Quantifiers

From Icelandic Parsed Historical Corpus (IcePaHC)
Jump to: navigation, search

Head types

Q - quantifier, positive: fáir 'few'

QR - correlative: færri 'fewer'

QS - superlative: fæstir 'fewest'

Q+NUM - HVORTVEGGJA

ONE+Q - EINHVER

Q+ADV - HVERGI

Q-heads, other than Q+ADV, are case marked: Q-N, Q-A, Q-D, Q-G; QR-N, QR-A, QR-D, QR-G; QS-N, QS-A, QS-D, QS-G, Q+NUM-N, Q+NUM-A, Q+NUM-D, Q+NUM-G, etc.

Treatment of individual quantifiers

ANNAR, tagged as OTHER; AÐRIR tagged as OTHERS (See also Treatment of individual words)

EINHVER and eitthvað, etc, are tagged ONE+Q (when the meaning is "some"), not ONE+WPRO

HVATKI 'nothing' (sometimes 'whatever') is tagged Q.

HVORTVEGGJA and hvorirtveggja, hvorartveggja, etc., are tagged Q+NUM

HVERGI Q+WADV

Quantifiers without a "special treatment", i.e. tagged as Q:

ALLUR 'everyone', BÁÐIR 'both', EKKERT 'nothing', ENGINN 'no one', HVER, HVOR, SÉRHVER, NEINN 'no one, (not) any' (HVER 'who' is WPRO), NOKKUR (and NEKKVAR/NEKKVER) 'any, anybody' (usually Q; see Treatment of individual words), MARGUR 'many', MIKILL and LÍTILL (usually Q; see Treatment of individual words

EKKI is usually tagged NEG, but with the meaning EKKERT it is tagged Q.

LÍTT 'not so much', tends to project an NP-MSR phrase.

MANNGI, MANGI 'no one'

Floated quantifiers

Floated quantifiers are tagged Q without projecting a phrase. They are parsed without a trace. See PPCME/PPCEME guidelines.


( (IP-MAT-SPE (CONJ og-og)
	      (NP-SBJ (NP-POS (PRO-G hans-hann))
		      (NS-N innyfli-innyfli))
	      (BEPI eru-vera)
	      (ALSO og-og)
	      (Q-N öll-allur)
	      (RP út-út)
	      (VAN hlaupin-hlaupa)
	      (. .-.)))

Floated quantifier modifying a trace (*T*)


		  (CP-REL (WNP-2 (WPRO-N hverjir-hver))
			  (C eð-eð)
			  (IP-SUB (NP-SBJ *T*-2)
				  (Q-N allir-allur)
				  (RDDI urðu-verða)
				  (ADJP (ADJ-N heilbrigðir-heilbrigður)))))
	  (. .-.))
  (ID 1540.NTACTS.REL-BIB,239.234))

Extraction out of NP

Topicalization out of NP, stylistic fronting of quantifiers out of NP, and other extraction of a subpart of an NP:

See in the PPCME/PPCEME Guidelines

When quantifiers move out of another NP, it is parsed like this:

						    (IP-SUB (QP-3 (Q-N engi-enginn))
							    (RDDI varð-verða)
							    (NP-SBJ (QP *ICH*-3)
								    (ADJ-N líkamleg-líkamlegur)
								    (N-N munúð-munúð))
							    (PP (P í-í)
								(NP (N-D holdi-hold)
								    (NP-POS (PRO-G hennar-hún))))