Difference between revisions of "MD"

From Icelandic Parsed Historical Corpus (IcePaHC)
Jump to: navigation, search
 
(18 intermediate revisions by 3 users not shown)
Line 1: Line 1:
 
Only verbs with modal meaning that take bare infinitives are tagged as MD:
 
Only verbs with modal meaning that take bare infinitives are tagged as MD:
  
gera/gjöra 'do' (found in Old(er) Icelandic), mega 'may', munu 'will', skulu 'shall', vilja 'want, will'
+
<pre>
 +
mega 'may', munu 'will', skulu 'shall', vilja 'want, will'
 +
</pre>
  
This includes [[GETA]] when it functions as a modal and takes a past participle complement (see [[Treatment of individual words]]).
+
This includes [[GETA]] when it functions as a modal and takes a past participle complement, and [[FÁ]] when it has the same meaning as GETA and functions the same (see [[Treatment of individual words]]).  
 +
 
 +
[[VILJA]] does not always function as a modal; e.g. in cases where it takes a direct object, it is not tagged MD: ''Ég vil þetta''.  There is considerable variation in earlier texts as to whether VILJA takes að-marked infinitives or not, and individual texts/authors are not expected to be consistent in this (occasionally this happens with SKULU as well, and very very occasionally with MUNU: see Ævisaga Jóns Steingrímssonar for MUNU as VB).  For this reason, we tag VILJA as MD when it takes a bare infinitive, and as VB elsewhere.  Analysts are cautioned to look for both taggings of VILJA in studying the history of Icelandic modals!
  
 
Verbs with modal meaning that take að-marker (to) and an infinitive are not tagged as MD but as main verbs taking IP-INF:
 
Verbs with modal meaning that take að-marker (to) and an infinitive are not tagged as MD but as main verbs taking IP-INF:
  
 +
<pre>
 
eiga 'ought to', fá 'be allowed to', hljóta 'must', kunna 'can', þurfa 'need to', verða 'must', ætla 'intend'
 
eiga 'ought to', fá 'be allowed to', hljóta 'must', kunna 'can', þurfa 'need to', verða 'must', ætla 'intend'
 +
</pre>
 +
 +
KUNNA can (rarely) take a bare infinitive, in which case it may be ambiguous between being a modal and being equivalent to "be able to VB" with an IP-INF complement (see Ævisaga Jóns Steingrímssonar and Jón Indíafari for examples).  In this case, we continue to tag it VB and treat it as taking an IP-INF complement.
 +
 +
If a verb that usually takes the að-marker for some reason does not, and does not have modal meaning, it is still parsed as taking IP-INF.
 +
 +
<pre>
 +
( (IP-MAT (NP-SBJ (PRO-A Hann-hann))
 +
          (VBPI langar-langa)
 +
          (IP-INF (VB fara-fara))))
 +
</pre>
 +
 +
Infinitive of some pretero-present has a special past marking (Hann sagðist mundu/skyldu gera þetta). This is not tagged especially: we tag it as MD and '''not''' MDD.
 +
 +
===Verbs elided after modals===
 +
See [http://www.ling.upenn.edu/~beatrice/annotation/syn-empty.htm#generic_empty_category PPCEME guidelines].
 +
 +
===VB before VB*===
 +
When a verb is fronted and becomes before a finite verb other than a modal, parse it as IP-INF.
  
If a verb that usually takes bare infinitive takes the að-marker, it is still marked as MD and vice versa, i.e. if a verb that usually takes the að-marker for some reason does not, it is still parsed as taking IP-INF.
+
<pre>
 +
  (ADVP-TMP (ADV þá-þá)
 +
    (CP-REL (WADVP 0)
 +
    (C er-er)
 +
    (IP-SUB-SPE (IP-INF-SPE (VB nálgast-nálgast))
 +
(VBDI tók-taka)
 +
(NP-SBJ (NP-POS (N-G fyrirheits$-fyrirheit) (D-G $ins-hinn))
 +
(N-N tími-tími)))))
 +
</pre>

Latest revision as of 13:07, 23 September 2010

Only verbs with modal meaning that take bare infinitives are tagged as MD:

mega 'may', munu 'will', skulu 'shall', vilja 'want, will'

This includes GETA when it functions as a modal and takes a past participle complement, and when it has the same meaning as GETA and functions the same (see Treatment of individual words).

VILJA does not always function as a modal; e.g. in cases where it takes a direct object, it is not tagged MD: Ég vil þetta. There is considerable variation in earlier texts as to whether VILJA takes að-marked infinitives or not, and individual texts/authors are not expected to be consistent in this (occasionally this happens with SKULU as well, and very very occasionally with MUNU: see Ævisaga Jóns Steingrímssonar for MUNU as VB). For this reason, we tag VILJA as MD when it takes a bare infinitive, and as VB elsewhere. Analysts are cautioned to look for both taggings of VILJA in studying the history of Icelandic modals!

Verbs with modal meaning that take að-marker (to) and an infinitive are not tagged as MD but as main verbs taking IP-INF:

eiga 'ought to', fá 'be allowed to', hljóta 'must', kunna 'can', þurfa 'need to', verða 'must', ætla 'intend'

KUNNA can (rarely) take a bare infinitive, in which case it may be ambiguous between being a modal and being equivalent to "be able to VB" with an IP-INF complement (see Ævisaga Jóns Steingrímssonar and Jón Indíafari for examples). In this case, we continue to tag it VB and treat it as taking an IP-INF complement.

If a verb that usually takes the að-marker for some reason does not, and does not have modal meaning, it is still parsed as taking IP-INF.

( (IP-MAT (NP-SBJ (PRO-A Hann-hann))
          (VBPI langar-langa)
          (IP-INF (VB fara-fara))))

Infinitive of some pretero-present has a special past marking (Hann sagðist mundu/skyldu gera þetta). This is not tagged especially: we tag it as MD and not MDD.

Verbs elided after modals

See PPCEME guidelines.

VB before VB*

When a verb is fronted and becomes before a finite verb other than a modal, parse it as IP-INF.

	  (ADVP-TMP (ADV þá-þá)
		    (CP-REL (WADVP 0)
			    (C er-er)
			    (IP-SUB-SPE (IP-INF-SPE (VB nálgast-nálgast))
					(VBDI tók-taka)
					(NP-SBJ (NP-POS (N-G fyrirheits$-fyrirheit) (D-G $ins-hinn))
						(N-N tími-tími)))))