Language typology Basic word order The two types

Language typology Basic word order

The two types of syntactic typology on word order types contentive typology alignment patterns

What is basic word order? • basic word order at the clausal level consists of three major parts: Subject (S), Object (O) and Verb (V) • basic word order can be found in indicative transitive clauses • basic word order = the ordering of the 3 major constituents (S, O and V)

A definition of basic word order [It is generally thought that] basic word order at the clausal level is found in stylistically neutral, independent, indicative clauses with full noun phrase (NP) participants, where the subject is definite, agentive and human, the object is a definite semantic patient, and the verb represents an action, not a state or an event. ” (Siewierska 1988: 8. in Song 2001: 49. )

Basic word order types • How can these 3 elements (S, O, V) appear in a clause? • How many ordering types are possible? • There are 6 logically possible ways of ordering:

(1) Turkish (Comrie 1981: 81 -82. ) (SOV) Hasan öküz-ü aldi. Hasan ox-ACC bought ’Hasan bought an ox. ’ (2) English (SVO) The farmer killed the duckling. (3) Welsh (VSO) Lladdodd y ddraig y dyn. killed the dragon the man’ ’The dragon killed the man. ’

(4) Malagasy (VOS) Nahita ny mpianatra ny vehivavy. saw the student the woman ’The woman saw the student. ’ (5) Hyxkariana (OVS) Toto yahosiye kamara. man grabbed jaguar ‘The jaguar grabbed the man. ’

(6) Nadëb (Song 2001: 2. ) samÚÚy yi (OSV) qa-wù howler-monkey people eat ‘People eat howler-monkeys. ’ • As it can be seen there are cross-linguistic examples of all 6 types • SOV, SVO, VSO, VOS, OVS and OSV

Some criteria about defining basic word order • pragmatic neutrality – perhaps the most important • textual frequency – the more frequent is the basic, but: the most frequent word order may vary from text to text depending on different text types – it isn’t a very reliable parameter • formal markedness – the unmarked is the basic – exceptions from different languages – this is the most unreliable parameter • these parameters are irrelevant to flexible word order languages

Non-clausal patterns • basic word order is also found in non-clausal patterns (i. e. in phrases) • e. g. ordering of: • adposition (Adp) and noun (N) • genitive (G) and noun (N) • relative clause (Rel) and noun (N) • What else could be a pattern like these? • adjective (A) and noun (N) • article (Art) and noun (N) • auxiliary verb (Aux) and main verb (V)

Inception of word order typology • Joseph H. Greenberg – 1963, the first work on basic word order from a typological viewpoint • he established a new type of universal statement, the implicational universal • e. g. x y (read: if the x exists, than this implies the existence of y) • Greenberg’s (45) implications are unilateral!- x y≠y x

Some greenbergian universal statements UNIVERSAL 1 „ In declarative sentences with nominal subject and object the dominant order is almost always one in which the subject precedes the object. ” UNIVERSAL 2 „In languages with prepositions, the genitive almost always follows the governing noun, while in languages with postpositions it almost always precedes. ” UNIVERSAL 3 „Languages with dominant VSO order are always prepositional. ” UNIVERSAL 27 „If a language is exclusively suffixing, it is postpositional; if it is exclusively prefixing, it is prepositional. ” (Greenberg 1963. in Song 2001: 53 -55. )

Examples for UNIVERSAL 1 A fiú kenyeret eszik. Poika syö leipää. ‘The boy is eating bread. ’ (SO) UNIVERSAL 2 a fa alatt/puun alla ‘under the tree’; az anya lánya/äidin tytär ‘the girl of the mother’ (NPp&GN) UNIVERSAL 25 Hungarian and Finnish are both suffixing languages and the most frequently used adposition types are postpositions. Do you think that these universal statements exist in your native language? Examples?

Reflections on universal 1 • Universal 1 says that in languages with nominal subject and object the ordering is SO. • Referring to universal 1 one has to say that OS order is almost impossible. • This means that the existance of VOS, OVS and OSV languages is almost impossible. • BUT: there are data opposed to this universal • AND: in universal 1 it is said that the S almost always precedes the O – this is not an exceptionless universal • It has been pointed out that the number of VOS, OVS and OSV languages is quite reduced (there about 4 -8 OVS languages. cf. WALS)

Tripartite system • Analyse of the three languages in which the S precedes the O – SOV, SVO, VSO • Verb based typology: serializing ordering types based on the position of verb • VSO – verb in the first position relative to the position of S and O • SVO – verb is in the second position • SOV – verb is in the final position • This verb based typology had a main role in further research

The verb based typology • W. P. Lehmann (1973) – Fundamental Principle of Placement (FPP) • the primary syntactic construction: is the verb and the object • they are ‘primary concomitants’ of each other in the sentence • subject is left out consideration in Lehmann’s work, because subject is not so important part of the sentence • as an argue for this theory: subjectless sentences like Lat. pluit ‘it rains’ • thus Greenberg’s tripartite system is reduced to OV and VO word order types

The FPP • The FPP says that: „…the modifiers are placed on the opposite side of a basic constituent, V or O, from its primary concomitant. ” (cf. Song 2001: 56. ) • depending on whether the analyzed language is OV or VO type, we can predict: in OV languages the verbal elements (e. g. negation, causative, etc. ) appear on the right side of the verb, while nominal elements (e. g. genitive, adjective, numeral) appear on the left side of the noun • What can one predict about the position of verbal and nominal elements in VO languages? • in VO: verbal elements are placed on the left side of the verb, whereas nominal elements are placed on the right side of the noun

The FPP • Lehmann has analyzed Greenberg’s data (a 30 language sample) using the predictions of the FPP • he has pointed out that on the basis of the FPP one could make predictions about word order properties at the morphologic level as well • FPP does not use the distinction of heads (main parts) and modifiers (dependent parts), FPP makes predictions about word order, depending on the status of the modifiers (verbal or nominal) • there a lot of languages which do not behave as the FPP predicts, but Lehmann says that the so called ambivalent languages are in a typological change from OV to VO or from VO to OV (further research has pointed out that only the former type of historical change is known)

An excercise you have the following parameters: OV, VO, NA, AN, V-vm, vm-V (where vm is ‘a verbal modifier’) What is the ordering of N and A, V and vm in OV and in VO languages? AN & V-vm in OV NA & vm-V in VO

Conclusion • One has to consider the FPP as a generalization about the greenbergian universal statements

Venemann’s theory • Th. Vennemann states that the subject has no importance in the notion of basic bord order, thus he uses the categories: OV and VO • his aim is to make a general explanation about Greenberg’s universal statements • Vennemann’s theory is the Principle of Natural Serialization (PNS) which can be seen as a generalization on greenbergian universals • the basis of the explanation is categorial analogy

Some features of the PNS • The PNS states that: the order of operators (i. e. modifiers or dependent parts) and operands (i. e. modified or head parts) tends to be serialized in one direction • in practice: operators before operands OR operands before operators operator {operand} operator [operand] in OV operand [operator] in VO

PNS • the following cathegories are used in the PNS (Vennemann 1974) OPERATOR OPERAND object adverbial main verb adjective relative clause genitive numeral determiner adjective standard of comparison noun phrase verb auxiliary noun noun comparison marker comparative adjective adposition

Some critical notes about the PNS • the status of some categories are questionable • for instance: the auxiliary (Aux) is regarded as a verbal modifier in traditional grammar, but in the PNS Aux is the head of the content (main) verb • this system is reduced correlated to Greenberg’s work • Vennemann’s implications are bilateral: p q = q p • Greenberg (UNIVERSAL 3): Languages with dominant VSO order are always prepositional. • How can we read this universal according to the PNS? • VO&Pr = OV&Pp • In Greenberg’s work one could make distinctions about universals: there are weak and strong ones BUT: the PNS cannot make distinctions like these • BUT: Vennemann’s universals are statistical → if more than the half part of the languages behave as the PNS predict then that feature is applicable for generalization

Hawkins’ theory on word order • his aim was to making language universals exceptionless • exceptionless universals help for characterizing possible human language • Hawkins states that statictical (restriceted) universals can be converted into exceptionless (unrestricted) universals

An example of converting • a. b. c. (a) (b) (c) • Statistical universals: Pr (NA NG) Pr (NDem NA) Pr (NNum NA) if a language has preposition word order, then if the adjective follows the noun, the genitive follows the noun if a language has preposition word order, then if the demonstrative determiner follows the noun, the adjective follows the noun if a language has preposition word order, then if the numeral followsthe noun, the adjective follows the noun Exceptionless universals: a. b. Pr & -SVO (NDem NG) Pr & -SVO (NNunm NG) • Hawkins creates exceptionless universals by increasing the conditioning property from one to two

The HSP • Hawkins found that only 7 of the 32 mathematically possible co-occurances of the five nominal modifiers ar attested in Greenberg’s data and in his own as well a. Pr & NDem & NNum & NA & NG & NRel b. Pr & Dem. N & NNum & NA & NG & NRel c. Pr & NDem & Num. N & NA & NG & NRel d. Pr & Dem. N & Num. N & NA & NG & NRel e. Pr & Dem. N & Num. N & AN & NG & NRel f. Pr & Dem. N & Num. N & AN & GN & NRel g. Pr & Dem. N & Num. N & AN & GN & Rel. N

The HSP • e. g. if the Rel is placed before the noun the G, A, Num, Det are placed as well • conclusion: the nominal modifiers are placed before the N „in a fixed and predictable pattern: first the demonstrative determiner or the numeral, then both, then the adjective, then the genitive, finally the relative clause” (Hawkins 1983: 75) • H. predicts that modifiers behave so preciesly because some modifiers are heavier or lighter than others, and because heavier modifiers tend to occur to the right of lighter ones • E. g. the G is heavier than the Num

The HSP • Heaviness Serialization Principle (HSP): Rel ≥R G ≥R A ≥R {Dem, Num} (the ≥R means ‘exhibits more or equal rightward positioning relative to head noun across languages’ BUT: the HSP alone cannot explain the distribution of the nominal modifiers in postpositional languages Thus H. invokes another theory

The MP • In order to explain the exceptions (of OV languages) referring to the HSP, Hawkins invokes the Mobility Principle (MP), which claims that the Dem, the Num, and the A are mobile than the G and the Rel, and thus are able to move around their heads more easily • Mobility Principle (MP) • {A, Dem, Num} ≥M {Rel, G} ( the ≥M means that exhibits greater or equal mobility away from the adposition + NP serialization) • the MP predicts when modifiers move around their heads, they will move in that direction which is expected by the HSP i. e. lighter elements on the left

Language frequency • Greenberg uses in his universals the following words for instance: ‘always, almost always, with overwhelmingly greater than chance frequency’ • these expressions are meant to distinguish different levels of language frequency • this is a crucial point of the analyses, because the number of exceptions that statistical universals can accomodate must be limited • in order to quantify language frequency more objectively H. puts forward the PCCH

The PCCH • Principle of Cross-Category Harmony (PCCH): „The more similar the position of operands relative to their operators across different operand categories considered pairwise (verb in relation to adposition order, noun in relation to adposition order, verb in relation to noun order), the greater the percentage numbers of exemplifying languages. ” (Hawkins 1983: 134)

The PCCH • In other words: if in a given co-occurance all operators are serialized in one direction, then that co-occurance is represented by more languages. • H. demonstrates some cognitive argues in connection with the HSP: until the head is processed, all its dependents have to be held in short-time memory • C. f. the cheese that the mouse that the cat chased ate was rotten

Frequency of the different basic word order types • The aim of Tomlin’s work (1986) was to determine the frequency of the six basic word order types and to provide an explanation of the relative frquencies of these word order types. (sample of 402 languages) • SOV >SVO >VSO >VOS >OVS >OSV • there was no statistical difference between the frequency of SOV and SVO, neither in VOS and OVS – thus: • SOV =SVO >VSO >VOS=OVS >OSV

Three functional principles referring to frequency 1. Theme First Principle (TFP): in clauses information that is relatively more thematic precedes information that is less so 2. Animated First Principle (AFP): in simple basic transitive clauses the NP which is more animated will precede the NPs which are less animated 2 a. Animacy Hierarchy: human >other animate >inanimate 2 b. Semantic Roles Hierarcy: Agent >Instrumental >Benefactive/Dative >Patient 3. Verb-Object Bonding Principle (VOB): the object of a transitive verb is more tightly bonded to the verb than is its subject These principles alone are not applicable for a general explanation about basic word order frequencies

Summary of different theories about word order • Greenberg – implicational universal statements • Lehmann – OV-VO typology, FPP • Vennemann – OV-VO typology, PNS • Hawkins – HSP, MP, PCCH • Tomlin – TFP, AFP, VOB