." # where the headverb nanapaka has DURATIVE aspect; yet its complement 'ity # hazo ity' is associated with PUNCTUAL aspect! # Incompatibility type 2 - Ileana Paul (2000:1) example (1)c # Nanapahan'i Sahondra ity hazo ity ny antsy. (0! 0 0 0) # pst.CT.cut.gen.Sahondra this tree this det knife #'The knife was used by Sahondra to cut the tree." # Nanapahana = n-an-(t)apaka-ana 'past-pass-be-cut-pass' (note the # circumfix an...an(a) has DURATIVE aspect, which in conjunction with # Spec-Head in Malagasy will force a PARTITIVE reading of the subject NP # "ny antsy" 'the knife', yielding something like 'a portion of the knife # was used by Sahondra to cut this tree.' # Ungrammatical Nanapaka ity hazo ity tamin'ny antsy i Sahondra. (0! 0 0 0) # N-an-(t)apaka ity hazo ity t-amin' ny antsy i Sahondra. # past-prf-be.cut DX.sg tree(s) DX.sg perf-with Det knife/ves D.sg Sahondra # "Sahondra cut this tree with the knife." # This sentence comes from Ileana Paul (2000:1) example (1)c. # The headverb nanapaka has DURATIVE aspect; yet its complement 'ity # hazo ity' is associated with PUNCTUAL aspect! # Ungrammatical Nanapahan'i Sahondra ity hazo ity ny antsy. (0! 0 0 0) # N-an-(t)apah-an' i Sahondra ity hazo ity ny antsy. # past-pass-be.cut-pass D.sg Sahondra DX.sg tree(s) DX.sg Det.unspec knife/ves #'The knife was used by Sahondra to cut the tree."(original translation) # The circumfix an...an(a) has DURATIVE aspect, which in conjunction with # Spec-Head in Malagasy will force a PARTITIVE reading of the subject NP # "ny antsy" 'the knife', yielding something like 'a portion of the knife # was used by Sahondra to cut this tree' or 'some of the knives were used # by Sahondra to cut this tree." Nikasa ny hividy boky i Jeanne. # N-i-kasa ny h-i-vidy boky i Jeanne. # past-prf-root.intend comp fut-prf-root.buy a book(s) D.sg Jeanne # 'Jeanne intended to buy a book/books.' # The subject Control predicate nikasa constituent-selects the future tense # on the embedded verb hividy 'will buy'. # Note the presence of the complementizer 'ny', typical of subject control # predicates. However comp 'ny' is totally incompatible with an object # control verb: Manosika anay ny hiditra ianareo.(0! 0 0 0) from Ileana # Paul (1998:115) example (11)b. # For further detail about comp 'ny' see Randriamasimanana (1986: 501-503). # Ungrammatical Manosika anay ny hiditra ianareo.(0! 0 0 0) # M-an- (t)osika anay ny h-iditra ianareo.(0! 0 0 0) # pres-prf-root.urge acc.us comp fut-enter you.pl # "You urge us to go in." # This sequence is from Matthew Pearson (2001:119,ex. 89a), # The matrix verb 'manosika' is an object control predicate and as such # can only constituent-select a comp like 'mba' (see below). It simply # cannot constituent-select the subject Control comp 'ny'. Nikasa hividy ilay boky i Jeanne. # N-i-kasa h-i-vidy ilay boky i Jeanne. # past-prf-root.intend fut-prf-root.buy DX.sg book(s) D.sg Jeanne #'Jeanne intended to buy the (previous mention) book.' # The subject Control predicate nikasa constituent-selects the future tense # on the embedded verb hividy 'will buy'. # Nikasa & hividy have DURATIVE, i.e. not PUNCTUAL aspect. # Note the absence of comp 'ny'. # Note the absence of the accusative marker 'an'. Nikasa ny hividy ilay boky i Jeanne. # N-i-kasa ny h-i-vidy ilay boky i Jeanne. # past-prf-root.intend comp fut-prf-root.buy DX.sg book(s) D.sg Jeanne #'Jeanne intended to buy the (previous mention) book.' # The subject Control predicate nikasa constituent-selects the future tense # on the embedded verb hividy 'will buy'. # Nikasa & hividy have DURATIVE, i.e. not PUNCTUAL aspect. # Note the presence of comp 'ny'. # Note the absence of the accusative marker 'an'. Nikasa hividy an'ilay boky i Jeanne. # N-i-kasa h-i-vidy an' ilay boky i Jeanne. # past-prf-root.intend fut-prf-root.buy acc DX.sg book(s) D.sg Jeanne #'Jeanne intended to buy the (previous mention) book.' # The subject Control predicate nikasa constituent-selects the future tense # on the embedded verb hividy 'will buy'. # Nikasa & hividy have DURATIVE, i.e. not PUNCTUAL aspect. # Note the absence of comp 'ny'. # Note the presence of the accusative marker 'an'. Nikasa ny hividy an'ilay boky i Jeanne. # N-i-kasa ny h-i-vidy an' ilay boky i Jeanne. # past-prf-root.intend comp fut-prf-root.buy acc DX.sg book(s) D.sg Jeanne #'Jeanne intended to buy the (previous mention) book.' # The subject Control predicate nikasa constituent-selects the future tense # on the embedded verb hividy 'will buy'. # Nikasa & hividy have DURATIVE, i.e. not PUNCTUAL aspect. # Note the presence of comp 'ny'. # Note the presence of the accusative marker 'an'. Nokasain'i Jeanne hovidiana ilay boky. # No-kasa-in' i Jeanne ho-vidi-ana ilay boky. # past.pass-root.intend-pass-by D.sg Jeanne fut.pass-root-pass DX.g book # Lit."Was intended by Jeanne the (previous mention) book will be bought." # "Jeanne intended to buy the (previous mention) book." # Note that the subject Control predicate still constituent-selects the future # tense on the embedded verb even though the two verbs are in the passive voice: # no...in(a) passive on matrix verb; ho...ina passive on embedded verb. # Double passive. # Nokasaina & hovidiana have PUNCTUAL aspect; PUNCTUAL event # interpretation of the proposition is obligatory. # no...in(a) & ho...in(a) forms of passive are the essential ingredients # of Malagasy passive (totally overlooked in Ileana Paul's 2000 PhD # dissertation and apparently replaced by a- passive). However see # restriction on a- passive below. This restriction was already explicitly # illustrated for example in Randriamasimanana (1986: 79 example 114b). # Note the absence of comp 'ny' between the two passives. # For further detail about comp 'ny' see Randriamasimanana (1986: 501-503). Nitady handeha i Jeanne. # N-itady h-andeha i Jeanne. # past-seek fut-go D.sg Jeanne # 'Jeanne was trying to leave.' # The subject Control predicate nitady constituent-selects the future tense # on the embedded verb handeha 'will go'. # Nitady & handeha have DURATIVE aspect. # Note the absence of comp 'ny'. Nitady ny handeha i Jeanne. # N-i-tady ny h-andeha i Jeanne. # past-prf-root.seek comp fut-go D.sg Jeanne # 'Jeanne was trying to leave.' # The subject Control predicate nitady constituent-selects the future tense # on the embedded verb handeha 'will go'. # Note the possibility of comp 'ny' with a subject control matrix # predicate. # The presence of comp 'ny' makes passive and fronting possible: # Ny handeha no notadiavin'i Jeanne. # Lit. "It was to go was wanted by Jeanne." "Jeanne really wanted to go" # with some emphasis on 'to go.' # Nitady & handeha have DURATIVE aspect. Ny handeha no notadiavin'i Jeanne. # Ny h-andeha no no- tadi-a-v-in' i Jeanne. # Comp fut-go focus past.passive-seek-a-v-pass D.sg Jeanne # Lit. "It was to go was wanted by Jeanne." "To go, that was what Jeanne # really wanted to do" with some emphasis on 'to go.' # Nitady & handeha have DURATIVE aspect. # Note the presence of comp 'ny'. This presence allows passive and # fronting. # Epenthetic a + v after root 'tadi' are provided by the passive imperative # 'tadi-av-o''root.tadi-av-by.you': The suffix 'o' has been replaced by the # 'in(a) suffix in the discontinuous form 'no...ina'. Nitady handefa ilay ankizy i Jeanne. # N-i-tady h-andefa ilay ankizy i Jeanne. # past-prf-root.seek fut-send DX.sg child D.sg Jeanne # "Jeanne was seeking to send the (previous mention) child." # handefa = h - an - lefa. # fut-prf- root.send, where l > d # The subject Control predicate nitady constituent-selects the future tense # on the embedded verb handefa 'will send'. # Nitady & handefa have DURATIVE aspect. # Interesting illustrative example from Ileana Paul (2000: 27) example # (12)b. of the use of a- passive: # (12)a. Nandidy ny hena tamin'ny antsy Rasoa. # pst.AT.cut det meat pst.gen.det knife Rasoa # "Rasoa cut the meat with the knife." # (12)b. Adidy ny hena ny antsy. instrument # a.cut det meat det knife # "The knife is used to cut the meat." # My remark: (12)b is only acceptable in a context where an electric knife # was used to give an initial impulse to the cutting! Now in Madagascar # an electric knife is almost totally unknown! # For further illustration, see Randriamasimanana (1986:44-46). Nitady handefa ilay ankizy i Jeanne. # N-i-tady h-andefa ilay ankizy i Jeanne. # past-prf-root.seek fut-send DX.sg child D.sg Jeanne # "Jeanne was seeking to send the (previous mention) child." # handefa = h - an - lefa. # fut-prf- root.send, where l > d # The subject Control predicate nitady constituent-selects the future tense # on the embedded verb handefa 'will send'. Notadiavin'i Jeanne halefa ilay ankizy. # No-tadi-av-in' i Jeanne h-a-lefa ilay ankizy. # past.pass-root.seek-pass-by D-sg Jeanne fut-pass-root.go DX.sg child # Lit. "was intended by Jeanne the child will be sent." "Jeanne intended to # send the (previous mention) child." # Epenthetic a + v after root tadi are provided by the passive imperative # 'tadi-av-o''root.tadi-av-by.you': The suffix 'o' has been replaced by the # 'in(a) suffix in the discontinuous form 'no...ina'. # The matrix subject Control predicate notadiavina constituent-selects the future # tense on the embedded verb halefa, itself also in the passive voice; # no...in(a) passive on matrix verb; a- prefix passive on the embedded verb. # Double passive. # Note the following restriction affecting the a- prefix form of # passive containing the feature BALLISTIC: Only a referent that (a) is # capable of autonomy --for an animate-- and (b) that can be tossed around # or catapulted for a non-animate-- can be subject NP. # Assumption: The child can walk on their own unaided and does not have to # be carried around (autonomy)." # Notadiavina & halefa have PUNCTUAL aspect; PUNCTUAL event interpretation # of the proposition is obligatory. # Note the absence of comp 'ny'. Nitady ny handefa an'ilay ankizy i Jeanne. # N-i-tady ny h-andefa an' ilay ankizy i Jeanne. # past-prf-root.seek comp fut-send acc DX.sg child D.sg Jeanne # "Jeanne was seeking to send the (previous mention) child." # handefa = h - an - lefa. # fut-prf- root.send, where l > d # The subject Control predicate nitady constituent-selects the future tense # on the embedded verb handefa 'will send'. # Assumption: The child can walk on their own unaided and does not have to # be carried around (autonomy)." # Nitady and handefa have DURATIVE, i.e. not PUNCTUAL aspect. # Note the presence of comp 'ny'. # Note the presence of the accusative marker 'an'. Nitady handefa ilay taratasy i Jeanne. # N-i-tady h-andefa ilay taratasy i Jeanne. # past-prf-root.seek fut-send DX.sg letter D.sg Jeanne # "Jeanne was seeking to send the (previous mention) letter." # handefa = h - an - lefa. # fut-prf- root.send, where l > d # The subject Control predicate nitady constituent-selects the future tense # on the embedded verb handefa 'will send'. # Nitady and handefa have DURATIVE, i.e. not PUNCTUAL aspect. # Note the absence of comp 'ny'. # Note the absence of the accusative marker 'an'. Nitady ny handefa ilay taratasy i Jeanne. # N-i-tady ny h-andefa ilay taratasy i Jeanne. # past-prf-root.seek comp fut-send DX.sg letter D.sg Jeanne # "Jeanne was seeking to send the (previous mention) letter." # handefa = h - an - lefa. # fut-prf- root.send, where l > d # The subject Control predicate nitady constituent-selects the future tense # on the embedded verb handefa 'will send'. # Nitady and handefa have DURATIVE, i.e. not PUNCTUAL aspect. # Note the presence of comp 'ny'. # Note the absence of the accusative marker 'an'. Nitady handefa an'ilay taratasy i Jeanne. # N-i-tady h-andefa an' ilay taratasy i Jeanne. # past-prf-root.seek fut-send acc DX.sg letter D.sg Jeanne # handefa = h - an - lefa. # fut-prf- root.send, where l > d # "Jeanne was seeking to send the (previous mention) letter." # The subject Control predicate nitady constituent-selects the future tense # on the embedded verb handefa 'will send'. # Nitady and handefa have DURATIVE, i.e. not PUNCTUAL aspect. # Note the absence of comp 'ny'. # Note the presence of the accusative marker 'an'. Notadiavin'i Jeanne halefa ilay taratasy. # No-tadi-av-in' i Jeanne h-a-lefa ilay taratasy. # past.pass-root.seek-pass-by D.sg Jeanne fut-pass-root.go DX.sg letter # "Jeanne was seeking to send the (previous mention) letter." # Epenthetic a + v after root tadi are provided by the passive imperative # 'tadi-av-o''root.tadi-av-by.you': The suffix 'o' has been replaced by the # 'in(a) suffix in the discontinuous form 'no...ina'. # The matrix subject Control predicate notadiavina constituent-selects the future # tense on the embedded verb halefa, itself also in the passive voice; # no...in(a) passive on matrix verb; a- prefix passive on the embedded verb. # Double passive. # Note the following restriction affecting the a- prefix form of # passive containing the feature BALLISTIC: Only a referent that (a) is # capable of autonomy --for an animate-- and (b) that can be tossed around # or catapulted for a non-animate-- can be subject NP. # Notadiavin' & halefa have PUNCTUAL aspect; PUNCTUAL event interpretation # of the proposition is obligatory. # Note the absence of comp 'ny'. # Ungrammatical Miditra ny handeha i Soa. (0! 0 0 0) # M-i-ditra ny h-andeha i Soa. # pres-prf-insist comp fut-go DX-.g Soa # 'Soa insists on going." # This sequence is from Ileana Paul (1998:115, ex.11a). # Even though a matrix predicate like 'miditra' may strictly subcategorise # for a future tense on the embedded verb; it does not always allow a comp # 'ny'. If that was the case, the following should be grammatical: # Ny handeha no idiran'i Soa. (0! 0 0 0) following the earlier pattern above # Nitady ny handeha i Jeanne. # N-i-tady ny h-andeha i Jeanne. # past-prf-root.seek comp fut-go D.sg Jeanne # 'Jeanne was trying to leave.' # Note the possibility of comp 'ny' with a subject control matrix predicate. # The presence of comp 'ny' makes passive and fronting possible: # Ny handeha no notadiavin'i Jeanne. # Ny h-andeha no no-tadi-av-in' i Jeanne. # Comp fut-go focus past.pass-root.seek-av-pass.by D.sg Jeanne # Lit. "It was to go was wanted by Jeanne." "Jeanne really wanted to go" with # some emphasis on 'to go.' Miditra handeha i Soa. # M-i-ditra h-andeha i Soa. # pres-prf-insist fut-go DX.sg Soa # 'Soa insists on going." # Note the absence of comp 'ny'. Ny handeha no notadiavin'i Jeanne. # Ny h-andeha no no-tadi-av-in' i Jeanne. # Comp fut-go focus past.pass-root.seek-av-pass.by D.sg Jeanne # Lit. "It was to go was wanted by Jeanne." "Jeanne really wanted to go" with # some emphasis on 'to go.' # Epenthetic a + v after root tadi are provided by the passive imperative # 'tadi-av-o''root.tadi-av-by.you': The suffix 'o' has been replaced by the # 'in(a) suffix in the discontinuous form 'no...ina'. # Ungrammatical Ny handeha no idiran'i Soa. (0! 0 0 0) # Ny h-andeha no i-dir-an' i Soa. # Comp fut-root.go focus pref-insist-pass D.sg Soa # 'Trying to leave, Soa was.' # Note the possibility of comp 'ny' with a subject control matrix predicate. # The presence of comp 'ny' makes passive and fronting possible. # Control structures # Object Control Niangavy an'i Jeanne mba handeha i Marie. # N-i-angavy an' i Jeanne mba h-andeha i Marie. # past-prf-request acc D.sg Jeanne comp fut-go D.sg Marie # "Marie asked Jeanne to go." # 'Niangavy' constituent-selects the future tense on the embedded verb handeha. # The object control predicate 'niangavy' allows a complementizer 'mba' and rules out # complementizers like 'fa' = that or 'ny' of subject control. See below. # The complementizer 'mba' is optional. # See further detail in Randriamasimanana (1986: 513-515). # 'Niangavy' and 'handeha' have DURATIVE aspect. # Note the marked word order SVO in the embedded structure 'i Jeanne # handeha' with apparently the lower specifier exceptionally case-marked # by the matrix verb. Niangavy an'i Jeanne handeha i Marie. # N-i-angavy an' i Jeanne h-andeha i Marie. # past-prf-request acc D.sg Jeanne fut-go D.sg Marie # "Marie asked Jeanne to go." # 'Niangavy' constituent-selects the future tense on the embedded verb handeha. # The object control 'niangavy' allows a complementizer 'mba' and rules out # complementizers like 'fa' = that or 'ny' of subject control. See below. # The complementizer 'mba' is absent. Niangavian'i Marie handeha i Jeanne. # N-i-angavi-an' i Marie h-andeha i Jeanne. # past-prf-pass.stem.request-pass.by D.sg Marie fut-go D.sg Jeanne. # Lit. "Was requested by Marie Jeanne will go.""Marie asked Jeanne to go." # 'Niangavian' has PUNCTUAL aspect while 'handeha' has DURATIVE aspect. Nanery an'i Jeanne hipetraka i Marie. # N-an-(t)ery an' i Jeanne h-ipetraka i Marie. # past-prf-root.force acc D.sg Jeanne fut-sit D.sg Marie # "Marie forced Jeanne to sit down." # 'Nanery' constituent-selects the future tense on the embedded verb hipetraka. # 'Nanery' and 'hipetraka' have DURATIVE aspect. # Note the marked word order SVO in the embedded structure 'i Jeanne # hipetraka'. Nanery an'i Jeanne hitondra ilay fiara i Marie. # N-an-(t)ery an' i Jeanne h-itondra ilay fiara i Marie. # past-prf-root.force acc D.sg Jeanne fut-drive the car D.sg Marie # "Marie forced Jeanne to drive the (previous mention) car." # 'Nanery' constituent-selects the future tense on the embedded verb hitondra. # Note the marked word order SVO in the embedded structure 'i Jeanne # hitondra ilay fiara'. # Note the marked word order SVO in the embedded structure 'i Jeanne # hitondra ilay fiara' with apparently the lower specifier exceptionally # case-marked by the matrix verb. Noteren'i Marie hitondra ilay fiara i Jeanne. # No-tere-n' i Marie h-itondra ilay fiara i Jeanne . # past.pass-root.force-pass.by D.sg Marie fut-drive the car D.sg Jeanne # "Marie was forced by Jeanne to drive the (previous mention) car." # The stem 'tere' comes from the passive imperative form 'tere-o' from which # the clitic pronoun suffix 'o''by.you' is dropped replaced by the passive # suffix '(a)na' in the discontinuous form 'no...ana'.. Naniraka an'ilay vehivavy hividy mofo ianareo. # N-an-iraka an' ilay vehivavy h-ividy mofo ianareo. # past-prf-root.send acc DX.sg woman/women fut-buy bread you.pl # "You sent the (previous mention) woman to buy bread." # 'Naniraka' constituent-selects the future tense on the embedded verb hividy. # Maniraka mividy mofo an'ilay vehivavy ianareo. (0! 0 0 0) # This sequence is from Ileana Paul (1998: 114, ex. 10e.), where # 'ilay vehivavy' follows the second NOT the first verb and # where the latter is in the present tense instead of the expected future tense. # Maniraka an'ilay vehivavy mividy mofo Rasoa.(0! 0 0 0) from Matthew Pearson # (2001: 116, ex. 81a.), where the embedded verb is in the present tense. # Note the absence of comp 'mba'. # Note the marked word order SVO in the embedded structure 'i Jeanne # hividy mofo'. # Ungrammatical Maniraka mividy mofo an'ilay vehivavy ianareo. (0! 0 0 0) # M-an-iraka m-ividy mofo an' ilay vehivavy ianareo. # pres-prf-send pres-buy bread acc DX.sg woman you.pl # "You sent the (previous mention) woman to buy bread." # This sequence is from from Ileana Paul (1998: 114, ex. 10e.), # where 'ilay vehivavy' follows the second verb and # where the latter is in the present tense instead of the required future # tense. See further examples in Randriamasimanana (1986: 510-515). # Maniraka an'ilay vehivavy mividy mofo Rasoa.(0! 0 0 0) from Matthew Pearson # (2001: 116, ex. 81a.), where the embedded verb is also in the present tense. # Ungrammatical Maniraka an'ilay vehivavy mividy mofo Rasoa.(0! 0 0 0) # M-an-iraka an' ilay vehivavy m-ividy mofo ianareo. # pres-prf-root.send acc DX.sg voman/women pres-buy bread you.pl # "You send the (previous mention) woman to buy bread." # This sequence is from Matthew Pearson (2001: 116, ex. 81a.), where the # embedded verb is in the present tense instead of the required future tense. # See further examples of embedded future tense under an object control verb # in Randriamasimanana (1986: 510-515). Naniraka an'ilay vehivavy mba hividy mofo ianareo. # N-an-iraka an' ilay vehivavy mba h-ividy mofo ianareo. # past-prf-root.send acc DX.sg voman/women comp fut-buy bread you.pl # "You sent the (previous mention) woman to buy bread." # 'Naniraka' constituent-selects the future tense on the embedded verb hividy. # Note the optional comp 'mba', characteristic of object control verbs. # Note the presence of comp 'mba'. Naniraka an'ilay vehivavy hividy mofo ianareo. # N-an-iraka an' ilay vehivavy h-ividy mofo ianareo. # past-prf-root.send acc DX.sg voman/women fut-buy bread you.pl # "You sent the (previous mention) woman to buy bread." # 'Naniraka' constituent-selects the future tense on the embedded verb hividy. # Note the absence of comp 'mba', characteristic of object control verbs. # Note the marked word order SVO in the embedded structure 'ilay vehivavy # hividy mofo'. # Ungrammatical Manosika anay ny hiditra ianareo.(0! 0 0 0) # M-an-(t)osika anay ny h-iditra ianareo.(0! 0 0 0) # pres-prf-urge acc.us comp fut-enter you.pl # "You urge us to go in." (original translation) # This sequence is from Matthew Pearson (2001:119, ex. 89a), where # an object control predicate constituent-selects the comp 'ny' restricted to # subject control predicates; This is instead of the comp 'mba' as expected # for a putative object control construction like 'manosika''push'. # See explicit restriction stated in Randriamasimanana (1986: 513). # Ungrammatical Manosika anay mba hiditra ianareo.(0! 0 0 0) # M-an-(t)osika anay mba h-iditra ianareo. # pres-prf-coerce acc.us comp fut-enter you.pl # "You push us into entering." (revised translation) # Note the presence of comp 'mba'. 'Mba' is a word meaning something like # 'please', which presumably would be rather incompatible with the higher verb # 'manosika' meaning 'push', hence the dubious grammaticality of the sentence. # On the meaning of the particle 'mba', see Randriamasimanana (1986: 514-515). # Ungrammatical Manosika anay hiditra ianareo. # M-an-(t)osika anay h-iditra ianareo. # pres-prf-root.coerce acc.us fut-enter you.pl # "You push us into entering." (? revised translation) # Note the absence of comp 'mba'. # This sequence is of very doubtful grammaticality, given the meaning of # 'manosika' 'to physically push something.' # Typically the causative prefix 'an(a) is associated with a manipulative # meaning, as illustrated in Randriamasimanana (1986: 48-49). # Ungrammatical Anosehanareo anay ny hiditra. (0! 0 0 0) # An- (t)oseh(a)- an(a) -nareo anay ny h-iditra. # pass-stem.coerce-pass -by.you acc.us (excl) comp fut-go # "Going in, you urge us (to do it)." (original translation) # This sequence does not exist in Malagasy and simply canNOT be translated. # This sequence is from Matthew Pearson (2001:119, ex. 89b) originally # translated as "Going in, you urge us (to do it)." This is a logical # consequence of an earlier decision to use the subject control comp 'ny' # instead of the object control comp 'mba' under the putative object control # predicate 'manosika' "push'. Once comp 'ny' is present in the sequence, # passive and fronting can be applied. # See explicit restrictions stated in Randriamasimanana (1986: 46-48) as # well as in Randriamasimanana (1986: 501-502). # Ungrammatical Anosehan'i Sahondra anay ny hiditra. (0! 0 0 0) # An- (t)oseh(a)- an(a) -n' i Sahondra anay ny h-iditra. # pass-stem.coerce-pass -by D.sg S acc.us (excl) comp fut-go # "Sahondra urges us to go in."( original translation) # This sequence is from Ileana Paul (1998:117, ex. 17b) originally # translated as "Sahondra urges us to go in." This is a logical consequence # of an earlier decision to use the subject control comp 'ny' instead of the # object control comp 'mba' under the object control predicate 'manosika' # "push'. Once comp 'ny' is present in the sequence, passive and # fronting can be applied, yielding the above irretrievably ungrammatical # sequence. This sequence does not exist in Malagasy and simply canNOT be translated. # See explicit restrictions stated in Randriamasimanana (1986: 46-48) as # well as in Randriamasimanana (1986: 501-502). # Ungrammatical Manosika miditra anay i Sahondra.(0! 0 0 0) # M-an-(t)osika m-iditra anay i Sahondra. # pres-prf-root.coerce pres-enter acc.us D.sg Sahondra # "Sahondra urges us to go in." (original translation) # This sequence is from Ileana Paul (1998:117, ex. 17a). # This sequence does not exist in Malagasy and simply canNOT be translated. # Note that the two verbs 'manosika' and 'miditra' are adjacent and have the # same present tense-markers on them: This suggests that manosika 'urge' is # a verb of the same subcategory as manandrana 'to try'; now manandrana 'to try' # as a verb requires an adjunct with the other verb in the same tense as itself; # now manosika as it is intended here is rather an object Control predicate. # This sequence does not exist in Malagasy and simply canNOT be translated. # Typically the causative prefix 'an(a) is associated with a manipulative # meaning, as illustrated in Randriamasimanana (1986: 48-49). # Distinction between argument vs adjunct and CONTROL # Given a PP (a prep or an adverb phrase), the latter is likely to be # an argument if the governing verb contains the positively valued feature # [+ CONTROL]; whereas it is likely to be an adjunct if the verb has the # negatively valued feature [- CONTROL]. See further elaboration and more # illustrative examples in Randriamasimanana (1999: 515-516, 522-525) as well # as in Randriamasimanana (2004: 346-350. Faits de Langues, Editions Orphrys. # Paris, France). Nandidy hena tamin'ny antsy i Bakoly. # N-an-didy hena t-amin' ny antsy i Bakoly. # past-prf-root.cut meat perf-with Det.unspec knife D.sg Bakoly # "Bakoly cut meat with the (previous mention) knife." # 'amin(a)' is a general preposition, which has several potential meanings. # Here it means 'with' and accompanies the verb 'nandidy''past-cut', which # contains the positively valued feature CONTROL. As a direct result of # this, PP 'amin(a)' is an argument of the verb 'nandidy'. Consequence: # When the PP 'tamin'ny antsy' is fronted, the verb must be in the passive # voice. # Note the argument status of PP 'tamin(a) indicated by the gloss 'perf' # for the 't' prefix on the particle. Tamin'ny antsy no nandidian'i Bakoly hena. # T-amin' ny antsy no n- an- didi-an' i Bakoly hena. # perf-with Det.unspec knife focus past.pass-root.cut-pass.by D.sg B meat # "It is with the (previous mention) knife that Bakoly cut meat." # The PP 'tamin'ny antsy' is an argument of the verb 'nandidy''past-cut' # since the latter contains the positively valued feature CONTROL: To cut # is a DELIBERATE kind OF ACTIVITY. # When an argument is fronted, the verb must be in the passive voice. # Note the argument status of PP 'tamin(a) indicated by the gloss 'perf' # for the 't' prefix on the particle. # See additional examples in Randriamasimanana (1986: 462, ex 53a. and 466, # ex 63a. and 63b.) showing that the instrument oblique is an argument. # See further elaboration in Randriamasimanana (2004: 348-349. Faits de Langues, # Editions Orphrys. Paris, France). Ungrammatical Tamin'ny antsy no nandidy hena i Bakoly.(0! 0 0 0) # T-amin' ny antsy no n- an- didy hena i Bakoly. # perf-with Det.unspec knife focus past-prf-root.cut meat D.sg Bakoly # "It is with the (previous mention) knife that Bakoly cut meat." # Adapted from Ileana Paul (2000: 38, ex 26b). # The verb nandidy 'past-cut' with the feature [+ CONTROL] takes the PP # 'tamin'ny antsy' as an argument. Yet when the PP-argument is fronted, # the verb does not take the passive voice: an illegal move. # Note the argument status of PP 'tamin(a) indicated by the gloss 'perf' # for the prefix 't' on the particle. # See additional examples in Randriamasimanana (1986: 462, ex 53a. and 466, # ex 63a. and 63b.) showing that the instrument oblique is an argument. # See further elaboration in Randriamasimanana (2004: 349. Faits de Langues, # Editions Orphrys. Paris, France). Ungrammatical Amin'ny antsy no mamono ny akoho ny mpamboly.(0! 0 0 0) # ø -amin' ny antsy no m- am-(v)ono ny akoho # nonperf-with Det.unspec knife focus pres-prf-root.kill Det.unspec chicken # ny mpamboly. # Det.unspec farmer # "It is with the knife that the farmer is killing the chickens". # (Original translation) from Matthew Pearson (AFLA 10, 28 March 2003 # example 14a handout). # The verb mamono 'pres-kill' with the feature [+ CONTROL] takes the PP # 'amin'ny antsy' as an argument. Yet when the PP-argument is fronted, # the verb does not take the passive voice: an illegal move. # Note the argument status of PP ø -amin(a) indicated by the gloss 'nonperf' # for the prefix 'ø' on the particle. # There is also another source for the ungrammaticality of this sequence: # 'M-amono''pres-kill' is associated with HABITUAL aspect and unspecified # Det 'ny' refers to a specific set of chickens that presumably the # previously mentioned set of farmers keep killing over and over! # See additional examples in Randriamasimanana (1986: 462, ex 53a. and 466, # ex 63a. and 63b.) showing that the instrument oblique is an argument. # See further elaboration in Randriamasimanana (2004: 349. Faits de Langues, # Editions Orphrys. Paris, France). Nirofotra teo amin'€™ny tarehin'i Soa ny mony. # N-i-rofotra t-eo amin'ny tarehi-n'i Soa ny mony. # past-prf-explode past-there on Det.unspec face-of D.sg Soa Det.un. pimples # 'The pimples exploded on Soa'€™s face.' # This sequence has been adapted from Rabenilaina. Lexique-Grammaire du # Malgache. PhD dissertation, Université Paris 7, France.1985. # The PP 't-eo amin'ny tarehin'i Soa''past-on Soa's face' is an adjunct to # the STATIVE verb 'nirofotra''past-explode': This verb clearly contains the # negatively valued feature [- CONTROL]. # Note the adjunct status of PP 'teo amin'ny tarehin'i Soa' indicated by the # gloss 'past' for the prefix 't' on the particle 'teo'. # A 'past...past' configuration is the hallmark of adjunction, as explained # and illustrated in Randriamasimanana (1999: 525). Teo amin'€™ny tarehin'i Soa no nirofotra ny mony. # T-eo amin'€™ ny tarehi-n' i Soa # past-there on Det.unspec face-of D.sg Soa # no n- i- rofotra ny mony. # focus past-prf-root.explode Det.unspec pimples # 'It was on Soa'€™s face that the pimples exploded.'€™ # This sequence was first presented and accounted for by Randriamasimanana in # his invited talk to AFLA V (Department of Linguistics, University of Hawaii # at Manoa, March 26-29, 1998). # The PP 't-eo amin'ny tarehin'i Soa''past-on Soa's face' is an adjunct to # the verb 'nirofotra''past-explode': When the PP is fronted, the verb does # remain in the active voice. # Note the adjunct status of PP 'teo amin'ny tarehin'i Soa' indicated by the # gloss 'past' for the prefix 't' on the particle 'teo'. # A 'past...past' configuration is the hallmark of adjunction, as explained # and illustrated in Randriamasimanana (1999: 525). # Ungrammatical Teo amin'€™ny tarehin'i Soa no nirofotran' ny mony.(0! 0 0 0) # T-eo amin'€™ ny tarehi-n'€™ i Soa # perf-there on Det.unspec face-of D.sg Soa # no n- i- rofotra-n' ny mony. # focus past-pass-explode-pass.by Det.unspec pimples # 'It was on Soa’s face that the pimples exploded.'€™ # The verb 'nirofotra''past-eplode' is a verb with the feature [-CONTROL]: # It is a STATIVE kind of predicate. Yet an attempt is made to passivize on # it with the discontinuous form 'i...an(a)' of passive. # The form 'nirofotrana' simply does not exist in Malagasy. # The PP 'teo amin'ny tarehin'i Soa' is wrongly assumed to be an argument # of the STATIVE verb 'nirofotra''past-explode'. # Word order --in particular, to the right of the subject-- can also signal # adjunct status. Namono tena izy noho ny fitiavana. # N-am-(v)ono tena izy noho ny fitiavana. # past-prf-kill self s/he because.of Det.unspec love # "S/he killed him/herself because of love." # This sequence comes from 'Malagasy and Formosan Languages: A Comparison' # by Randriamasimanana (2004: 422, ex 28a). # Note the occurrence of 'noho ny fitiavana' 'because of love' to the right # of the subject 'izy''s/he', showing that it is an adjunct. # Ungrammatical Ny fitiavana no namonoany tena.(0! 0 0 0) # Ny fitiavana no n- am- (v)ono- a- ny tena. # Det.unspec love focus past-pass-root.kill-pass-by.him/her self # "He killed himself for love" (Original translation) # This sequence has been adapted from Matthew Pearson (2001:33, ex 30c). # Note the attempt to passivize on the adjunct 'noho ny fitiavana'. # See further elaboration in Malagasy and Formosan Languages: A Comparison. # by Randriamasimanana. (2004:421-422). # Verbs with an optional feature [+ CONTROL] or [- CONTROL] Mianatra ao an-davarangana i Soa. # M- i- anatra ø-ao an-davarangana i Soa. # Pres-prf-root.study nonpast-there on-balcony D.sg Soa # ‘Soa (usually) studies on the balcony.' # This sequence was first presented and accounted for by Randriamasimanana in # his invited talk to AFLA V (Department of Linguistics, University of Hawaii # at Manoa, March 26-29, 1998). # In this first interpretation, the verb 'mianatra' 'pres-study' simply # describes an habitual state of affairs: It therefore is deemed to contain # the feature [- CONTROL]. # PP 'ø-ao an-davarangana' 'on the balcony' is an adjunct. # Note the 'ø' glossed as 'nonpast' = non-past tense, somewhat # parallel to the prefix 'm' for 'present tense' on the verb. Ao an-davarangana no mianatra i Soa. # ø-ao an-davarangana no m-i-anatra i Soa. # nonpast-there on-balcony focus pres-prf-root.study D.sg Soa # "It is on the balcony that Soa (usually) studies." # This sequence was first presented and accounted for by Randriamasimanana in # his invited talk to AFLA V (Department of Linguistics, University of Hawaii # at Manoa, March 26-29, 1998). # Note the active voice on the verb with the fronted adjunct PP with this # first interpretation of the verb as containing the feature [- CONTROL]. # Note the 'ø' glossed as 'nonpast' = non-past tense, somewhat # parallel to the prefix 'm' for 'present tense' on the verb. Mianatra ao an-davarangana i Soa. # M-i-anatra ø-ao an-davarangana i Soa. # Pres-prf-study nonperf-there on-balcony D.sg Soa # ‘Soa, (go and) study on the balcony!'[ an injunction] # This sequence was first presented and accounted for by Randriamasimanana in # his invited talk to AFLA V (Department of Linguistics, University of Hawaii # at Manoa, March 26-29, 1998). # In this second interpretation, the verb 'mianatra' 'pres-study' represents # an injunction: This is an order given to Soa by someone. # It therefore is deemed to contain the feature [+ CONTROL]. # PP 'ao an-davarangana' 'on the balcony' is an argument of the verb. # Note the 'ø' glossed as 'nonperf' = non-perfective aspect. # For further elaboration on this meaning, see Randriamasimanana, Charles. 1985. # Tense/Aspect and Locatability in Malagasy. In Working Papers in Linguistics, # University of Melbourne, Melbourne, Australia. Volume 11, December 1985. # Pp 109-136. Ao an-davarangana no ianaran'i Soa. # ø-ao an-davarangana no i- anar- an' i Soa # nonperf-there on-balcony focus pass-stem.study-pass.by D.sg Soa # "Soa, (go) on the balcony and study there!" # Note the passive voice on the verb 'ianaran(a)' with the fronted PP # argument 'ao an-davarangana'. # This sequence was first presented and accounted for by Randriamasimanana in # his invited talk to AFLA V (Department of Linguistics, University of Hawaii # at Manoa, March 26-29, 1998). # This is the second interpretation of the same verb 'mianatra''pres-study' # containing the feature [+ CONTROL]. # The PP 'ø-ao an-davarangana' 'on-balcony' is therefore an argument of # the verb, which must be in the passive voice when PP is fronted. # Note the 'ø' glossed as 'nonperf' = non-perfective aspect. # The form of the passive is the discontinuous 'i...an(a)'. # The stem 'anar' is derived from the passive imperative 'i-anar-o''pref- # stem.study-by.you' from which the suffix 'o''by.you' is dropped, replaced # by the passive suffix 'ana' of the discontinuous form 'i...ana'. # To see the contrast between a [ + CONTROL and a [ - CONTROL] interpretations # consult Randriamasimanana (2002: 48-49). Amin'ity savony ity no manasa lamba ny vehivavy. # ø-amin' ity savony ity no m- an-(s)asa lamba ny vehivavy. # nonpast-with D.sg soap D.sg focus pres-prf-wash linen Det.unspec women # "All of you women, with this soap (go and) wash linen." # Originally this was a sequence from Edward L. Keenan (1976:271, ex # 63a) whose intended reading was "It is with this soap that the woman # is washing clothes." # For further elaboration, see Randriamasimanana (2002: 55). # However the above reading is impossible as PP '-amin'ity savony ity' # 'with this soap' is an adjunct to the verb 'manasa', which contains # the feature [ - CONTROL]. # Note that the verb 'manasa' is in the active voice. # Note the 'ø' glossed as 'nonpast' = non-past tense, somewhat # parallel to the prefix 'm' for 'present tense' on the verb. Amin'ity savony ity no anasan'ny vehivavy lamba. # ø-amin' ity savony ity no ø- an- (s)as(a)- an' # nonperf-with D.sg soap D.sg focus nonpast-pass-root.wash-pass.by # ny vehivavy lamba. # Det.unspec woman/women linen # "It is with this soap that the women are washing clothes." # The verb 'manasa''pres-wash' has the feature [ + CONTROL]. # As a result PP 'ø-amin'ity savony ity''nonperf-with this soap' # is an argument of the verb. # Note the passive voice 'an...an(a)' on the verb with the fronted PP. # Note the 'ø' glossed as 'nonperf' = non-perfective aspect. # For a similar construction, see Randriamasimanana (2004: 349-350. # Faits de Langues, Editions Orphrys. Paris, France). # However note that this optionality is NOT always available with all verbs. # Ungrammatical Amin’ny langilangy no mamaky ny tavoahangy i Noro.(0! 0 0 0) # ø-amin' ny langilangy no m-am-(v)aky ny tavoahangy # nonperf-with Det.unspec stick focus pres-prf-break Det.unspec bottle(s) # i Noro. # D.sg Noro # 'It's with the stick that Noro breaks the bottle.' (Original translation) # The above sequence has been adapted from Ileana Paul (2000:98, ex 9e). # The verb 'mamaky"'pres-break' caontains the feature [ +CONTROL] and # therefore the PP 'ø-amin'ny langilangy''with the stick' is an argument # of the verb 'mamaky'; when this PP-argument is fronted, the verb must be # in the passive voice. # See additional examples in Randriamasimanana (1986: 462, ex 53a. and 466, # ex 63a. and 63b.) # Ungrammatical Tamin'ny rano no nameno ny tavoahangy i Sahondra.(0! 0 0 0) # T-amin' ny rano no n- am- (f)eno ny tavoahangy # perf-with Det.unspec water focus past-prf-root.full Det.unspec bottle(s) # i Sahondra. # D.sg Sahondra # "It is with water that Sahondra filled the bottle". (original translation) # The above sequence was adapted from Ileana Paul (2000:38, ex 26a). # The verb 'nameno' 'past-fill' has the feature [ +CONTROL]; as a result # the PP 't-amin'ny rano''perf-with the water' is an argument of the verb # 'nameno' and not just an adjunct. # Therefore the fronted PP-argument requires the passive voice on the verb. # See additional examples in Randriamasimanana (1986: 462, ex 53a. and 466, # ex 63a. and 63b.) # Incorporation. A constituent which at first sight appears to be an adjunct # seems to have become an argument of the verb: Verbal aspect is responsible for # this apparent shift. See Comrie 1976 for Aspect & Baker 1988 for Incorporation. # However note that adjacency of the incorporating headverb and its putative # complement is absolutely crucial: The main reason for this is that Malagasy is # not a 'feature spread' language, but rather a 'missing feature' language. # See further elaboration in Malagasy and Formosan Languages: A Comparison. # by Randriamasimanana. (2004:421-422). Tamin'ny herin'ny sabotsy lasa no nialan-dRabe teto. # T-ami-n' ny heri-n' ny sabotsy lasa no # perf-on-of Det.unspec next-of Det.unspec saturday gone focus # n- i- ala- n- d Ra-be t-eto. # past-pass-root.leave-pass-d-Ra-be perf-here. # DURATIVE # "Samedi dernier, il y a huit jours que Rabe etait parti".(Original # translation) English:"As of last saturday, Rabe (familiar) would have # been gone from here a week". # This sentence comes from Jacques Dez (1990: 166, ex 713. Cheminements # Linguistiques Malgaches. Asie et Monde Insulindien 20. Editions Peeters # SELAF 307.Paris. # Note the DURATIVE aspect on the passive voice verb. # The complex verb 'n-iala t-eto''past-leave from here' can accommodate a # DURATIVE interpretation if it involves a DELIBERATE kind of ACTIVITY. # DURATIVE = Internal speaker perspective. See Bernard Comrie (1976: 3-6). # The PP 't-ami-n'ny heri-n'ny sabotsy lasa' 'perf-on-of Det.unspec next-of # saturday gone' is an argument of the verb 'niala' [ + DURATIVE]. # When the PP-argument is fronted, the passive voice is mandatory. Tamin'ny herin'ny sabotsy lasa no niala teto Rabe. # T-ami-n' ny heri-n' ny sabotsy lasa # past-on-of Det.unspec next-of Det.unspec saturday gone # no n-iala t-eto Ra-be. # focus past-leave perf-here Ra-be # PUNCTUAL # "One week before last saturday, Rabe (familiar) left from here". # Note that 'n-iala t-eto''past-leave from here' has PUNCTUAL aspect. The # accompanying Time-PP 't-amin'ny herin'ny sabotsy''one week before last # saturday' is an adjunct, as is usually the case when the verb is not of # DURATIVE aspect. Furthermore the morpheme t- on the PP indicates tense, # not verbal aspect; this is an explicit feature of adjunction. # Note the PUNCTUAL aspect of the active voice verb. # PUNCTUAL = External speaker perspective.See Bernard Comrie (1976: 3-6). # The PP 't-ami-n'ny heri-n'ny sabotsy lasa' 'past-on-of Det.unspec next-of # saturday gone' is an adjunct of the verb 'niala teto' [ - DURATIVE]. # When the PP-adjunct is fronted, the verb remains in the active voice. Niala teto Rabe tamin'ny herin'ny sabotsy lasa. # N- iala t-eto Rabe t- ami-n'ny heri-n' # past-leave perf-here Ra-be(familiar) past-on-of Det.unspec next-of # ny sabotsy lasa. # Det.unspec saturday gone # V S PP # "Rabe (familiar) left from here one week before last saturday". # Note the PP 't-ami-n'ny heri-n'ny sabotsy lasa' 'past-on-of Det.unspec next-of # saturday gone' to the right of the subject 'Rabe': The morpheme t- on PP is a # tense --not aspect-- marker; this suggests that the PP is an adjunct. # In other words this PP is NOT adjacent to its putative complex headverb. Niala teto tamin'ny herin'ny sabotsy lasa Rabe. # N-iala t-eto t-ami-n' ny heri-n' ny # past-leave perf-here perf-on-of Det.unspec next-of Det.unspec # sabotsy lasa Rabe. # saturday gone Ra-be(familiar) # V PP S # "Rabe (familiar) left from here one week before last saturday". # Note the PP 't-ami-n'ny heri-n'ny sabotsy lasa' 'perf-on-of Det.unspec next-of # saturday gone' to the left of the subject 'Rabe'; the morpheme t- on the PP # is a verbal aspect-marker. # This means that the PP is now adjacent to the complex headverb 'niala teto', # which if it contains the feature [ + DURATIVE] can incorporate the PP. # The PP then becomes a PP-argument. # Ungrammatical Ny fitiavana no namonoany tena.(0! 0 0 0) # Ny fitiavana no n- am- (v)ono- a- ny tena. # Det.unspec love focus past-pass-root.kill-pass-by.him/her self # "He killed himself for love" (Original translation) # This sequence has been adapted from Matthew Pearson (2001:33, ex 30c). # Note the attempt to passivize on the PP-adjunct 'noho ny fitiavana', which # as shown on the sentence immediately below is located to the right of the # grammatical subject, i.e. NOT adjacent to its putative incorporating head. # See further elaboration in Malagasy and Formosan Languages: A Comparison. # by Randriamasimanana. (2004:421-422) as to why adjacency is required in # Malagasy for Incorporation. Namono tena izy noho ny fitiavana. # N-am-(v)ono tena izy noho ny fitiavana. # past-prf-kill self s/he because.of Det.unspec love # V S CAUSAL-PP # "S/he killed him/herself because of love." # This sequence comes from 'Malagasy and Formosan Languages: A Comparison' # by Randriamasimanana (2004: 422, ex 28a). # Note the occurrence of 'noho ny fitiavana' 'because of love' to the right # of the subject 'izy''s/he', suggesting that it is more than likely to be # an adjunct. # Ungrammatical Namono tena noho ny fitiavana izy .(0! 0 0 0) # N-am-(v)ono tena noho ny fitiavana izy. # past-prf-kill self because.of Det.unspec love s/he # V CAUSAL-PP S # "S/he killed him/herself because of love." # Note the PP adjacent to the verb 'n-am-(v)ono''past-kill'. Amin'ny alarobia no handehananay. # ø- ami-n'ny alarobia no h- an- deha-na- nay. # nonperf-on-of Det.unspec wednesday focus fut-pass-go- pass-by.us # DURATIVE # "We will leave on Wednesday". (Original translation) In fact "It is on # Wednesday we will be leaving". # This sequence is from Matthew Pearson (2001: 33, ex 30a). # Note the DURATIVE aspect associated with the passive voice verb. # The PP 'ø-amin'ny alarobia''non-perf-on-of Det.unspec wednesday' is # an argument of the verb 'h-andeha' [ + DURATIVE]. # When the PP-argument is fronted, the verb must be in the passive voice. Amin'ny alarobia no handeha izahay. # ø- ami-n'ny alarobia no h- an-deha izahay. # nonperf-on-of Det.unspec wednesday focus fut-prf-go we.exclusive # PUNCTUAL # "It is on Wednesday that we will leave". # Note the PUNCTUAL aspect associated with the active voice verb 'handeha' # [ - DURATIVE]. # The PP 'ø-amin'ny alarobia''non-perf-on-of Det.unspec wednesday' is # an adjunct of the verb 'handeha' [ - DURATIVE]. # When the PP-adjunct is fronted, the verb remains in the active voice. Handeha izahay amin'ny alarobia. # h- an-deha izahay ø- ami-n'ny alarobia. # fut-prf-go we.exclusive nonperf-on-of Det.unspec wednesday # V S PP # "We will leave on wednesday". Handeha amin'ny alarobia izahay. # h- an-deha ø- ami-n'ny alarobia izahay . # fut-prf-go nonperf-on-of Det.unspec wednesday we.exclusive # V PP S # "We will leave on wednesday". # This sequence is marginally grammatical, as a marked structure. # Assessment of recent 'Malagasy' Control structures: # Finally a so-called Backward Control theory has been proposed for # Malagasy and a movement theory of Control has been based on some crucial # 'Malagasy' data, with dire consequences for the structure of Malagasy as # a natural language. My purpose here is to show that many of the proposed # sentences are simply irretrievably ungrammatical. # What follows is intended to be merely illustrative and does not constitute # an exhaustive listing of all published 'Malagasy' Control sequences. # Irretrievably ungrammatical sequence from (Law (1995), (8)): kasain-dRasoa ho-sasa-ko ny zaza (0! 0 0 0) # intend.PASS.by-Rasoa FUT.wash.PASS.by-me the child # 'It is intended by Rasoa that the child will be washed by me.' # The above sentences is NOT Malagasy, or to put it plainly, it does NOT exist # either in standard Malagasy nor in any of the known dialectal variety for # exactly the same reasons as the ones mentioned below relative to the first set # of 'Malagasy' data, which I wish to consider at the conclusion of this document # on Control in Malagasy. # To help the reader understand to what extent the above sequence is totally alien # to Malagasy, I reproduce immediately below one crucial example using the same # passive voice predicate borrowed from Rajemisa-Raolison (1995:481). Rakibolana # malagasy. [ Malagasy encyclopedia ] Ambozontany, Fianarantsoa, Madagascar. # Rajemisa-Raolison is an authority as far as Malagasy language and grammar is # concerned. Here is one crucial example from him: Raharaha inona no kasainao hatao rehefa lehibe? # Raharaha inona no ø- kasa- in(a) /nao # occupation what focus pass.pres-stem.kasa-suff.ina/clitic.by.you # h-a- tao rehefa lehibe? # fut-pass.pref.a-root.tao when grown-up # Lit.:"What occupation is intended by you will be done -[ no clitic]- when grown-up?" # English:"What will you do when you grow up?" # adapted from Rajemisa-Raolison. 1995:481. # In both (Law (1995)(8)) and Rajemisa-Raolison (1995:481), we have a Malagasy Double Passive # contruction involving the matrix passive voice form 'kasaina''ø-stem.kasa-suff.ina''is.intended'. # In (Law (1995)(8)) a clitic,i.e. 'ko''by.me' is appended to the embedded passive verb and # this has the immediate effect of rendering the sequence irretrievably ungrammatical. # On the other hand, in Rajemisa-Raolison (1995:481), only the matrix passive voice verb takes # a clitic; the embedded passive voice verb, i.e. here 'h-a-tao''future-passive.prefix.a-root.tao' # does NOT take a clitic at all: The sentence involving a double Passive in Rajemisa-Raolison # (1995:481) is perfectly grammatical. One simple explanation for this difference in grammaticality # lies in the fact that the Malagasy predicate 'm-i-kasa''present-active.prefix.i-root.kasa' # requires the same grammatical subject for the matrix as well as the embedded verbs. In its # passive voice counterpart, we have ONE genitive clitic on the matrix passive and NO clitic # at all on the embedded passive predicate. # My immediate concerns here are the following two sets of 'Malagasy' sentences # proposed as typical Control structures for this language. It should be noted once # again that the sentences in (28)a., (28)b., (3)a. & (3)b. below are NOT Malagasy # and simply canNOT be assigned any meaning at all. # I. First set of data: # From a handout given at Stanford University, Workshop on Austronesian syntax # and semantics, February 5, 2005. Some Remarks on Austronesian Clause Structure: # Implications from Malagasy Control Complements. Maria Polinsky, University of # California, San Diego & Eric Potsdam, University of Florida. # Apparently (28)a. & (28)b. were already presented on February 27, 2004 by Eric # Potsdam as sequences (8)a & (8)b. on his handout entitled 'Patterns of Control # in Malagasy and Their Theoretical Implications', Department of Linguistics, # University of Maryland. # Irretrievably ungrammatical sequences where the original numbers, glosses and # translations are kept intact from the February 5, 2005 Stanford University handout: mikasa ny mpianatra [fa izaho no hangalatra ny toaka] (0! 0 0 0) #(28) a. mikasa ny mpianatra [fa izaho no hangalatra ny toaka] CP (0! 0 0 0) # intend the student that I FOCUS steal the booze # 'The student intends that I steal the booze.' mikasa ahy [hangalatra ny toaka] ny mpianatra (0! 0 0 0) # b. mikasa ahy [hangalatra ny toaka] ny mpianatra SOR (0! 0 0 0) # intend me steal the booze the student # 'The student intends me to steal the booze.' # Neither (28)a. nor (28)b is Malagasy. In other words, and to put it rather plainly # sentence (28)a. and sentence (28) b. simply do NOT exist in Malagasy, either in the # standard language nor in any of the known dialectal varieties. #'M-i-kasa' 'present-prefix.i-root.kasa' is a Subject Control predicate i.e. Equi-1 # or Same Subject Control structure in the terminology used in The Causatives of # Malagasy, University of Hawaii Press, Honolulu, Hawai 1986, which not only constituent # -selects the future tense-marker on the embedded verb, i.e in the present case # 'h-angalatra''future-will.steal', but which can also accommodate a complementiser like # 'ny' but which NEVER subcategorizes for a factive complementiser like 'fa'. Furthermore # a Subject Control predicate does NOT behave like a Verb of Saying and NEVER allows its # sentential complement to be separated from its active voice head verb by a grammatical # subject.[ Also contra Polinsky, M. & E. Potsdam, April 23, 2006 Abstract (ex.5) for Zentrum # für Allgemeine Sprachwissenschaft, Typologie und Universalienforschung (ZAS) - Workshop # Control verbs in cross-linguistic perspective ]. # The properties of Subject Control predicates alluded to above were already described # in some detail and in English under the label 'Equi-1' in 'The Causatives of Malagasy', # University of Hawaii, Honolulu, Hawaii (1986: 495-509), which is conveniently ignored # by Maria Polinsky, Department of Linguistics, University of California, San Diego, USA # & Eric Potsdam, Program in Linguistics, University of Florida, Gainesville, Florida, USA. # Also, consult the on-line document relative to Verbs of Saying on the Oxford University # website dedicated to Malagasy Language & Linguistics and available since 2005 at # http://users.ox.ac.uk/%7Ecpgl0015/pargram/data/Vsay-testfile.txt. # In addition, is available from both of the websites cited earlier a paper called # Randriamasimanana, Charles. 2002. ‘Binary Branching and Null Subjects in Malagasy’. # (Volume 3, Issue 1, 2002). Language and Linguistics, Institute of Linguistics # (Preparatory Office), Academia Sinica, Taipei, Taiwan. In its section 4.2 AGR and # Control Structures, we have a different example (28)a with another Subject Control # predicate. # Under Point # 3.3 of Part1/March 30, 2001 Academia Sinica handout on the Malagasy # Verbal Voice System, examples (15) and (18) illustrate the use of the Subject Control # predicate 'mikasa''present-intend'. If the reader wants to study those Malagasy sentences, # s/he can go either (a) to http://www.ratsimandresy.org/book_ls.html and click on the # following title 'Ch.4 - Malagasy Verbal voice System - Part 1, March 30, 2001 handout' # to download the corresponding pdf-formatted paper, (b) alternatively, s/he can go to # the other site at http://folk.uio.no/janengh/gassisk/ and click on the following title # '4.Malagasy Verbal voice System - Part 1. March 30, 2001' to download the relevant paper. # II.Second set of data: # From an abstract entitled 'Backward Object Control:  Against an Empty Category Analysis' # submitted by Eric Potsdam & accepted by West Coast Conference on Formal Linguistics 25, # presentation scheduled for April 28, 2006. # Apparently similar sequences were presented by Eric Potsdam, University of Florida & # Maria Polinsky, University of California, San Diego at AFLA 12, Department of Linguistics # University of California at Los Angeles, California, USA on Sunday May 1, 2005 # on a handout entitled 'Finite Control in Malagasy': The sequences involving the predicate # 'nampahatsiahy' were the irretrievably ungrammatical (18)b, the irretrievably ungrammatical # (24)b., the irretrievably ungrammatical (50)b., the irretrievably ungrammatical (51)b, # and the irretrievably ungrammatical (59)b. # In addition, of course, to support the so-called 'Backward control theory' for Malagasy, # Polinsky, M. & E. Potsdam in their April 23, 2006 Abstract for their paper 'Control and # control-like constructions in Malagasy' accepted by Zentrum für Allgemeine Sprachwissenschaft, # Typologie und Universalienforschung (ZAS), Jägerstraße 10-11, 10117 Berlin (Mitte)Berlin, # Germany. Workshop Control verbs in cross-linguistic perspective, show 'Passive Backward # Control Malagasy' sentence (6). Consult the relevant abstract at this website: # http://www.zas.gwz-berlin.de/home/stiebels/kontrollverbenabstracts/polinsky.html # At any rate, concentrating on our second set of recent 'Malagasy' examples, we note the # following irretrievably ungrammatical Malagasy sequences where the original numbers, glosses # and translations are kept intact from the April 28, 2006 abstract submitted by Eric Potsdam # and accepted by West Coast Conference on Formal Linguistics 25, Department of Linguistics # University of Washington, Seattle, WA 98195-4340, USA. nampahatsiahivin- dRasoa ahy [hohidiana ny varavarana ] (0! 0 0 0) #(3) a. nampahatsiahivin- dRasoa ahy [hohidiana ny varavarana ] ORDINARY CONTROL (0! 0 0 0) # remind.OT Rasoa me lock.TT the door nampahatsiahivin- dRasoa [hohidia-ko ny varavarana ] (0! 0 0 0) # b. nampahatsiahivin- dRasoa [hohidia-ko ny varavarana ] BACKWARD CONTROL (0! 0 0 0) # remind.OT Rasoa lock.TT I the door # '€˜Rasoa reminded me to lock the door.' # Neither (3)a. nor (3)b is Malagasy: The passive voice predicate 'nampahatsiahivina' # requires either a nominal Definite Phrase Subject, i.e. with Determiner 'ny' or a Sentential # Subject with obligatory FACTIVE Complementiser 'fa', as will be illustrated immediately below. # Malagasy Factive Comp 'fa', of course, casts serious doubts on so-called 'Finite Control' # structures posited in Maria Polinsky, University of California, San Diego & Eric Potsdam, University # of Florida,'Finite Control in Malagasy', handout given Sunday May 1, 2005.AFLA 12, UCLA as well as # in Polinsky, M. (UC San Diego) and E. Potsdam (University of Florida). Control and control-like # constructions in Malagasy. Abstract of paper scheduled for April 23, 2006. # Malagasy translation of equivalent structure ‘Rasoa recalled that I should lock the door.’ # Fully grammatical Malagasy sequence: nampahatsiahivin-dRasoa fa tokony hidiko ny varavarana. #(3')a. nampahatsiahivin-dRasoa fa tokony hidiko ny varavarana. # n- amp-aha-tsiahiv- in- d- Rasoa # past-amp-aha-stem.tsiahiv-suffix.in(a)-ep.d-Rasoa # fa tokony h-id(i)-ina/ko ny varavarana. # FACTIVE.Comp.fa should fut-stem.hidi-suff.ina/clitic.by.me Det door # Lit."Was recalled by Rasoa that should be locked by me the door." # English:"It was recalled by Rasoa that I should lock the door." # Note that in (3')a., the passive voice predicate 'nampahatsiahivina''was.recalled' # takes a Sentential Subject, i.e. "fa tokony hidiko ny varavarana''that should be locked # by me the door" with an obligatory FACTIVE Complementiser 'fa'. # Malagasy translation of equivalent structure ‘Rasoa brought to my attention the fact # that the door should be locked.’ Rather doubtful grammatical Malagasy sequence, but # acceptable in Colloquial Malagasy: nampahatsiahivin-dRasoa ahy fa tokony hidina ny varavarana. # "I was reminded by Rasoa that the door should be locked." #(3')b. nampahatsiahivin-dRasoa ahy # n- amp-aha-tsiahiv- in- d- Rasoa ahy # past-amp-aha-stem.tsiahiv-suffix.in(a)-ep.d-Rasoa acc.me # fa tokony hidina ny varavarana. # fa tokony h-id(i)-ina ny varavarana. # Comp.fa should fut-stem.hidi-suff.ina Det door # Lit."Was recalled by Rasoa to me that should be locked the door." # English:"I was reminded by Rasoa that the door should be locked." # The structure shown in (3')b still has the passive voice predicate # 'nampahatsiahivina''was.recalled' with the Sentential Subject "fa tokony # hidina ny varavarana''that should be locked the door." # However, note that in addition it also involves Incorporation (as explained # earlier on in this document) and relies on the advancement of an initial # Oblique-Adjunct to a derived Direct Object position, i.e. 'ahy''accusative.me". # Crucially, the sentence in (3')b. does NOT mean at all that I will necessarily # be the one to lock the door! # There is some evidence to substantiate the claim about (3')b. from the following piece # of Malagasy data borrowed from http://fr.groups.yahoo.com/group/serasera/message/25806. # Note that in this particular case, the Sentential Subject comprises the nominalising # Complementiser 'ny' --different from the 'Subject Control' Comp 'ny' -- instead of the # usual FACTIVE Complementiser 'fa'. # De : "Rabesata Arima" # Date : Vendredi 10, Février 2006 7:02 # Objet : FOTOANA GASY ve ? rabesata@... Mety tokoa ny hanazavana azy ireny sy ampahatsiahivina amintsika mianakavy! # English: "It is indeed proper that those will.be.recalled to.us as.a.family." # Mety tokoa ny hanazavana azy ireny sy ampahatsiahivina amintsika mianakavy! # Mety tokoa ny h-ana-zav(a)-ana azy ireny # Mety tokoa ny h-ana-zav(a)-ana azy ireny # "suitable indeed Comp.ny fut-Caus-stem.(a)zava-suff.ana them.nonhuman.pronoun" # sy amp- aha-tsiahiv- ina amintsika mianakavy! # sy ø-amp-aha-stem.tsiahiv-suff.ina ø-amintsika family # and is.being.recalled nonpast-to.us family #" and is.being.recalled to us.as.a.family." # English: "It is indeed proper that those will.be.recalled to.us as.a.family." # Note the Passive.1 form with suffix.ina and THEME Sentential Subject # 'ny hanazavana azy ireny', with optional Oblique-ADJUNCT 'amintsika mianakavy' # 'nonpast-to.us as.a.family' attached to the passive predicate 'ampahatsiahivina'. # As pointed out relative to Preposition Phrases in Malagasy, a tense-marker # is an absolute barrier to Incorporation. However, the morpheme ø is # ambivalent between a nonpast tense-marker and a nonperfective aspect-marker. # This is one area of Malagasy grammar where we would expect some variation; # but unfortunately NOT to the extent portrayed in Eric Potsdam (April 28, 2006). # Last but not least, note that Comp 'ny' in the sequence above is different from # Comp 'ny' constituent-selected by a Subject Control predicate: In the case of # a Subject Control predicate, Comp 'ny' accommodates an active voice verb in the # embedded position, i.e. comprising a root and NOT a stem; whereas in the case of # nominalising Comp 'ny', as used in 'ny hanazavana azy ireny', the accompanying # embedded verb canNOT be in the active voice; it MUST be in the Passive.2 voice, # i.e. comprise a stem and NOT a root. # Here is a prototypical use of the predicate 'ampahatsiahivina', where it takes # a Sentential Subject, borrowed from the official site of the 'Eglise Evangélique # luthérienne de France, Paroisse protestante Saint-Matthieu, Pontault-Combault, # France' at http://www.saintmatthieu-protestant.org/culte_malagasy.html Ampahatsiahivina fa misokatra hoan’ny rehetra ny fotoam-pivavahana amin'€™ny teny frantsay isan'alahady maraina amin'€™ny 10 h 30. # "(It) is recalled that is open to all the religious service # in the French language every Sunday morning at 10:30." # Ampahatsiahivina fa misokatra hoan’ny rehetra ny fotoam-pivavahana # ø- amp-aha-tsiahiv- ina fa m-isokatra # pass.pres-amp-aha-stem.tsiahiv-suff.ina Comp.fa pres-be.open # "Is/are being.recalled that open/s.as.a.process" # ho an'€™ ny rehetra ny fotoam-pivavahana # prep.for Det all.pro Det religious.service # "for all the religious service(s)." # "Is/are.being.recalled that open.process to all the religious service(s)" # (amin’ny teny frantsay) isan’alahady maraina amin’ny 10 h 30. # (ø-amin'€™ ny teny frantsay) isan'€™ alahady maraina ø-amin'€™ ny 10 h 30. # (nonpast-in Det language French) every sunday morning nonpst-at Det 10 h 30. # "(in the French language) every sunday morning at 10 h 30." # This piece of Malagasy data clearly casts serious doubts on Eric Potsdam # (April 2006, examples 3.a & 3.b) as Potsdam posits a human THEME Direct Object, # where none is required in addition to the fact that Comp 'fa' is mandatory # in the Sentential Subject of the passive predicate 'ampahatsiahivina'. # Here is another prototypical use of the predicate 'ampahatsiahivina', where it # takes a nominal DP Subject, i.e. with Determiner 'ny' from the following Malagasy # language site to be found at http://fr.groups.yahoo.com/group/serasera/message/25806 De : FPMC Date : Vendredi 24, Fé©vrier 2006 21:30 Objet : JOBILY FAHA-10 TAONA, VOLANA FAHAROA fpm_c Indro averina ampahatsiahivina etsy ambany ny fandaharam-potoana mandritra ny taona 2006. # English "Below is given once again, recalled the timetable for the year 2006." # Indro ø- a- verina ø- amp-aha-tsiahiv- ina # here pass.pres-pref.a-root.verina pass.pres-amp-aha-stem.tsiahiv-suff.ina # "Here is being.repeated, being.recalled" # ø-etsy ambany ny fandaharam-potoana # nonpast-there below ny f-andahara(n>m)(a)-(f>p)otoana # # there below Det nominaliser.f-stem.andahar-suff.ana schedule # mandritra ny taona 2006. # during Det year 2006 # "for the year 2006." # Relevant & simplified structure without the adjuncts: ampahatsiahivina ny fandaharam-potoana # "The timetable is being recalled." # ø- amp-aha-tsiahiv- ina ny fandaharam-potoana ... # pass.pres-amp-aha-stem.tsiahiv-suff.ina Det schedule # "Is being recalled the schedule." # In this instance, we have Passive.1 with suffix.ina and where DP Subject # 'ny fandaharam-potoana''the schedule' is a nominal THEME, which by the way # is NOT human. # In the active voice, the predicate 'mampahatsiahy' also requires a non-human # THEME Object, as is obvious from the following title from the on-line site of # 'L'Express de Madagascar', Actualité© du 10 mai 2004 n°2789 from the website at # http://www.lexpressmada.com/ Nampahatsiahy ny fanajana ny fitsipi-pifehezana ny minisitra Razakanirina Lucien Victor. # English:"Minister Razakanirina Lucien Victor recalled # the importance of the respect for the rule of law." # N- amp-aha-tsiahy ny fanajana ny fitsipi-pifehezana # past-amp-aha-root.tsiahy Det respect Topic regulations # "Recalled the respect for regulations" # ny minisitra Razakanirina Lucien Victor. # Det minister Razakanirina Lucien Victor # "the minister Razakanirina Lucien Victor." # Lit.:"Recalled the respect for regulations the minister # Razakanirina Lucien Victor." # English: "Minister Razakanirina Lucien Victor recalled the # importance of the respect for the rule of law." # In this sequence, the verb 'mampahatsiahy' is in the active voice # and its direct object is a non-human THEME. # This piece of genuine Malagasy data clearly casts serious doubts on Eric Potsdam # (Abstract, April 2006, example 6), for instance: Not only is Complementiser 'fa' # required by the predicate absent from Potsdam sentence (6), but not even the non-human # nature of the THEME Direct Object is respected. Relative to sentence (5) on his abstract, # suffice to say that it is NOT even a 'plausible' Malagasy sentence.