Armin Buch Linguistic Universals Greenberg’s universals Conclusions Linguistic Universals Armin W. Buch 1 2012/11/28 1 Relying heavily on material by Gerhard Jäger and David Erschler
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Linguistic Universals
Armin W. Buch1
2012/11/28
1Relying heavily on material by Gerhard Jäger and David Erschler
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Linguistic universals
I Properties shared by all languagesI Trivial: all languages have consonants and vowelsI More interesting: Do properties correlate?I Something like “If a language has a phoneme /u/, than it
has neutral gender” (wrong!)I implicational universals
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Example of an implicational universal
I If a language has first/second person reflexives, it has thirdperson reflexives.
I Modern English: 1st, 2nd, 3rd person (‘myself . . . ’)I Modern French, German: only 3rd person (‘sich’)I Old English: no dedicated forms for reflexivesI 1/2 reflexives, but no 3 reflexives: unattested
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg’s universals
I Joseph Greenberg, 1963: “Some Universals of Grammarwith Particular Reference to the Order of MeaningfulElements”
I a relatively small sample of geographically andgenetically diverse languages
I investigated correlations between features of word orderI in a somewhat hard to define “basic word order”I Also keep in mind languages with no fixed word order
(Slavic l., Latin, . . . )
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Problems
I Most universals are only statistical, not absoluteI Correlations might be due to:
I genetic relationshipI areal distributionI chance: all instances of a rare feature might co-occur with
a certain very frequent feature
I So it is not very clear how to count independent events.I But still, universals give a general feeling of what to
expect
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 1
I In declarative sentences with nominal subject and object,the dominant order is [. . . ]2 one in which the subjectprecedes the object.
I WALS mapI Mind the “nominal”: pronominal forms might behave
differently
2I’m leaving out statements of frequency (“almost always” etc.).
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 2
I In languages with prepositions, the genitive [. . . ] followsthe governing noun, while in languages with postpositionsit [. . . ] precedes it.
I WALS mapI diachronic explanation: Genitive constructions are a
possible source for adpositions (source)I English “in front of” etc.I Japanese postpositions:
(1) Teburu-noTable-GEN
ue-nitop-LOC
aruexists
“It’s on the table.”
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 3
I [. . . ] languages with dominant VSO order are alwaysprepositional.
I WALS map
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 4
I Languages with normal SOV order are postpositional.I WALS mapI Well known exception: Latin (SOV, prepositions)
(2) remthing.ACC
publicampublic.ACC
universamwhole.ACC
petisattack.2SG
‘You are attacking the whole republic.’ (Cic. Cat1.)
(3) exfrom
urbecity.ABL
‘out of the city’
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 5
I If a language has dominant SOV order and the genitivefollows the governing noun, then the adjective likewisefollows the noun.
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 8
I When a yes-no question is differentiated from thecorresponding assertion by an intonational pattern, thedistinctive intonational features of each of these patternsare reckoned from the end of the sentence rather than fromthe beginning.
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 9
I [. . . ] when question particles or affixes are specified inposition by reference to the sentence as a whole, if initial,such elements are found in prepositional languages, and, iffinal, in postpositional.
I WALS mapI weak evidence
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 12
I If a language has dominant order VSO in declarativesentences, it always puts interrogative words or phrasesfirst in interrogative word questions; if it has dominantorder SOV in declarative sentences, there is never such aninvariant rule.
I WALS map
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 17
I [. . . ] languages with dominant order VSO have theadjective after the noun.
I WALS mapI “There is no evidence of any relationship between the
order of Verb and Object and the order of Adjective andNoun.” (Dryer 1988:191, Dryer 1986:98; as cited here)
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 18
I When the descriptive adjective precedes the noun, thedemonstrative and the numeral, [. . . ] do likewise.
I WALS map
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 19
I When the general rule is that the descriptive adjectivefollows, there may be a minority of adjectives whichusually precede, but when the general rule is thatdescriptive adjectives precede, there are no exceptions.
I French: petit, grand
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 20
I When any or all of the items (demonstrative, numeral, anddescriptive adjective) precede the noun, they are alwaysfound in that order. If they follow, the order is either thesame or its exact opposite.
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 25
I If the pronominal object follows the verb, so does thenominal object.
I Not vice versa!I Example from Russian
(4) Rybakfisherman
pojmalcaught
rybufish
‘The fisherman caught a fish.’
(5) Rybakfisherman
jejoher
pojmalcaught
‘The fisherman caught it.’
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 27
I If a language is exclusively suffixing, it is postpositional;if it is exclusively prefixing, it is prepositional.
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 28
I If both the derivation and inflection follow the root, orthey both precede the root, the derivation is alwaysbetween the root and the inflection.
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 29
I If a language has inflection, it always has derivation.
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 34
I No language has a trial number unless it has a dual. Nolanguage has a dual unless it has a plural.
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 36
I If a language has the category of gender, it always has thecategory of number.
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 37
I A language never has more gender categories innonsingular numbers than in the singular.
I German: No morphological gender in plural(demonstratives, determiners, adjectives)
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Greenberg 42
I All languages have pronominal categories involving atleast three persons and two numbers.
I 4 persons: inclusive (“We’re going to the cinema. . . ” —“Great!”) vs. exclusive ‘we’ (“. . . without you.”)
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Explanations for universals
I At least some of universals are fairly robust, and thus needto be explained
I Are they wired into our brain (universal grammar)?I A priori, this is at contradiction with the statistical nature
of the findings.I What would a language be like if it violated a universal?
I Harder to learn/speak/process?I Not even a language?
I In word order, uniform headedness seems to be preferred
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Structures that are easier to process will begrammaticalized before the grammar sanctions amore difficult structure of the relevant type. Thisresults in implicational dependencies such as “ifSOV, then postpositions in PP,” a word-orderco-occurrence that will be argued to be optimal forprocessing, in contrast to SOV and prepositions.
Many of the universals are explained by “head ofphrase” generalization. Head-final languages willhave SOV, postpositions, genitive-noun order etc,whereas in head-initial languages the situation isreverse.
(Hawkins 1983, 1994)
Armin Buch
Linguistic Universals
Greenberg’suniversals
Conclusions
Sources
I World Atlas of Language StructuresI Universals ArchiveI Das grammatische Raritätenkabinett