1. Two modes of case-assignment: evidence from two stages of Greek
Elena Anagnostopoulou & Christina Sevdali
University of Crete Ulster university
13th International CONFERENCE IN GREEK LINGUISTICS
London, 7-9TH September 2017, UNIVERSITY OF Westminster

2. The issues
Properties of dative/genitive objects in Classical and Standard Modern Greek (CG vs. SMG) as a window to the nature of dative and genitive case cross linguistically;
The transition from one system of case assignment to another;
The relationship between the diachrony of morphological case and the diachrony of prepositions.

3. In this talk
We will first discuss the known CG facts (Anagnostopoulou & Sevdali, 2015) on the environments and behaviour of argumental datives and genitives;
Then we will focus on the properties of the SMG system;
We will propose an analysis whereby CG dative/genitive is best analysed as lexical/prepositional case, while SMG genitive is best analysed as dependent case in the sense of Baker (2015);
We will discuss a consequence of this proposal vis á vis the well-known property of CG of passivization of the indirect object (IO);
We will present an overview of the diachrony of morphological case and prepositions in Greek and outline a path of change.

4. Classical Greek
Five morphological cases (nominative, accusative, dative, genitive and vocative);
The case system of CG is the result of a reduction from the original Indo-European (IE) system (Luraghi, 2003 among many others), something that has led to CG cases being syncretic: the genitive had resulted from merging the IE genitive and ablative, and the dative was the result of merging the IE dative, locative and instrumental;
Nominative case is reserved for subjects of finite clauses;
Accusative is the most common case for objects (Delaunois, 1988);
The choice of dative or genitive case can be partially reduced to semantic generalizations (Luraghi, 2010: 64-67; Anagnostopoulou & Sevdali, 2015: ) and are also partially idiosyncratically determined by particular verbs or prepositions (but see also Stolk, 2015 for a different view that focuses on different “construal of events” among different predicates).

5. Classical Greek Mono-transitives:
There are verb classes selecting for dative DP objects:
(e.g. armozo: ‘is appropriate’, omoiazo: ‘resemble’, isoumai ‘be equal to’, epikouro: ‘assist’, timo:ro: ‘punish’, peithomai ‘trust, obey’, epomai ‘follow’, and CRUCIALLY complex verbs prefixed with the prepositions en- ‘in’, sun- ‘with’, epi- ‘on’, para- ‘next to’, hupo- ‘under’, and the adverb omou ‘similarly);
There are other verb classes selecting for genitive DP objects:
(e.g. mimne:iskomai ‘remember’, archo: with the meaning ‘begin’, epimelomai ‘take care of’, epithumo: ‘want, desire’, ero: ‘love’, steromai ‘lose’ etc. ).
The number of mono-transitive verbs selecting for non-accusative objects in CG is large.
For datives: Verbs denoting appropriateness, Equality/agreement, Friendly or adversarial feeling or action, Persuasion, submission, meeting and Complex verbs with the prepositions en- ‘in’, sun- ‘with’, epi- ‘on’, para- ‘next to’, hupo- ‘under’, and the adverb omou ‘similarly’ (omo-noo: ‘agree’, sun-eimi ‘coexist’, sun-oiko: ‘cohabit’, sum-pratto: ‘assist’, emmeno: ‘inhabit’, em-pipto: ‘attack’, epi-cheiro: ‘attempt’, par-istamai ‘present’, hupo-keimai ‘be placed below’, etc.)
For genitives: Memory (mimne:iskomai ‘remember’, epilanthanomai ‘forget’, etc.).
Beginning/ending (archo: with the meaning ‘begin’, pauomai ‘finish’).
Taking care of (epimelomai ‘take care of’, amelo: ‘neglect’, kataphrono: ‘look down upon’, etc.).
Wanting, enjoyment, being part of (epithumo: ‘want, desire’, ero: ‘love’, koino:no: ‘have a share of, take part in’, etc.).
Losing, needing (steromai ‘lose’, aporo: ‘wonder’, deo:/deomai ‘need’)
Feeling/perception (aptomai ‘touch’, akouo: ‘listen’, etc.).
Attempt, success/failure (peiro:/peiromai ‘try’, apotugchano: ‘fail’, etc.).
Ruling (archo: with the meaning ‘rule, govern’, turanno: ‘be a monarch’).
Comparison ( pleonekto: ‘exceed’, pro:teuo: ‘come first’ , meionekto: ‘be worse than’, etc.).

6. Classical Greek
Case arrays in Ancient Greek ditransitives (Anagnostopoulou & Sevdali 2015: 456)
Accusative IO – Accusative DO (e.g. ero:to: ‘ask’, didasko: ‘teach’, o:phelo: ‘benefit’ etc. )
Dative IO – Accusative DO (e.g. hupischnoumai ‘promise’, dido:mi ‘give’ and complex verbs with the prepositions epi- ‘on’, en- ‘in’, sun- ‘with’ like epitasso:‘assign/enjoin’, epitrepo: ‘entrust/transfer’ etc. )
Genitive IO – Accusative DO (e.g. estio: ‘feed’, apotemno: ‘cut off’, lambano: ‘receive’, akouo: ‘listen’ etc.)
Dative IO – Genitive DO (e.g. phthono: ‘envy’, metecho:/koino:no: ‘take part in’, metadido:mi ‘transmit’ etc.)
Dative and genitive in the different arrays are subject to thematic and morpho-syntactic generalizations (Anagnostopoulou & Sevdali 2015: 457): for example, goals tend to be dative, sources and possessors tend to be genitive, and verbs prefixed by dative or genitive assigning prepositions must assign dative or genitive to the goal.
NB. ( Following standard practice in the literature, we call themes DOs, and goals, sources, possessors IOs).

7. Standard Modern Greek Basic characteristics of the SMG system:
Reduction of morphological cases (loss of dative);
Systematic replacement of dative by either genitive or accusative depending on the syntactic environment (ditransitives vs. transitives) and the dialect (Southern Greek vs. Northern Greek) (Michelioudakis and Sitaridou, 2013);
Accusative surfaces on almost all objects of transitive verbs in all dialects of Modern Greek and the vast majority of the verbs that selected for dative and genitive objects in CG now take accusative objects. Notice that the examples in (1) feature exactly the same verb:
(1) a. Ho Odusse-us ephthon-e:se Palame:d-ei dia sophia-n.
the Ulysses-nom envy-3sg.aor.act Palamedes-Dat because wisdom
‘Ulysses was jealous of Palamedes because of his wisdom.’ (CG)
b O Odiseas fthonise ton Palamidi gia tin sofia tu
the Ulysses-nom envy-3sg.aor.act Palamedes-Acc because the wisdom his
‘Ulysses was jealous of Palamedes because of his wisdom.’ (SMG)

8. Standard Modern Greek
Re ditransitives, MG has a dialect split on the case realization of the IO:
Northern Greek (NG) has ditransitives where both objects bear accusative morphology (Dimitradis, 1999, inter alia )
Central and Southern Greek (e.g. the dialects spoken in Athens, Peloponnese and many of the islands), as well as in Standard Modern Greek (SMG; based on Southern dialects, see Mackridge, , 2009), the IO bears genitive and the DO accusative, (Anagnostopoulou, 2003, Michelioudakis, 2012, Georgala, 2012, i.a.):
(2) Edhosa ton Petro / tu Petru ena paghoto NG/SMG
Gave-1sg.pst.act the Peter-Acc /Peter-Gen an icecream-Acc
‘I gave Peter an ice cream.’

9. SMG vs. CG datives/genitives: different systems
Difference #1: very few SMG mono-transitives assign genitive to their complements, including ones prefixed with archaic prepositions:
(3) Tilefonisa/milisa tu Petru
Called/ talked.1sg.pst the Peter-gen
‘I called Peter/ talked to Peter’
(4) O proedros iper-aminthike tis epilogis tu
The president- nom defended the choice his-gen
‘The president defended his choice’
 This is not the case for CG (cf. slide 5)

10. SMG vs. CG datives/genitives: different systems
Difference #2: Genitive is not related to a specific theta-role
Genitive has been generalized to all IOs in SMG, regardless of whether they are goals (with ‘give’), sources (with ‘steal’) or beneficiaries (with ‘buy’)
(5) Edhosa tis Marias to vivlio Goal
Gave-1sg the Mary-Gen the book-Acc
‘I gave Mary the book’
(6) Eklepsa tis Marias to vivlio Source
Stole-1sg the Mary-Gen the book-Acc
‘I stole the book from Mary’
(7) Eftiaksa tis Marias pagoto Beneficiary
Made-1sg the Mary-Gen ice cream-Acc
‘I made Mary ice cream’
 This is not the case in CG (cf. slide 6)

11. Lexical/Inherent case cannot account for the CG vs. SMG differences
Anagnostopoulou (2003, 2005), Michelioudakis (2012) and Georgala (2012) all analyze SMG IOs as bearing inherent/quirky Case assigned by an applicative head in a structure like (8), in agreement with Woolford (2006) and others:
(8) vapplP 3
IO-GEN vappl’
vappl’ RootP
Root DO-ACC

12. CG vs. SMG differences
But on the basis of Woolford’s criteria CG dative/genitive is also lexical/inherent, since:
In monotransitives they are idiosyncratically assigned by particular verbs and prepositions (lexical);
In ditransitives they are inherent Cases, as they are systematically associated with specific thematic roles, such as ‘goal’, ‘source’, ‘possessor’ (inherent).
But if both CG and SMG have inherent dative/genitive Case, then the differences discussed above are accidental.
Syncretism due to loss of morphological dative played a role but cannot explain sensitivity to the environment, i.e. why genitive was generalized as the regular case for IOs and accusative was generalized as the regular case for DOs in SMG.

13. Our proposal
There are two ways of assigning dative case cross-linguistically:
As lexical/inherent case á la Woolford (2006); following many authors in the literature we assume that this is case assigned by a zero P (Rezac 2008, Pesetsky 2013, Baker 2015 and many others).
Configurationally á la Baker (2015) building on Marantz (1991): with dative as dependent case. According to Baker, the crucial property of dative is that it is assigned in opposition to a lower argument in the VP domain, unlike accusative which is assigned in opposition to a higher argument (the external argument, EA) in the IP/CP domain.
General Dependent Case rule (adapting Marantz 1991) (Baker 2015: 79, 111)
(9) If XP bears c-command relationship Y to ZP in local domain WP, then assign case V to
XP.
For Dative: (Baker 2015: 131)
(10) If XP c-commands ZP in VP, then assign U (dative) to XP
NB. We assume that XP and ZP in (10) are DPs and VP in (10) is vAPPLP (see tree (8), slide (11)).

14. Our proposal
Proposing that dative/genitive is assigned differently in CG from SMG can prima facie explain the differences between the two systems;
 but it opens new questions:
What is the actual independent evidence synchronically that SMG genitive is dependent case;
What about passivisation of IO in CG (Anagnostopoulou & Sevdali 2015), a hallmark property of structural case?
In the rest of the talk, we first focus on evidence for dependent case from SMG;
Then we discuss the passivisation facts ;
And finally we move to the transition between the two systems and how the diachrony of morphological case interacts with changes in prepositions in Greek.

15. SMG genitive as dependent case
Baker & Bobaljik (2017) distinguish between dependent and inherent case based on two types of evidence:
This case must be sensitive to the presence of another argument, and NOT to a specific theta role;
Lack of this case must be reduced to a lack of another nominal in the local domain.
We already showed data that provide an argument in favour of the first point (slide 10) where we saw that SMG IOs in ditransitives appear in genitive regardless of their theta role (source, goal, beneficiary) contrary to what happens in CG.
Further indicative evidence were seen to come from the lack of productivity of genitive case in monotransitives and the replacement of CG datives and genitives by accusatives in this context.
In the next section we provide some further arguments that show that the SMG genitive is indeed sensitive to the presence of a lower argument in the vAPPLP domain.

16. SMG genitive as dependent case
Argument #1: Dyadic unaccusative verbs with genitives and nominatives
Psychological predicates corresponding to Italian ‘piacere’ verbs (Belletti & Rizzi 1988, i.a., see Anagnostopoulou 1999 for SMG), sensation predicates, different types of possessive and change of location verbs productively select for a genitive experiencer, possessor or goal argument and a nominative theme, as shown in (11) and (12):
(11) Tu Petru tu aresi i musiki
The Peter-gen cl-gen please-3sg the music-nom
‘Peter likes music’
(12) Tu Petru tu xriazete/lipi enas anaptiras
The Peter-gen cl-gen need-3sg/lack.3sg a lighter-nom
‘Peter needs/lacks a lighter’

17. SMG genitive as dependent case
The vAPPLP in these constructions contains a lower theme object introduced at the Root-level.
 Genitive case is assigned in opposition to the lower argument
Case assignment proceeds just as in ditransitives except that the theme bears nominative morphology since there is no EA unlike in ditransitives where a transitive v/Voice head is present (Voice in Alexiadou et al. 2006, 2015, Schäfer 2008, i.a.). GEN is assigned in opposition to the lower argument not bearing case at the vAPPLP level. At the TP/CP level, the DO is assigned environment- sensitive (unmarked) NOM case.
(13) vapplP 3
Exp/Poss/Goal-GEN vappl’
vappl’ RootP
Root Theme-NOM

18. SMG genitive as dependent case
Argument # 2: Genitive alternates with nominative in the absence of a theme argument with sensation verbs like ‘hurt’, ‘be cold’ etc. The existence of such alternations presents evidence that GEN is not linked to a particular theta-role and that its assignment depends on the presence of a lower argument:
(14) O Janis ponai /krioni
The Janis-nom hurt-3sg / is cold-3sg
‘Jianis hurts’ / ‘Jianis is cold’
(15) Tu Jani tu ponai o lemos tu
The Peter-gen cl.gen hurt-3sg the throat-nom-his-GEN
‘Jianis has a sore throat’
(16) Tu Jani tu kriosan ta cheria tu
The Peter-gen cl.gen are cold-3pl the hands-nom-his-GEN
‘Jianis’s hands got cold’

19. SMG genitive as dependent case
The experiencer receives dependent genitive in opposition to the lower vP-internal o lemos tu ‘his throat’ ta cheria tu ‘his hands’;
On the other hand, the experiencer is the single argument of the Root+vAPPL complex and receives unmarked/ environment sensitive Nom since dependent genitive cannot be assigned, because crucially there is no lower argument to be assigned in opposition to.
Note that the lower NOM argument in (15) and (16) contains a possessor, providing evidence against a possessor-raising analysis of these constructions (clitic doubling within DPs is impossible in Greek and hence, (15) and (16) could not involve possessor raising +clitic doubling).
By contrast, even though Greek has high applicatives with static predicates (17a) it does not allow them with unergatives (17b) because there is no lower argument for GEN to be assigned:
a. Kratisa tis Marias tin tsanda gia na vgali to palto tis
Held-1sg the Mar-GEN the bag-ACC to na take.out-3sg the coat her-GEN
‘I held Mary’s bag (for her) to take out her coat’
b. *Etreksa/*Kolimpisa/*Perpatisa tu Petru
Run-1SG/Swam-1SG/Walked-1SG the Peter-GEN
*’I run/swam/walked for Peter’

20. SMG genitives
Argument #3: Direct object incorporation and genitive/accusative alternation
Theme incorporation in (low) goal applicatives yields realization of the IO as ACC. This is evidence that GEN is not tied to the theta-role goal (as the inherent case theory would predict) but to the presence of a lower DP in the vAPPLP domain (as the dependent case theory would predict):
(18) Dino trofi tu ftochu
Give-1sg food-acc the poor-gen
‘I give food to the poor’
(19) Trofodoto ton ftocho
Food.give-1sg the poor-acc
‘I feed/cater for the poor’
Here someone could notice and ask that the same alternation applies to se-goals:Dino trofi s-ton ftoxoTrofodoto ton ftoxo.If you receive this question, you will say that the existence of this alternation would lead us to analyse se-PPs as DPs bearing dependent case. Say that in my paper in (2005) I provide evidence that se-PPs in Greek are in fact introducing double object datives (this evidence comes from benefactive constructions and so called Oehrle's contexts, maybe you should read this paper) and that the existence of these alternations suggests that se-is becoming a case marker (in my paper I said that it is a P, this would have to be reconsidered). Do not put it in the slides, only if someone asks you. We can talk about it briefly on the phone.

21. SMG genitives summary
SMG genitives are sensitive to the existence of a lower argument inside the vAPPLP domain in a variety of environments like ditransitives (ex. 5-7), dyadic unaccusatives (ex ) and unaccusative sensation verbs (ex ).
When there is no local lower argument, the same DP appears in nominative (ex. 14), accusative (ex. 19) or is banned (ex. 17b with high applicatives).
These facts strongly suggest that GEN in SMG is NOT inherent case sensitive to particular theta roles but sensitive to the presence of a lower DP in the vAPPLP domain.
 SMG genitive is dependent case in the sense of Baker (2015)

22. CG datives/genitives They are subject to semantic generalizations;
They are idiosyncratically determined by particular verbs or prepositions;
They are linked to certain theta-roles (goals tend to be dative, sources and possessors tend to be genitive, and verbs prefixed by dative or genitive assigning prepositions must assign dative or genitive to the goal)
They are licensed as internal argument to both monotransitives and ditransitives  they are NOT sensitive to the presence of another argument inside the vP.
 They satisfy all criteria for lexical/inherent cases.
But then how do we account for their ability to alternate with nominatives in passives (Anagnostopoulou & Sevdali 2015)?

23. CG Passivisation of datives and genitives
IOs are not allowed to alternate with nominative in MG, (20) is ungrammatical in all dialects. Thus, even though NG ditransitives contain two accusative objects, they cannot be assimilated to English double object constructions which freely permit passivization of the IO,
(20) a. *O Petros dothike ena pagoto All dialects of Greek
The Peter.Nom gave.NAct an ice cream.Acc
‘Peter was given an ice cream’
b. He was given an ice cream English
This difference between Greek and English follows if SMG GEN and NG ACC are sensitive to the presence of a lower DP, while English IOs (at least pronouns that surface with overt ACC and NOM) bear dependent case sensitive to a higher DP (see Baker 2015, ch. 6 on the mechanics of this in English-type systems).
But in CG, DAT and GEN do become NOM in passives, even though they were seen to qualify as lexical/inherent cases w.r.t. the criterion of theta-related/ idiosyncratically assigned case.

24. CG passivization of datives/genitives Anagnostopoulou & Sevdali 2015, Alexiadou, Anagnostopoulou & Sevdali 2014
(21) dat-nom alternations in transitives:
a. Athe:nai-oi epibouleu-ousin he:m-in.
Athenians-nom betray-3sg.prs.act us-dat
‘The Athenians are betraying us.’
b. He:m-eis hup’ Athe:nai-o:n epibouleu-ometha.
we-nom by Athenians-gen betray-1pl.prs.pass
‘We are betrayed by the Athenians.’(Thucydides, Historia I: 82. 1)
(22) gen-nom alternations in transitives:
a. Katapse:phe:z-o: tin-os.
condemn-1sg.prs.act someone-gen
‘I condemn someone.’
b. Ekeino-s katepse:phis-the:.
he-nom condemn-3sg.aor.pass
‘He was condemned.’ (Xenophon, Historia V: 2. 36)

25. CG passivization of datives/genitives Anagnostopoulou & Sevdali 2015, Alexiadou, Anagnostopoulou & Sevdali 2014
(23) dat-nom alternations in ditransitives:
a. All-o ti meiz-on hum-in epitaks-ousin (Active: acc-dat)
something.else-acc bigger-acc you-dat order-3pl.prs.act
‘They will order you to do something else bigger/greater.’
b. All-o ti meiz-on hum-eis epitachthe:s-esthe (Passivized: acc-nom)
something.else-acc bigger-acc you-nom order-2pl.prs.pass
‘You will be ordered to do something else, bigger.’ (Thucydides, Historia I: )
(24) gen-nom alternations in ditransitives:
a. Apetem-on to:n strate:g-o:n tas kephal-as (Active: gen-acc)
cut.off-3pl.aor.act the generals-gen the heads-acc
‘They cut the heads from the generals.’
b. Hoi strate:g-oi apetme:th-e:san tas kephal-as. (Passivized: nom-acc)
the generals-nom cut.off-3pl.aor.pass the heads-acc
‘The generals were beheaded/The generals had their heads cut off.’ (Xenophon, Anabasis II: 6. 29)

26. CG passivization of datives/genitives
The core observation we will build on is that all verbs in the examples (21 – 24) above, and in general the majority of verbs that allow passivization of dative/genitive IOs in CG are prefixed (cf. also Michelioudakis 2012):
(25) a. epi-bouleuometha ‘be betrayed’
b. kat-epse:phis-the: ‘be condemed’
c. epi-tachthe:s-esthe: ‘be ordered’
d. ap-etme:th-e:san: ‘be cut off’
These prefixes are homophonous to the corresponding prepositions which retain their case- assigning properties when they are prefixed to the verbs.

27. CG passivization of datives/genitives
Following AAS (2014) we propose that genitives and datives in CG are always contained within PPs, overt, as in the prefixed examples discussed here, or covert with non-prefixal verbs;
Prefixal verbs in CG are formed by P(reposition)-incorporation of the prepositions introducing the dative and genitive objects;
Generally speaking, PPs are phases, and, DPs contained within them are inaccessible to operations triggered by higher heads; i.e. opaque to operations like Agree, Move;
However, there are strategies by which such PPs become transparent (Rezac (2008) for different dialects of Basque);
One major strategy leading to the transparency of dative and genitive objects is P incorporation into a higher head, the complex V-Voice.
We propose that P-incorporation in CG makes the relevant PPs transparent, and the dative and genitive objects are allowed to passivize, as shown in (21-24) above.
CG datives and genitives do not appear as such when they become subjects of passives posssibly because CG allows case-stacking (Richards, 2012)
The phase-lifting effect of P-incorporation follows from the hypothesis that head-movement of certain phase heads extends the phase to the higher projection, as proposed by den Dikken (2007), Gallego (2006, 2010), Gallego & Uriagereka (2006), see also Bobaljik & Wurmbrand (2013), in the spirit of Baker’s (1988) Government Transparency Corollary.

28. On the diachrony of Greek cases and prepositions
Proposal so far:
CG datives/genitives are lexical/inherent cases
they are hidden PPs
SMG genitives bear dependent GEN
they are DPs
SMG genitives bear dependent GEN
they are DPs (in most cases; not in archaic monotransitives, non-archaic monotransitives might be analysed as hidden ditransitives with a zero theme pro that counts for dependent case assignment).
Question: how did this change take place?
How does it relate to the loss of morphological dative from Greek?—
We will present evidence that changes in the PP-system played a key role.

29. Loss of morphological dative
Replacement of (morphological) dative in Standard Modern Greek
a) Accusatives (as objects of transitive verbs, as IOs of ditransitive verbs in Northern Greek).
b) Genitives as IOs in SMG.
c) PPs (for locative, instrumental and other adverbial uses of CG datives).
Argumental dative loss timeline:
3rd – 9th/10th centuries Humbert (1930) or 4th – 8th century A.D. Luraghi (2003)
Humbert (1930): the three main uses of the dative: locative, instrumental and argumental have three very separate diachronic paths.

30. Loss of morphological dative
From the 3rd – 8th centuries AD the dative was variably replaced by the genitive and the accusative case . Browning (1983: 42 – 43): datives and genitives are found coordinated as complements of the same verb. Stolk (2015) also reports what she calls “the mixed construction”:
(26) aneste:sa emauto:i kai Eias te:s sumbiou
set.up.aor.1sg myself.dat and Eia.gen the.gen wife.gen
‘I set this up for myself and for my wife, Eia’
Last step of dative loss happened around the 9th – 10th century AD Humbert (1930);
Lendari & Manolessou (2003) argue that there was a long period of variability, resolved in around the 15th century with the dialectal split that we observe today; This “instability” was in fact a classic case of “grammars in competition” (Kroch, 1989, Pintzuk and Kroch, 1999, Santorini, 1992 and Taylor, among many others) observed in literary vs. non-literary texts;
The situation with ditransitives that display a dialectal split today, is a result of two separate cases of analogy:
Accusative replaces the dative, due to a generalization of the ACC-ACC frame (Horrocks, 2007);
Genitive replaces the dative, due to the influence of possessor raising constructions (Gianollo, , Cooper & Georgala, 2012 and Stolk, 2015) .
In P.Oxy (4TH AD) we see dative replaced by acc and gen in the same text.
Later in New Testament we see the same P selecting for acc and dat in the same text – John

31. Diachrony of prepositions
Prepositions in CG function as head of PPs, but also as prefixes. A key difference in the properties of CG and SMG prepositions is that the former could assign all three oblique cases of CG (accusative, dative and genitive), whereas prepositions in SMG have lost their case assigning capacities, assigning only accusative (exceptions: archaic prepositions reflecting the CG system due to the “katharevousa” register metaksi ‘between’ , enantion ‘against’ , kata ‘against’);
There are two interrelated changes in the prepositional system of Greek that concern us here:
Prepositions are first used to “reinforce” (Bortone, 2010) / double the semantic role of Greek cases, leading ultimately to the replacement of oblique cases by prepositions;
The prepositions themselves lose their case-assigning abilities and are only able to assign accusative case (Hatzidakis, 1892, Bortone, 2010).

32. Diachrony of prepositions
Prepositons are starting to lose their capacity of assigning dative and genitive from Koiné (Lavidas, 2010) and this change concludes during the early Medieval Greek period (Horrocks, ) or the 10th century (Browning 1983: 42 – 43), i.e. Ps had lost their idiosyncratic case- assigning capacity.
This change is especially interesting if viewed from the perspective of Pesetsky’s (2013) proposal about the nature of oblique case in connection to prepositions:
According to Pesetsky assigning an oblique case is indeed a defining characteristic of prepositions.
In his system, oblique case is a P- feature assigned by a P to its complement.

33. Diachrony of Greek prepositions, DOs & IOs
Classical Greek: Prepositions assign oblique  learners acquire them as prepositions;
Dative/genitive IOs and DOs involve a covert preposition  learners can acquire them as PPs exactly because they bear overt oblique case morphology.
 (Prediction: every oblique phrase in CG is a PP with a null preposition).
 Hellenistic Greek /Koiné: Prepositions start losing their oblique case assigning capacity  learners cannot straightforwardly acquire them as such (Grammars in competition)
Medieval Greek: Prepositions have lost their oblique case assigning capacity
 in argumental position Ps cannot be analysed as prepositions anymore;
 learners reanalyse DOs and IOs as DPs;
 These DPs cannot receive case by anything phrase internal anymore;
 Greek IOs and DOs get reanalysed as DPs receiving dependent case, in opposition to a lower DP argument within the VP.

34. Summary
CG dative/genitives that function as DOs and IOs are inherent/lexical cases, with a prepositional structure, they are PPs, while on the other hand, SMG genitives are DPs that receive dependent case configurationally within the VP.
What we normally refer to as inherent case-marked nominal is a hidden PP.
The change in Greek is a category change in the structural makeup of Greek DOs and IOs, from PPs to DPs.
Trigger for change: changes in the case-assigning properties of prepositions;
the loss of morphological dative case and its syncretism with genitive and accusative was not the main cause and only indirectly affecting the observed syntactic change;
Major consequence for Greek: extensive loss of null prepositions and generalization of dependent case assignment to internal arguments.
On the three genitive-assigning SMG prepositions <metaxy, enantion, kata> : they are a direct effect of katharevusa making the system conservative and introducing grammars in competition.
Milao tu Petru, Tilefonao tu Petru, can be analysed as concealed ditransitives with a null cognate theme DO: Milao tu Petru kati, TIlefonao tu Petru ena tilefonima.

35. Ευχαριστούμε!

36. Appendices Accusative case assignment?
Other arguments in favour of SMG dependent genitive
The three genitive assigning prepositions and the few monotransitives that assign genitive
On the three genitive-assigning SMG prepositions <metaxy, enantion, kata> : they are a direct effect of katharevusa making the system conservative and introducing grammars in competition.
Milao tu Petru, Tilefonao tu Petru, can be analysed as concealed ditransitives with a null cognate theme DO: Milao tu Petru kati, TIlefonao tu Petru ena tilefonima.