Corpus of Historical Low German

HeliPaD: Syntactic annotation


Preliminaries

The HeliPaD parsing closely follows the principles of the York-Toronto-Helsinki Parsed Corpus of Old English Prose (YCOE), and this manual broadly follows the structure of that one. On occasion, my practice deviates from that of the YCOE and follows that of the Icelandic Parsed Historical Corpus (IcePaHC) or the other Penn historical corpora instead, or adopts a different solution entirely.

Where the HeliPaD follows other Penn historical corpora, the text will be marked like this.

Where the HeliPaD does its own thing, the text will be marked like this.

There's also a page with a summary of these differences.

Examples used in this manual are from HeliPaD version 0.9. You may also be interested in the principles of tokenization, POS-tagging, or the list of tags and categories.


General issues

Phrase structure

The annotation scheme used for the Penn historical corpora and for the HeliPaD uses labelled bracketing, which is easily translatable into tree notation. The trees thus derived, however, are substantially flatter than those usually found in GB or Minimalist syntactic theory: multiple branching is permitted, and many phrase types - especially clauses - end up with quite flat structure. This is not intended as a theoretical commitment, but rather is done for ease of annotation and retrieval.

In particular, there is no VP in the Penn historical corpora, or the HeliPaD. The verb and all its objects and adjuncts are immediately dominated by IP and are sisters there. Here's a simple example:

( (IP-MAT-SPE (CODE <C>)
              (ADVP-TMP (ADV Thuo-tho))
              (VBDI^3^PL slogun-slahan)
              (ADVP-LOC (ADV thar-thar))
              (ADVP-TMP (ADV eft-eft))
              (NP-SBJ (N^N^PL crud-krud))
              (PP (P an-an)
                  (NP (N^A^SG gimang-gimang)))
              (. ,-,))
  (ID OSHeliandC.1291.2409))

All phrase types (except IP and CP, on which more here) have essentially the same structure in the corpus. The key points are the following:

  • Intermediate phrase levels such as bar-levels are not used. Phrasal nodes (e.g. NP, ADVP, NUMP) always immediately dominate the head category (e.g. N, ADV, NUM).
  • Structure is endocentric: heads project a corresponding phrasal node. This is not exceptionless: some heads, such as verbs, as well as determiners (D) and particles (RP), never project a phrase. Single-word modifiers also do not project.
  • Complements, whether single-word or multi-word, always project. This includes possessive genitive phrases, but not possessive pronouns.

In the HeliPaD, single-word modifiers do not project a phrase, even when they follow the head.

Heads, modifiers and complements

Since there are no bar levels, both modifiers and complements are in a sisterhood relation to the head, and the phrase immediately dominates all three. Phrases (except IP) are, in general, the same category as their head (e.g. NP dominates N).

In some cases, phrases are not headed in this way. In some cases, this is because the head categories are viewed as subclasses of one another. For instance, pronouns (PRO) and proper nouns (NPR) are seen as a subcategory of N:

(NP-SBJ (PRO^N^3^PL sie-he))

(OSHeliandC.4.12-15)

In other cases, the absence of a head may be due to elision, which is not represented directly in the corpus as a process. The annotation is neutral as to which of these possibilities is the correct one in any given case: the user is free to judge.

It is in nominals where the greatest discrepancy between head label and phrase label can be found. In addition to nouns (N, NPR), the following categories may serve as "heads":

  • pronouns (PRO, PRO$, MAN)
  • determiners (D)
  • adjectives (ADJ, ADJR, ADJS)
  • quantifiers (Q, QR, QS)
  • numbers/numerals (NUM)
  • inflected participles (VNI, VGI, etc.)

The default treatment of pronouns and determiners is as heading the phrases that immediately dominate them, whereas the other categories listed above are normally treated as modifiers. As such, these other categories may project their own phrase level (e.g. ADJP) when they are themselves modified. When they are not modified, an NP may dominate two potential heads:

(NP-SBJ (PRO^N^3^SG it-it)
        (Q^N^SG all-al))

(OSHeliandC.1303.2428-2430)

(NP-SBJ (D^N^PL thiu-the)
        (NUM^N^PL fiui-fif))

(OSHeliandC.12.47)

If an NP immediately dominates only a "modifier" in this sense, the modifier is treated as the head, and the extra level omitted. For instance, in the following example, no QP is present:

(NP-POS (ADV so-so)
        (Q^G^PL managaro-manag))

(OSHeliandC.974.1862-1865)

ADJPs can be headed only by adjectives, inflected participles, and possessive pronouns, and not by participle phrases or by quantifiers as in the YCOE.

PPs are always headed by prepositions, ADVPs are always headed by adverbs, and QPs are always headed by quantifiers. NUMPs may contain more than one NUM, in which case these together are treated as the head, much as in compounds.

Modifiers, like complements, are daughters of the phrasal node and sisters to the head. Modifiers only project a phrase level when they are multi-word and/or themselves modified. The following examples show a construction with multiple adjectives, then one with a multi-word modifier:

(NP-SBJ-PRN (ADJ^N^SG mari-mari)
            (ADJ^N^SG mahtig-mahtig)
            (NPR^N^SG Crist-Krist))

(OSHeliandC.1384.2574-2577)

(NP (N^D^PL uuordon-word)
    (QP (ADV so-so)
        (Q^D^PL filo-filo)))

(OSHeliandC.1001.1923-1926)

When modifiers are separated from a head with which they agree, these are traced to the head, as with alla below. Unlike in the YCOE, this is the case regardless of whether they are case-marked.

( (IP-MAT-SPE (CODE <C>)
              (INTJP (WADV Huat-hwe))
              (, ,-,)
              (NP-SBJ (PRO^N^1^PL uui-we)
                      (QP *ICH*-1))
              (NP-OB1 (D^A^SG thia-the))
              (ADVP-LOC (ADV hier-her))
              (VBPI^1^PL uuitun-witan)
              (QP-1 (Q^N^PL alla-al))
              (. ,-,))
  (ID OSHeliandC.1426.2654))

Complements, unlike modifiers, always have a phrase label. This includes single-word complements. Possessor NPs are treated as complements in all instances, but possessive pronouns are usually treated as modifiers.

(NP-SBJ (D^N^SG that-the)
        (NP-POS (N^G^PL manno-man))
        (N^N^SG folc-folk))

(OSHeliandC.1430.2658)

(NP-OB1 (PRO$^A^3^SG is-is)
        (N^A^SG gibodscipi-gibodskepi))

(OSHeliandC.1433.2659-2660)

NP arguments (including complements) are distinguished from adjuncts at the IP level, but there is an irreducible element of subjectivity to this task, and the corpus user interested in argument structure should check these distinctions carefully. No attempt has been made to do the same for PPs.

Within other phrases, excluding IPs, PTPs and CPs but including noun phrases, NPs are indicated as possessive (NP-POS) if this is their function (as it usually is within NPs themselves), and unmarked otherwise (as is always the case within PPs). This differs from the approach taken in the YCOE, which labels constituents for case. The default treatment of NPs is to attach them as high as possible in the structure: this means treating constituents as NP-ADT rather than arguments of a non-verbal element.

Sentence fragments (FRAG) and restarts

The exocentric phrasal category FRAG is used for fragmentary clauses or utterances that were written as such in the text, not for those that have been distorted by textual problems or omission.

In HeliPaD 0.9, FRAG is only used for restarts, i.e. for a clause that breaks off and then begins again. The incomplete beginning is included within the FRAG as an IPX- with no equal-sign index, and the FRAG itself is included within the complete clause.

( (IP-MAT (CODE <R_507>)
          (FRAG (IPX-MAT (NP-SBJ (PRO^N^3^SG Siu-siu))
                         (MDDI^3^SG muosta-motan)
                         (PP (P after-aftar)
                             (NP (PRO$^D^3^SG ira-iru)
                                 (N^D^SG magathedi-magathhed)))
                         (, ,-,)))
          (CODE <C>)
          (CP-ADV (ADVP-TMP (ADV sithor-sithor))
                  (C 0)
                  (IP-SUB (NP-SBJ (PRO^N^3^SG siu-siu))
                          (NP-PRD (N^G^SG mannes-man))
                          (RDDI^3^SG uuarth-werthan)
                          (CODE <R_508>)
                          (NP-PRD-PRN (N^G^SG erlas-erl))
                          (PP (P an-an)
                              (NP (N^D^SG ehti-eht)))
                          (CODE <C>)
                          (NP-SBJ-PRN (ADJ^N^SG eđili-ethili)
                                      (N^N^SG thiorna-thiorna))))
          (, ,-,)
          (CODE <R_509>)
          (ADVP (ADV so-so))
          (MDDI^3^SG muosta-motan)
          (NP-SBJ (PRO^N^3^SG siu-siu))
          (PP (P mid-mid)
              (NP (PRO$^D^3^SG iro-iru)
                  (N^D^SG brudigumen-brudigumo)))
          (CODE <C>)
          (NP-OB1 (N^G^PL bodlu-bodal))
          (GE+VB giuualdan-waldan)
          (CODE <R_510>)
          (NP-TMP (NUM^A^PL siƀun-sivun)
                  (N^A^PL uuinter-wintar))
          (ADVP (ADV saman-saman))
          (. .-.))
  (ID OSHeliandC.270.507-510))

In subordinate clauses, the subordinator is usually repeated.

( (IP-MAT-SPE (CODE <C>)
              (CONJ endi-endi)
              (NP-SBJ-X *exp*)
              (NP-OB2 (PRO^D^1^PL us-we))
              (BEPI^3^SG is-wesan)
              (NP-ADT (N^D^PL firinon-firina))
              (NP-PRD (N^N^SG tharf-tharf))
              (CODE <R_2429>)
              (CP-THT-SPE-X (FRAG (CP-THT-SPE (C that-that)
                                              (IPX-SUB-SPE (NP-SBJ (PRO^N^1^PL uui-we))
                                                            ...
                            (, ,-,)
                            (CODE <R_2430>)
                            (C that-that)
                            (IP-SUB-SPE (NP-SBJ (PRO^N^1^PL uui-we))
                                        (NP-OB1 (PRO^A^3^SG it-it))
                                        (PP (P an-an)
                                            (NP (D^D^SG theson-these)
                                                (N^D^SG lande-land)))
                                        (PP (P at-at)
                                            (NP (PRO^D^2^SG thi-thu)))
                                        (CODE <C>)
                                        (VB linon-linon)
                                        (MDPS^1^PL muotin-motan)))
              (. .-.)
              (' '-'))
  (ID OSHeliandC.1303.2428-2430))

Text problems (X)

The exocentric category X is used for ungrammatical structures that arise due to text problems (typically manuscript problems), usually alongside a comment that indicates what the problem is.

( (X (CODE <C>)
     (CODE <COM:OMISSION>)
     (NP-SBJ-X *exp*)
     (ADJP-PRD (ADJ^A^SG cuth-kuth))
     (CODE <R_5891>)
     (NP-OB2 (D^D^PL them-the)
             (N^D^PL liudon-liudi))
     (PP (P after-aftar)
         (NP (D^D^SG them-the)
             (N^D^SG lande-land)))
     (CODE <C>)
     (CP-THT-X (C that-that)
               (IP-SUB (NP-SBJ (PRO^N^3^PL sia-he))
                       (NP-OB1 (ADJ^A^PL sulica-sulik)
                               (N^A^PL lugina-lugina))
                       (MDDI^3^PL uuoldun-willian)
                       (CODE <R_5892>)
                       (VB ahebbian-ahebbian)
                       (PP (P be-bi)
                           (NP (D^A^SG than-the)
                               (ADJ^A^SG helagan-helag)
                               (N^A^SG drohtin-drohtin)))))
     (. .-.))
  (ID OSHeliandC.3489.5890-5892))

Direct speech (-SPE and QTP)

Direct speech is marked in the corpus by adding the label extension -SPE to any IPs and CPs that are spoken. Any IP or CP dominated by an IP or CP labelled -SPE is also labelled -SPE.

( (IP-MAT-SPE (CODE <R_118>)
              (NP-SBJ (PRO$^N^2^SG thin-thin)
                      (N^N^SG theonost-thionost))
              (BEPI^3^SG is-wesan)
              (NP-OB2 (PRO^D^3^SG im-he))
              (CODE <MS_8a>)
              (PP (P an-an)
                  (NP (N^D^SG thanke-thank)))
              (, ,-,)
              (CODE <C>)
              (CP-THT-SPE (C that-that)
                          (IP-SUB-SPE (NP-SBJ (PRO^N^2^SG thu-thu))
                                      (NP-OB1 (ADJ^A^SG sulica-sulik)
                                              (N^A^SG githaht-githaht))
                                      (HVPI^2^SG haƀes-hebbian)
                                      (CODE <R_119>)
                                      (PP (P an-an)
                                          (NP (NP-POS (PRO^G^3^SG is-he)
                                                      (NUM^G^SG enes-en))
                                              (N^A^SG craft-kraft)))))
              (. .-.))
  (ID OSHeliandC.48.118-119))

Non-clausal sequences of direct speech are labelled as QTP, an exocentric phrasal category.

( (IP-MAT (CODE <C>)
          (CONJ endi-endi)
          (NP-SBJ *con*)
          (NP-OB2 (D^D^SG them-the)
                  (N^D^SG alouualden-alowaldo)
                  (CODE <R_2843>)
                  (ADJ^D^SG selƀem-self))
          (VBDI^3^SG sagda-seggian)
          (, ,-,)
          (CODE <C>)
          (CP-THT (C that-that)
                  (IP-SUB (NP-SBJ (PRO^N^3^PL sia-he))
                          (PP (P an-an)
                              (NP (PRO$^D^3^SG iro-iru)
                                  (N^D^SG gisithie-gisithi)))
                          (NP-OB1 (ADV than-than)
                                  (QR^A^SG mer-mer)
                                  (CODE <R_2844>)
                                  (NP-POS (ADJ^G^SG garoes-garu)))
                          (NEG ni-ne)
                          (HVDS^3^PL habdin-hebbian)
                          (CODE <C>)
                          (' '-')
                          (QTP (PP (P nouan-newan)
                                   (NP (NP (ADJ^A^PL gerstin-girstin)
                                           (N^A^PL bruod-brod)
                                           (CODE <R_2845>)
                                           (NUM^A^PL fiui-fif))
                                       (CONJP *ICH*-1)))
                               (PP (P an-an)
                                   (NP (PRO$^D^1^SG usero-usa)
                                       (N^D^SG ferde-fard)))
                               (CODE <C>)
                               (CONJP-1 (CONJ endi-endi)
                                        (NP (N^A^PL fiscos-fisk)
                                            (NUM^A^PL tuena-twene))))))
          (. .-.))
  (ID OSHeliandC.1544.2842-2845))

For the Heliand, the fact that it was originally an oral epic does not mean that all clauses are annotated as -SPE! Only those which are spoken by characters in the story are marked with -SPE or QTP.

In the YCOE, the first independent clause following a verb of saying is included in the parse as the complement of the verb of saying, whereas later independent clauses are treated as separate tokens. In the HeliPaD, these clauses are always treated as independent tokens.

Left-dislocation (-LFD) and resumption (-RSP)

Constituents in the left periphery of a clause that are later resumed by a coreferential phrase are given the extended label -LFD.

In the HeliPaD, unlike in the YCOE, the label -LFD is used systematically with all clausal categories, including CP-ADV (e.g. in if ... then constructions).

( (IP-MAT-SPE (CODE <C>)
              (CP-ADV-LFD-SPE (ADVP (ADV thoh-thoh))
                              (C 0)
                              (IP-SUB-SPE (NP-SBJ-1 (PRO^N^3^SG hie-he))
                                          (ADJP (ADV so-so)
                                                (ADJ^N^SG othi-othi)
                                                (IP-INF-SPE *ICH*-1))
                                          (NEG ni-ne)
                                          (BEPS^3^SG si-wesan)
                                          (CODE <R_1787>)
                                          (NP-OB2 (N^D^PL firihon-firihos))
                                          (IP-INF-SPE-1 (NP-OB1 *-1)
                                                        (TO te-te)
                                                        (VB faranne-faran))))
              (, ,-,)
              (CODE <C>)
              (ADVP-RSP (ADV thoh-thoh))
              (MDPI^3^SG scal-skulan)
              (NP-SBJ (PRO^N^3^SG hie-he))
              (PP (P te-te)
                  (NP (N^D^SG frumu-fruma)))
              (RD uuerthan-werthan)
              (CODE <R_1788>)
              (CP-FRL-OB2-SPE (WNP-SBJ-2 (ADV so-so)
                                         (WPRO^N hue-hwe))
                              (C so-so)
                              (IP-SUB-SPE (NP-SBJ *T*-2)
                                          (NP-OB1 (PRO^A^3^SG ina-he))
                                          (VBPI^3^SG thurugengit-thurhgangan)))
              (. ,-,))
  (ID OSHeliandC.940.1786-1788))

Resumptive elements (-RSP) are used either to match -LFD elements, as in the example above, or on a resumptive pronoun in an unbounded dependency context where a trace would be expected, as in the example below.

  (NP-PRD (D^N^SG that-the)
	  (N^N^SG barn-barn)
	  (NP-POS (NPR^G^SG godes-god)))
  (CODE <R_912>)
  (' '-')
  (CP-REL-SPE (WNP-SBJ-1 (D^N^SG that-that))
	      (C 0)
	      (IP-SUB-SPE (ADVP-LOC (ADV hier-her))
			  (ADVP-TMP (ADV lango-lango)
                                    (ADV iu-ju))
			  (' '-')
			  (IP-MAT-PRN (VBDI^3^PL quathun-kwethan)
				      (NP-SBJ (PRO^N^3^PL sea-he)))
			  (, ,-,)
			  (CODE <C>)
			  (' '-')
			  (NP-SBJ (N^N^PL liudi-liudi))
			  (VBDI^3^PL sagdun-seggian)
			  (CODE <R_913>)
			  (NP-SBJ-PRN (N^N^PL uueros-wer))
			  (ADVP (ADV uuarlico-warliko))
			  (, ,-,)
			  (CODE <C>)
			  (CP-THT-SPE (C that-that)
				      (IP-SUB-SPE (NP-SBJ-RSP-1 (PRO^N^3^SG hie-he))
						  (MDDS^3^SG scoldi-skulan)
						   ...
						  (VB cuman-kuman)))))

(OSHeliandC.489.911-913)

Appositives and parentheticals (-PRN)

Appositive or parenthetical constituents are given the extended label -PRN, which may occur with almost any phrase type. Appositive nominals are particularly common in the Heliand.

In the YCOE, nominal appositive constituents are either contained within, or indexed to, the constituent to which they are in apposition. In the HeliPaD, they are instead treated as sisters to that constituent (or indexed to a sisterhood position). This enables a much less cluttered clausal representation, and the intended apposition relations are usually very clear semantically.

( (IP-MAT-SPE (CODE <C>)
              (VBPI^3^SG farit-faran)
              (NP-OB1 (PRO^A^3^SG ina-he))
              (NP-SBJ (NP-POS (N^G^SG uuerodes-werod))
                      (ADJ^N^SG liut-lut))
              (, ,-,)
              (CODE <R_1783>)
              (NP-SBJ-PRN (ADJ^N^SG faho-faho)
                          (N^N^SG folcscepi-folkskepi))
              (. :-:))
  (ID OSHeliandC.937.1782-1783))

Clauses that are in apposition are contained within or indexed to the constituent to which they are in apposition.

( (IP-MAT-SPE (CODE <C>)
              (PP (P For-for)
                  (CODE <MS_53b>)
                  (NP (D^I^SG thiu-the)
                      (CP-THT-PRN-SPE *ICH*-1)))
              (NP-SBJ (PRO^N^2^PL gi-gi))
              (VB sorgon-sorgon)
              (MDPI^2^PL sculun-skulan)
              (CODE <R_1881>)
              (CP-THT-PRN-SPE-1 (C that-that)
                                (IP-SUB-SPE (NP-OB2 (PRO^D^2^PL iu-gi))
                                            (NP-SBJ (D^N^PL thia-the)
                                                    (N^N^PL man-man))
                                            (NEG ni-ne)
                                            (MDPS^3^PL mugin-mugan)
                                            (CODE <C>)
                                            (NP-OB1 (N^A^PL muodgithahti-modgithaht))
                                            (, ,-,)
                                            (CODE <R_1882>)
                                            (NP-OB1-PRN (N^A^SG uuilleon-willio))
                                            (VB auuendan-awendan)))
              (. .-.))
  (ID OSHeliandC.980.1880-1882))

Parenthetical clauses are also contained within the clause to which they are parenthetical. In most cases, these involve the word quathie:

( (CP-QUE-SPE (CODE <R_2025>)
              (' '-')
              (WNP-SBJ-1 (WPRO^N Huat-hwe))
              (IP-SUB-SPE (NP-SBJ *T*-1)
                          (BEPI^3^SG ist-wesan)
                          (NP-OB2 (NP-OB2 (PRO^D^1^SG mi-ik))
                                  (CONJP (CONJ endi-endi)
                                         (NP-OB2 (PRO^D^2^SG thi-thu))))
                          (' '-')
                          (IP-MAT-PRN (VBDI^3^SG quat$-kwethan)
                                      (NP-SBJ (PRO^N^3^SG $hie-he)))
                          (, ,-,)
                          (CODE <C>)
                          (' '-')
                          (PP (P umbi-umbi)
                              (NP (NP-POS (D^G^PL thesaro-these)
                                          (N^G^PL manno-man))
                                  (N^A^SG lith-lith)))
                          (, ,-,)
                          (CODE <R_2026>)
                          (PP (P umbi-umbi)
                              (NP (NP-POS (D^G^SG theses-these)
                                          (N^G^SG uuerodes-werod))
                                  (N^A^SG uuin-win))))
              (. ?-?))
  (ID OSHeliandC.1047.2025-2026))

-PRN IPs are also used in the annotation of right node raising.

Empty categories

There are seven types of empty category in the HeliPaD:

  • *con*: subject elided under co-ordination
  • *exp*: null expletive subject (may be co-referential with a clausal element)
  • *arb*: arbitrary subject in infinitivals
  • *pro*: null subject, none of the above (usually referential)
  • *T*: trace of wh-movement, always coindexed
  • *ICH*: trace of other movement, always coindexed
  • *: other empty category (e.g. in tough-movement)

In addition, empty complementizers and wh-operators are given a notional phonological content of 0.

Since all complete finite clauses and small clauses are required to have a subject, an empty subject is added if there is no overt one. Non-finite clauses (IP-INF) are not required to have a subject in the annotation, and in cases of raising-to-subject or subject control they will not have one. For expletive subjects, see Expletive constructions.

In cases of conjunction reduction, the elided subject is *con* in all clause types. This category can only be used when the subject is coreferential and identical in number; otherwise, *pro* is used. A conjunction itself is not necessary for *con* to be used: see below for examples both with and without conjunction.

( (IP-MAT (CODE <C>)
          (CONJ endi-endi)
          (NP-SBJ (PRO^N^3^SG hie-he)
                  (ADJP *ICH*-1))
          (ADVP-LOC (ADV thar-thar))
          (PP (P mid-mid)
              (NP (PRO$^D^3^PL is-is)
                  (N^D^PL fingron-fingar)))
          (ADVP-TMP (ADV thuo-tho))
          (CODE <R_2042>)
          (VBDI^3^SG segnoda-segnon)
          (ADJP-1 (ADJ^N^SG selbo-self))
          (CODE <C>)
          (NP-OB2 (PRO$^D^3^PL sinon-sin)
                  (N^D^PL handon-hand))
          (. ,-,))
  (ID OSHeliandC.1055.2041-2042))

( (IP-MAT (CODE <R_2043>)
          (NP-SBJ *con*)
          (VBDI^3^SG uuarahta-wirkian)
          (NP-OB1 (PRO^A^3^SG it-it))
          (PP (P te-te)
              (NP (N^D^SG uuine-win)))
          (. ,-,))
  (ID OSHeliandC.1056.2043))

( (IP-MAT (CODE <C>)
          (CONJ endi-endi)
          (NP-SBJ *con*)
          (VBDI^3^SG hiet-hetan)
          (IP-INF (IP-INF (NP-SBJ *arb*)
                          (NP-OB1 (PRO^G^3^SG is-it))
                          (PP (P an-an)
                              (NP (NUM^A^SG enn-en)
                                  (N^A^SG uuegi-wegi)))
                          (VB hladan-hladan))
                  (, ,-,)
                  (CODE <R_2044>)
                  (CONJP (IP-INF (NP-SBJ *arb*)
                                 (VB sceppian-skeppian)
                                 (PP (P mid-mid)
                                     (NP (NUM^D^SG enaro-en)
                                         (N^D^SG scalun-skala))))))
          (. ,-,))
  (ID OSHeliandC.1057.2043-2044))

Arbitrary PRO in ECM infinitives is indicated by *arb*, as it is in the PPCME2 and IcePaHC (but not the YCOE); see the above example.

Where *con*, *exp* and *arb* are all impossible, *pro* is used. This is not a commitment to a genuine pro-drop analysis; rather, *pro* is a dustbin category for empty subjects, some of which may represent genuine referential null subjects, others scribal error. Cases of conjunction reduction with number mismatch are also treated as *pro*. All empty subjects, whether *pro* or not, appear as early in the clause as possible after such elements as wh-traces, conjunctions, vocatives, interjections, and left-dislocated items.

( (IP-MAT (CODE <C>)
          (NEG Ne-ne)
          (BEDI^3^SG uuas-wesan)
          (ADVP-TMP (ADV io-eo))
          (NP-POS-1 (NPR^G^PL Iudeo-Judeo))
          (ADVP (ADV bethiu-bethiu))
          (CODE <R_2361>)
          (NP-SBJ (NP-POS *ICH*-1)
                  (NP-POS-PRN (N^G^SG lethes-leth)
                              (N^G^SG liudscipes-liudskepi))
                  (CODE <C>)
                  (N^N^SG giloƀo-gilovo))
          (ADJP-PRD (D^I^SG thiu-the)
                    (ADJR^N^SG bettera-bet))
          (CODE <R_2362>)
          (PP (P an-an)
              (NP (D^A^SG thena-the)
                  (ADJ^A^SG helagon-helag)
                  (NPR^A^SG Crist-Krist)))
          (. ,-,))
  (ID OSHeliandC.1258.2360-2362))

( (IP-MAT (CODE <C>)
          (CONJ ac-ak)
          (NP-SBJ *pro*)
          (HVDI^3^PL habdun-hebbian)
          (NP-OB2 (PRO^D^3^SG im-he))
          (NP-OB1 (ADJ^A^SG hardon-hard)
                  (N^A^SG muod-mod))
          (, ,-,)
          (CODE <R_2363>)
          (NP-OB1-PRN (ADJP (ADV suitho-switho)
                            (ADJ^A^SG starcan-stark))
                      (N^A^SG strid-strid))
          (. ,-,))
  (ID OSHeliandC.1259.2362-2363))

*pro* is also used in examples where a subjunctive is used with imperative force.

( (IP-MAT (CODE <R_3912>)
          (CP-FRL-LFD (WNP-SBJ-1 (ADV so-so)
                                 (WPRO^N hue-hwe))
                      (C so-so)
                      (IP-SUB (NP-SBJ *T*-1)
                              (ADVP-LOC (ADV thar-thar))
                              (PP (P mid-mid)
                                  (NP (N^I^SG thurstu-thurst)))
                              (CODE <C>)
                              (VN bithuungan-bithwingan)
                              (BEDS^3^SG uuari-wesan)))
          (, ,-,)
          (CODE <R_3913>)
          (' '-')
          (QTP (NP-SBJ-RSP *pro*)
               (ADVP (ADV so-so))
               (VBPS^3^SG gange-gangan)
               (NP-OB2-RFL (PRO^D^3^SG im-he))
               (ADVP-DIR (ADV herod-herod))
               (IP-INF-SPE (VB drincan-drinkan)
                           (PP (P te-te)
                               (NP (PRO^D^1^SG mi-ik)))
                           (NP-OB1 *ICH*-2))
               (' '-')
               (IP-MAT-PRN (VBDI^3^SG quat$-kwethan)
                           (NP-SBJ (PRO^N^3^SG $hie-he)))
               (, ,-,)
               (CODE <C>)
               (' '-')
               (NP-TMP (NP-POS (N^G^PL dago-dag))
                       (GE+WADJ^G^SG gihuilikes-gihwilik))
               (CODE <R_3914>)
               (NP-OB1-2 (ADJ^A^PL suotian-swoti)
                         (N^A^PL brunnon-brunno)))
          (. .-.))
  (ID OSHeliandC.2262.3912-3914))

Most complete CPs must have a complementizer (C), which is represented as phonologically 0 when not overtly present. Direct questions, yes/no-questions, and V1 conditionals do not require a C.

( (IP-MAT (CODE <C>)
          (ADVP (ADVR Uuirss-wirs))
          (BEPI^3^SG is-wesan)
          (NP-OB2 (D^D^PL them-the)
                  (ADJP (ADJ^D^PL ođron-othar)))
          (, ,-,)
          (CODE <R_1348>)
          (ADJP-PRD (ADJ^N^SG giƀiđig-gividig))
          (NP-SBJ (ADJR^N^SG grimmera-grim)
                  (N^N^SG thing-thing))
          (CODE <MS_39b>)
          (CODE <C>)
          (NP-OB2-PRN (D^D^PL them-the)
                      (CP-REL-SPE (WNP-SBJ-1 (D^N^PL thia-the))
                                  (C 0)
                                  (IP-SUB-SPE (NP-SBJ *T*-1)
                                              (ADVP-LOC (ADV hier-her))
                                              (NP-OB1 (N^A^SG guod-god))
                                              (VBPI^3^PL egun-egan)
                                              (, ,-,)
                                              (CODE <R_1349>)
                                              (NP-OB1-PRN (ADJ^A^SG uuidon-wid)
                                                          (N^A^SG uuerolduuelon-weroldwelo)))))
          (. :-:))
  (ID OSHeliandC.737.1347-1349))

( (CP-QUE-SPE (CODE <R_923>)
              (IP-SUB-SPE (BEPI^2^SG Bist-wesan)
                          (NP-SBJ (PRO^N^2^SG thu-thu))
                          (NP-PRD (Q^N^SG enig-enig)
                                  (NP-POS (D^G^PL thero-the)
                                          (CODE <C>)
                                          (CP-REL-SPE (WNP-SBJ-1 (D^N^SG thi-the))
                                                      (C 0)
                                                      (IP-SUB-SPE (NP-SBJ *T*-1)
                                                                  (ADVP-LOC (ADV hier-her))
                                                                  (ADVP-TMP (ADV er-er))
                                                                  (BEDS^3^SG uuari-wesan))))
                                  (CODE <R_924>)
                                  (NP-POS-PRN (ADJ^G^PL uuissaro-wis)
                                              (N^G^PL uuarsagono-warsago))))
              (. ?-?))
  (ID OSHeliandC.498.923-924))

In CP-CMP, CP-REL, CP-FRL and wh-CP-QUE, wh-operators must be present, and are represented as phonologically 0 when not overtly present. The clause from which the wh-operator is extracted must contain a trace of the same category, or else a resumptive. Unlike in the YCOE, this trace is marked with all the same extended labels as the operator itself. The extended labels of wh-operators are identical in every way to those of overt constituents.

( (IP-MAT-SPE (CODE <C>)
              (VBI^2^SG Sagi-seggian)
              (NP-OB2 (PRO^D^1^PL us-we))
              (CP-QUE-SPE (WNP-PRD-1 (WPRO^N huat-hwe)
                                     (NP-POS *ICH*-2))
                          (C 0)
                          (IP-SUB-SPE (NP-PRD *T*-1)
                                      (NP-SBJ (PRO^N^2^SG thu-thu))
                                      (NP-POS-2 (N^G^PL manno-man))
                                      (BEPS^2^SG sis-wesan)))
              (. !-!))
  (ID OSHeliandC.497.922))

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (VBDI^3^SG quat-kwethan)
          (CP-THT (C that-that)
                  (IP-SUB (NP-SBJ (D^N^PL thia-the)
                                  (CP-REL *ICH*-1))
                          (ADJP-PRD (ADJ^N^PL saliga-salig))
                          (BEDS^3^PL uuarin-wesan)
                          (CODE <R_1301>)
                          (NP-SBJ-PRN (N^N^PL mann-man))
                          (PP (P an-an)
                              (NP (D^D^SG thesaro-these)
                                  (N^D^SG middilgardun-middilgarda)))
                          (CODE <C>)
                          (CP-REL-1 (WNP-SBJ-2 (D^N^PL thia-the))
                                    (C 0)
                                    (IP-SUB (NP-SBJ *T*-2)
                                            (ADVP-LOC (ADV her-her))
                                            (PP (P an-an)
                                                (NP (PRO$^D^3^SG iro-iru)
                                                    (N^D^SG muode-mod)))
                                            (BEDS^3^PL uuarin-wesan)
                                            (CODE <R_1302>)
                                            (ADJP-PRD (ADJ^N^PL arma-arm))
                                            (PP (P thuru-thurh)
                                                (NP (N^A^SG ođmuodig-othmodi)))))))
          (. :-:))
  (ID OSHeliandC.712.1300-1302))

Scrambling and extraposition are indicated by *ICH* rather than *T*, and are only represented if the movement in question has taken the moved constituent out of its source category (that is, structurally vacuous "linear" scrambling and extraposition do not exist in this annotation scheme). Traces are always as early as possible in the source constituent if the movement is upward/to the left, and as late as possible if the movement is downward/to the right. With non-finite clauses, elements are traced to the position of the non-finite verb, which is taken as the position of the clause.

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Than-than))
          (MDPI^2^PL motun-motan)
          (NP-SBJ (PRO^N^2^PL gi-gi))
          (NP-OB1 (D^A^SG thia-the)
                  (N^A^SG fruma-fruma)
                  (CP-THT-SPE *ICH*-1))
          (VB egan-egan)
          (CODE <R_1461>)
          (CP-THT-SPE-1 (C that-that)
                        (IP-SUB-SPE (NP-SBJ (PRO^N^2^PL gi-gi))
                                    (MDPI^2^PL muotun-motan)
                                    (VB hetan-hetan)
                                    (CODE <C>)
                                    (NP-PRD (NP-POS (N^G^SG heƀancuninges-hevankuning))
                                            (N^N^PL suni-sunu))
                                    (, ,-,)
                                    (CODE <R_1462>)
                                    (NP-PRD-PRN (PRO$^N^3^PL is-is)
                                                (ADJ^N^PL blithi-blithi)
                                                (N^N^PL barn-barn))))
          (. .-.))
  (ID OSHeliandC.787.1460-1462))

An index number on the scrambled/extraposed constituent indicates that it does not belong to (i.e. originate in) the phrase that immediately dominates it.

( (IP-MAT-SPE (CODE <C>)
              (, –-–)
              (NP-SBJ (PRO^N^2^PL gi-gi))
              (VBDI^2^PL hordun-horian)
              (NP-OB1-1 (PRO^A^3^SG it-it))
              (ADVP-TMP (ADV oft-oft))
              (IP-INF-SPE (NP-OB1 *ICH*-1)
                          (VB sprekean-sprekan)
                          (CODE <P_102>)
                          (CODE <R_1433>)
                          (NP-ADT (N^D^PL uuordun-word))
                          (NP-SBJ (ADJ^A^PL uuisa-wis)
                                  (N^A^PL man-man)))
              (, –-–)
              (. :-:))
  (ID OSHeliandC.776.1432-1433))

Raising to subject is not usually explicitly represented in the HeliPaD.

The exception is tough-movement, which is represented as in the following example, using a default empty category:

(CP-ADV-LFD-SPE (ADVP (ADV thoh-thoh))
                (C 0)
                (IP-SUB-SPE (NP-SBJ-1 (PRO^N^3^SG hie-he))
                            (ADJP (ADV so-so)
                                  (ADJ^N^SG othi-othi)
                                  (IP-INF-SPE *ICH*-1))
                            (NEG ni-ne)
                            (BEPS^3^SG si-wesan)
                            (CODE <R_1787>)
                            (NP-OB2 (N^D^PL firihon-firihos))
                            (IP-INF-SPE-1 (NP-OB1 *-1)
                                          (TO te-te)
                                          (VB faranne-faran))))

(OSHeliandC.940.1786-1788)

Expletive constructions

Among overt elements, only the pronoun it is treated as an expletive; similar constructions involving that are treated as instances of apposition. Expletive coindexing involves a final -X, as in the YCOE (though it is upper case, for technical reasons). Extraposed clausal subjects (that-clauses, infinitives or questions) are coindexed to an overt (it) or null (*exp*) expletive. These extraposed clausal subjects occur in various constructions.

( (IP-MAT (CODE <C>)
          (NP-SBJ-X *exp*)
          (NP-OB2 (PRO^D^3^SG Im-he))
          (BEDI^3^SG uuas-wesan)
          (ADVP-TMP (ADV thuo-tho))
          (NP-PRD (N^N^SG uuilleo-willio)
                  (ADJ^N^SG mikil-mikil))
          (CODE <R_872>)
          (CP-THT-X (C that-that)
                    (IP-SUB (NP-SBJ (PRO^N^3^SG hie-he))
                            (PP (P fan-fan)
                                (NP (ADJ^D^PL solicon-sulik)
                                    (N^D^PL salđon-salitha)))
                            (CODE <C>)
                            (VB seggean-seggian)
                            (MDDS^3^SG muosti-motan)))
          (. .-.))
  (ID OSHeliandC.470.871-872))

( (IP-MAT (CODE <C>)
          (NP-SBJ-X *exp*)
          (BEDI^3^SG uuas-wesan)
          (NP-OB2 (PRO^D^3^PL im-he))
          (NP-PRD (NP-POS (PRO$^G^3^SG is-is)
                          (N^G^SG huldi-huldi))
                  (NP-POS-PRN (N^G^PL helpono-helpa))
                  (N^N^SG tharf-tharf))
          (CODE <R_1188>)
          (IP-INF-X (TO te-te)
                    (GE+VB githiononne-thionon))
          (. :-:))
  (ID OSHeliandC.650.1187-1188))

( (IP-MAT (CODE <C>)
          (NP-SBJ-X *exp*)
          (BEDI^3^SG uuas-wesan)
          (NP-OB2 (PRO^D^3^PL im-he))
          (NP-PRD (N^N^SG niud-niud)
                  (ADJ^N^SG mikil-mikil))
          (CODE <R_183>)
          (CP-QUE-X (WNP-OB1-1 (WPRO^A huat-hwe)
                               (NP-POS *ICH*-2))
                    (C 0)
                    (IP-SUB (IP-SUB-0 (NP-OB1 *T*-1)
                                      (NP-SBJ (PRO^N^3^SG he-he))
                                      (NP-OB2 (PRO^D^3^PL im-he))
                                      (NP-POS-2 (ADJ^G^SG suodlicas-sothlik))
                                      (CODE <C>)
                                      (VB seggean-seggian)
                                      (MDDS^3^SG uueldi-willian))
                            (, ,-,)
                            (CODE <R_184>)
                            (CONJP (IPX-SUB=0 (VB uuisan-wisian)
                                              (PP (P te-te)
                                                  (NP (N^D^PL uuarun-war)))))))
          (. .-.))
  (ID OSHeliandC.86.182-184))

Other expletive subjects occur in impersonal constructions, and are not coindexed.

( (IP-MAT (CODE <C>)
          (NP-SBJ *exp*)
          (ADVP-TMP (ADV Thuo-tho))
          (RDDI^3^SG uuarth-werthan)
          (NP-OB2 (Q^D^PL managon-manag))
          (ADJP-PRD (ADJ^N^SG cuth-kuth))
          (CODE <R_387>)
          (PP (P oƀer-ovar)
              (NP (D^A^SG thesa-these)
                  (ADJ^A^SG uuidun-wid)
                  (N^A^SG uuerold-werold)))
          (. ,-,))
  (ID OSHeliandC.213.386-387))

( (IP-MAT-SPE (CODE <R_418>)
              (' '-')
              (NP-SBJ *exp*)
              (NP-PRD (NP-PRD (N^N^SG Diuritha-diuritha))
                      (CONJP *ICH*-1))
              (BEPS^3^SG si-wesan)
              (ADVP-TMP (ADV nu-nu))
              (' '-')
              (IP-MAT-PRN (VBDI^3^PL quathun-kwethan)
                          (NP-SBJ (PRO^N^3^PL sia-he)))
              (, ,-,)
              (CODE <C>)
              (' '-')
              (NP-OB2 (N^D^SG drohtine-drohtin)
                      (ADJ^D^SG selƀon-self))
              (CODE <R_419>)
              (PP (P an-an)
                  (NP (D^D^SG thiem-the)
                      (ADJS^D^SG hohoston-hoh)
                      (CODE <C>)
                      (NP-POS (N^G^PL himilo-himil))
                      (N^D^SG rikea-riki)))
              (, ,-,)
              (CODE <R_420>)
              (CONJP-1 (CONJ endi-endi)
                       (NP-PRD (N^N^SG friđu-frithu)))
              (PP (P an-an)
                  (NP (N^D^SG erđu-ertha)))
              (CODE <C>)
              (NP-OB2-PRN (NP-POS (N^G^PL firio-firihos))
                          (N^D^PL barnon-barn))
              (, ,-,)
              (CODE <R_421>)
              (NP-OB2-PRN (ADJ^D^PL guoduuilligon-godwillig)
                          (N^D^PL gumon-gumo)
                          (, ,-,)
                          (CODE <C>)
                          (CP-REL (WNP-SBJ-2 (D^N^PL thie-the))
                                  (C thia-the)
                                  (IP-SUB (NP-SBJ *T*-2)
                                          (NP-OB1 (NPR^A^SG god-god))
                                          (VBPI^3^PL ankenneat-andkennian)
                                          (CODE <R_422>)
                                          (PP (P thuru-thurh)
                                              (NP (ADJ^A^SG hluttran-hluttar)
                                                  (N^A^SG hugi-hugi))))))
              (. .-.)
              (' '-'))
  (ID OSHeliandC.229.418-422))

Conjunction

Conjunction of everything other than matrix clauses with coreferential subjects is represented using CONJP, a phrasal category, usually endocentric with CONJ as its head. The basic structure is as follows:

(XP1 (XP2 first_conjunct)
     (CONJP (CONJ conjunction)
	    (XP3 second_conjunct)))

CONJP is sister to the first conjunct, with the second conjunct acting as complement within the CONJP (following the CONJ itself, if there is one).

(PP (PP (P mid-mid)
        (NP (N^D^PL uuordun-word)))
    (CONJP (CONJ endi-endi)
           (PP (P mid-mid)
               (NP (N^D^PL uuercun-werk)))))

(OSHeliandC.1.1-5)

Third and subsequent conjuncts are also within CONJPs and sisters to the first conjunct.

(PP (PP (P mid-mid)
        (NP (ADJ^I^SG leohtu-lioht)
            (N^I^SG hugi-hugi)))
    (CODE <C>)
    (CONJP (CONJ endi-endi)
           (PP (P mid-mid)
               (NP (N^D^SG giloƀon-gilovo)
                   (ADJ^D^SG guodon-god))))
    (CODE <P_24>)
    (CODE <R_291>)
    (CONJP (CONJ endi-endi)
           (PP (P mid-mid)
               (NP (ADJ^D^PL hluttron-hluttar)
                   (N^D^PL treuuon-treuwa)))))

(OSHeliandC.160.288-291)

In cases of "balanced" conjunction, i.e. where a conjunction precedes the first conjunct, it is included under the root node.

(NP-OB1-PRN (CONJ ge-ge)
            (NP-OB1-PRN (N^A^SG hosc-hosk))
            (CONJP (CONJ gi-ge)
                   (NP-OB1-PRN (N^A^SG harmquidi-harmkwidi))))

(OSHeliandC.987.1894-1896)

The HeliPaD's approach to conjunction differs in three important ways from that of the YCOE and other Penn corpora. First, single-word conjuncts are treated in exactly the same way as other conjuncts. Secondly, any extended labels borne by the root note are inherited by the two conjuncts. Both these points are illustrated by the above example. Thirdly, shared pre- and post-head modifiers are simply included under the root node, as illustrated by the below:

(NP-OB1 (NP-POS (NP-POS (Q^G^PL allaro-al)
                        (N^G^PL manno-man))
                (GE+WPRO^G gihues-gihwe))
        (CODE <C>)
        (NP-OB1 (N^A^PL muodgithahti-modgithaht))
        (, ,-,)
        (CODE <R_1926>)
        (CONJP (NP-OB1 (N^A^SG uuord-word)))
        (CONJP (CONJ endi-endi)
               (NP-OB1 (N^A^SG uuilleon-willio))))

(OSHeliandC.1001.1923-1926)

(NP-OB1-PRN (NP-OB1-PRN (N^A^PL scenkion-skenkio))
            (CONJP (CONJ endi-endi)
                   (NP-OB1-PRN (N^A^PL scapuuardos-skapward)))
            (, ,-,)
            (CODE <C>)
            (CP-REL (WNP-SBJ-1 (D^N^PL thia-the))
                    (C 0)
                    (IP-SUB (NP-SBJ *T*-1)
                            (ADVP-LOC (ADV thar-thar))
                            (MDDI^3^PL scoldun-skulan)
                            (NP-OB2 (D^D^SG thera-the)
                                    (N^D^SG scola-skola))
                            (CODE <MS_57b>)
                            (VB thionon-thionon))))

(OSHeliandC.1051.2032-2036)

When unlike constituents are conjoined, the dominating phrase will be the same as the first conjunct.

(QP-2 (QP (Q^A^SG manag-manag))
      (CONJP (CONJ endi-endi)
             (ADJP (ADJ^A^SG mislic-mislik))))

(OSHeliandC.985.1889-1892)

Finally, focus-sensitive adverbs and ADVPs such as ok (tagged ALSO), so self and so sama may be treated as part of a CONJP, whether preceding or following the conjunct.

(CONJP (CONJ endi-endi)
       (ALSO oc-ok)
       (NP-OB1 (D^A^SG that-the)
               (N^A^SG barn-barn)
               (NP-POS (NPR^G^SG godes-god)))

(OSHeliandC.257.478-480)

(PP (PP (P umbi-umbi)
        (NP (NP-POS (D^G^PL thesaro-these)
                    (N^G^PL liudio-liudi))
            (N^A^SG lif-lif)))
    (CODE <C>)
    (CONJP (CONJ endi-endi)
           (ALSO oc-ok)
           (PP (P umbi-umbi)
               (NP (D^A^SG thit-these)
                   (N^A^SG land-land)))
           (ADVP (ADV so-so)
                 (ADV samo-sama)))))

(OSHeliandC.2163.3755-3757)

Elision

Noun phrase elision is never indicated in the HeliPaD, but clauses may be complete (IP-, CP-) or incomplete (IPX-, CPX-). Elision is the usual source of such incomplete clauses; textual problems are treated differently, using the X phrase. Incomplete clauses arising from elision have equal-sign coindexing: the incomplete clause will bear =0 (or =00, =000 etc.), while the clause it is patterned on will be marked -0 (or -00, -000 etc.).

( (IP-MAT-SPE-0 (CODE <C>)
                (CONJ endi-endi)
                (VBPI^3^SG gangat-gangan)
                (NP-OB2 (PRO^D^3^SG im-he))
                (NP-SBJ (N^N^SG diuball-diuval))
                (ADVP-DIR (ADV ferr-fer))
                (, ,-,)
                (CODE <R_2481>)
                (NP-SBJ-PRN (ADJ^N^PL uuretha-wreth)
                            (CODE <MS_69b>)
                            (N^N^PL uuihti-wiht))
                (CODE <C>)
                (IPX-MAT-PRN-SPE=0 (CONJ endi-endi)
                                   (NP-SBJ (D^N^SG thie-the)
                                           (N^N^SG uuard-ward)
                                           (NP-POS (NPR^G^SG godes-god)))
                                   (CODE <R_2482>)
                                   (ADVP-DIR (ADVR nahor-nah)
                                             (ADJ^I^SG mikilu-mikil)))
                (CODE <C>)
                 ...
                (. ;-;))
  (ID OSHeliandC.1328.2480-2485))

VP conjunction is not distinguished from IP conjunction due to the lack of a VP node. It is treated as elision, using an IPX with equal-sign coindexing.

( (IP-MAT-SPE (CODE <C>)
              (IP-MAT-SPE-0 (ADVP-TMP (ADV Than-than))
                            (MDPI^2^SG scalt-skulan)
                            (NP-SBJ (PRO^N^2^SG thu-thu))
                            (ADVP-TMP (ADV eft-eft))
                            (NP-OB1 (uuord-word N^A^SG))
                            (VB sprekan-sprekan))
              (, ,-,)
              (CODE <R_169>)
              (CONJP (IPX-MAT-SPE=0 (HV hebbean-hebbian)
                                    (NP-OB1-PRN (NP-POS (PRO$^G^2^SG thinera-thin)
                                                        (N^G^SG stemna-stemna))
                                                (N^A^SG giuuald-giwald))))
              (. :-:))
  (ID OSHeliandC.78.168-169))

In right-node raising the second conjunct is placed in a parenthetical IPX, itself at IP level, without a CONJP.

( (IP-MAT-0 (CODE <R_1075>)
            (ADVP-TMP (ADV Thuo-tho))
            (VBDI^3^SG bigan-biginnan)
            (ADVP-TMP (ADV eft-eft))
            (VB niuson-niuson)
            (CODE <C>)
            (IPX-MAT-PRN=0 (CONJ endi-endi)
                           (ADVP-DIR (ADVR nahor-nah))
                           (VBDI^3^SG gieng-gangan))
            (CODE <R_1076>)
            (NP-SBJ (ADJ^N^SG unhiuri-unhiuri)
                    (N^N^SG fiond-fiund))
            (CODE <C>)
            (NP-TMP (ADJ^I^SG othar-othar)
                    (N^I^SG sithu-sith))
            (. ,-,))
  (ID OSHeliandC.586.1075-1076))

Unindexed IPX is used in certain types of comparative (with CPX) and in restarts.

CPs can also be incomplete (CPX) in certain types of comparative, and in that-clauses with an elided complementizer.

Negation

The sentential negator in Old Saxon is the particle ne (NEG), placed at IP level, usually directly preverbal. The negator ne plus eowiht is always parsed as a nominal constituent, and hence as involving constituent negation, regardless of whether or not it is inflected. In constituent negation, NEG is contained within the negated constituent, and often cliticized onto the following word (in the manuscript perhaps even more often than in the Sievers edition).

( (IP-MAT-SPE (CODE <C>)
              (NEG Ni-ne)
              (HVI^2^SG haƀi-hebbian)
              (NP-SBJ (PRO^N^2^SG thu-thu))
              (NP-OB1 (ADJ^A^SG uuekean-wek)
                      (N^A^SG hugi-hugi))
              (. ,-,))
  (ID OSHeliandC.139.262))

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (NEG ni-ne)
          (VBDI^3^SG antuuordida-andwordian)
          (NP-OB1 (NEG+N^A^SG niouuiht-eowiht))
          (CODE <R_5383>)
          (PP (P uuiđ-with)
              (NP (PRO$^A^3^PL iro-he)
                  (ADJ^A^PL uurethun-wreth)
                  (N^A^PL uuord-word)))
          (. :-:))
  (ID OSHeliandC.3187.5382-5383))

Clauses

Clause structure

As described under phrase structure, the clause structure representations in the HeliPaD are quite flat due to the lack of a VP node. In addition, there are a number of word-level categories that may be immediately dominated by the IP with no phrasal node intervening. These are:

  • All verbs, finite and non-finite
  • Particles (RP, ALSO)
  • Sentential conjunctions (CONJ)
  • Negation (NEG)
  • Single-word interjections (INTJ)
  • The word wita (UTP)
( (IP-MAT-SPE (CODE <C>)
              (CONJ endi-endi)
              (NP-SBJ *con*)
              (ADVP-TMP (ADV eft-eft))
              (RP upp-up)
              (GE+VBDI^3^SG gigeng-gangan)
              (. ,-,))
  (ID OSHeliandC.1288.2408))

( (IP-MAT-SPE (CODE <C>)
              (CONJ ac-ak)
              (IP-MAT-SPE-0 (UTP uuita-wita)
                            (NP-1 (PRO^D^3^SG im-he))
                            (VB uuonian-wonon)
                            (PP (NP *ICH*-1)
                                (P mid-mid)))
              (, ,-,)
              (CODE <R_3996>)
              (CONJP (IPX-MAT-SPE=0 (VB thuoloian-tholon)
                                    (PP (P mid-mid)
                                        (NP (PRO$^D^1^SG usson-usa)
                                            (N^D^SG thiodne-thiodan)))))
              (. :-:))
  (ID OSHeliandC.2317.3995-3996))

Any complete finite clause (IP-MAT*, IP-SUB*) will contain at least a subject and a finite verb. See Subjects for details of the former.

There are also two types of non-finite IPs: infinitives (IP-INF*) and small clauses (IP-SMC). The former are headed by a non-finite verb, and do not necessarily have a subject (though they may). The latter always contain a subject and a non-verbal predicate or participial verb form. In addition, participial phrases behave like clauses in some respects.

The structure of CPs is discussed under Subordinate clauses.

Clausal extended labels

Clausal phrases (IPs, CPs) must have one of the following extended labels:

For IPs:

  • -MAT: main clause
  • -SUB: subordinate clause, inc. direct questions
  • -INF: infinitival clause
  • -INF-NCO: non-complement infinitival clause
  • -SMC: small clause

For IPs, the extended label -SUB-CON is not used in the HeliPaD, as it is redundant and all instances of conjoined subordinate clauses can be retrieved in other ways. The label -ABS (for infinitival absolutes) is also not used.

For CPs:

  • -REL: relative clause
  • -FRL: free relative clause
  • -THT: that-clause
  • -ADV: adverbial clause
  • -CMP: comparative clause
  • -QUE: interrogative clause
  • -DEG: degree clause

For CPs, the HeliPaD does not include -CAR (clause-adjoined relatives), -CLF (clefts), -EXL (exclamatives), and -EOP (gapped infinitival relative/purpose clauses), as these structures are essentially not found in the Heliand.

Following these functional extended labels, an IP or a CP may additionally bear one or more of the following, in that order:

(CP-FRL-OB1-PRN-SPE (WNP-SBJ-1 (ADV so-so)
                               (WPRO^N huat-hwe))
                    (C so-so)
                    (IP-SUB-SPE (NP-SBJ *T*-1)
                                (NP-OB2 (PRO^D^2^SG thi-thu))
                                (ADJP-PRD (ADJ^N^SG gibidig-gividig))
                                (ADVP-DIR (ADV forth-forth))
                                (CODE <R_3379>)
                                (RD uuerthan-werthan)
                                (MDDS^3^SG scoldi-skulan)))

(OSHeliandC.1918.3377-3379)

Matrix clauses (IP-MAT)

The label for matrix clauses is IP-MAT. Almost all tokens are IP-MAT at the highest level.

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (VBDI^3^SG bigan-biginnan)
          (NP-OB2-RFL (PRO^D^3^SG im-he))
          (NP-SBJ (D^N^SG the-the)
                  (ADJ^N^SG uuiso-wis)
                  (N^N^SG mann-man))
          (, ,-,)
          (CODE <R_313>)
          (NP-SBJ-PRN (ADJP (ADV suitho-switho)
                            (ADJ^N^SG guod-god))
                      (N^N^SG gumo-gumo))
          (CODE <C>)
          (NP-SBJ-PRN (NPR^N^SG Ioseph-Joseph))
          (PP (P an-an)
              (NP (PRO$^D^3^SG is-is)
                  (N^D^SG muode-mod)))
          (CODE <R_314>)
          (VB thenkean-thenkian)
          (NP-OB1 (D^G^PL thero-the)
                  (N^G^PL thingo-thing))
          (, ,-,)
          (CODE <C>)
          (CP-QUE (WADVP-1 (WADV huo-hwo))
                  (C 0)
                  (IP-SUB (ADVP *T*-1)
                          (NP-SBJ (PRO^N^3^SG hie-he))
                          (NP-OB1 (D^A^SG thea-the)
                                  (N^A^SG thiornun-thiorna))
                          (ADVP-TMP (ADV tho-tho))
                          (CODE <R_315>)
                          (NP-ADT (N^D^PL listion-list))
                          (VBDS^3^SG forlieti-farlatan)))
          (. .-.))
  (ID OSHeliandC.174.312-315))

( (IP-MAT (CODE <R_334>)
          (NP-SBJ *con*)
          (VBDI^3^SG bisorogoda-bisorgon)
          (NP-OB1 (PRO^A^3^SG sea-siu))
          (PP (P an-an)
              (NP (PRO$^D^3^SG is-is)
                  (N^D^SG gisitha-gisith)))
          (. ,-,))
  (ID OSHeliandC.190.334))

Imperative clauses are also IP-MAT, but are distinguished by having an imperative verb form (e.g. VBI) and no requirement for a subject.

( (IP-MAT-SPE (CODE <C>)
              (NEG Ni-ne)
              (VBI^2^PL latat-latan)
              (IP-INF (NP-SBJ (PRO$^A^2^SG iuuan-iuwa)
                              (N^A^SG hugi-hugi))
                      (VB tuiflean-twiflian))
              (. !-!)
              (' '-'))
  (ID OSHeliandC.514.948))

( (IP-MAT-SPE (CODE <C>)
              (HVI^2^PL Hebbeat-hebbian)
              (NP-OB1 (PRO$^A^2^SG iuuan-iuwa)
                      (N^A^SG uuilleon-willio))
              (ADVP-DIR (ADV tharod-tharod))
              (, ,-,)
              (CODE <R_944>)
              (NP-VOC (N^N^PL liudi-liudi))
              (NP-OB1-PRN (PRO$^A^2^SG iuuan-iuwa)
                          (N^A^SG giloƀon-gilovo))
              (. :-:))
  (ID OSHeliandC.512.943-944))

( (IP-MAT-SPE (CODE <R_1067>)
              (GE+VBI^2^SG Giheli-helian)
              (NP-OB1 (PRO$^A^2^SG thinan-thin)
                      (N^A^SG hungar-hungar))
              (. .-.)
              (' '-'))
  (ID OSHeliandC.580.1067))

Clauses containing wita (UTP) as main verb are labelled IP-MAT, and like imperatives have no requirement for a subject to be present.

( (IP-MAT-SPE (CODE <C>)
              (UTP Uuita-wita)
              (VB kiesan-kiosan)
              (NP-OB2 (PRO^D^3^SG im-he))
              (NP-OB1 (ADJ^A^SG ođerna-othar)
                      (CODE <R_224>)
                      (ADJ^A^SG niudsamana-niudsam)
                      (N^A^SG namon-namo))
              (. :-:))
  (ID OSHeliandC.113.223-224))

Subordinate clauses

As a rule, subordinate clauses are CPs, at least when they are finite. All complete CPs are labelled for type and contain an IP-SUB or IPX-SUB complement. In wh-questions, relatives, free relatives, comparatives, etc., they will also contain a wh-phrase (overt or empty). All CPs contain a C head in the HeliPaD, except direct questions and conditional inversion.

The HeliPaD does not share the Penn corpora's strange fetish for treating subordinating conjunctions as prepositions.

(CP-CMP (WADVP-TMP-2 (ADV than-than)
                     (ADV lang-lang))
        (C the-the)
        (IP-SUB (ADVP-TMP *T*-2)
                (NP-OB1 (PRO^A^3^SG sia-siu))
                (NP-SBJ (NP-POS (N^G^PL firio-firihos))
                        (N^N^PL barn-barn))
                (CODE <R_4455>)
                (VB ardon-ardon)
                (MDPI^3^PL muotun-motan)))

(OSHeliandC.2580.4452-4456)

The position in front of the wh-word or complementizer ("SpecCP", though no theoretical significance should be attached to this term) can host elements such as adverbs, left-dislocated constituents, etc.

(CP-ADV-SPE (C huand-hwand)
            (IP-SUB-SPE (CP-FRL-LFD-SPE (WNP-SBJ-1 (ADV so-so)
                                                   (WPRO^N hue-hwe))
                                        (C so-so)
                                        (IP-SUB-SPE (NP-SBJ *T*-1)
                                                    (NP-OB1 (NP-POS (N^G^PL uuapno-wapan))
                                                            (N^A^SG nith-nith))
                                                    (, ,-,)
                                                    (CODE <R_4897>)
                                                    (NP-OB1-PRN (ADJ^A^SG grimman-grim)
                                                                (N^A^SG gerheti-gerheti))
                                                    (MDPI^3^SG uuil-willian)
                                                    (CODE <C>)
                                                    (ADVP (ADV gerno-gern))
                                                    (VB frummean-frummian)))
                        (, ,-,)
                        (CODE <R_4898>)
                        (IP-SUB-SPE (NP-SBJ-RSP (PRO^N^3^SG hie-he))
                                    (VBPI^3^SG sueltiđ-sweltan)
                                    (NP-OB2-RFL (PRO^D^3^SG im-he))
                                    (ADVP-TMP (ADV oft-oft))
                                    (CODE <C>)
                                    (NP-ADT (NP-POS (N^G^SG suerdes-swerd))
                                            (N^I^PL eggion-eggia)))
                        (, ,-,)
                        (CODE <R_4899>)
                        (CONJP (IP-SUB-SPE (NP-SBJ *con*)
                                           (VBPI^3^SG doit-doian)
                                           (NP-OB2-RFL (PRO^D^3^SG im-he))
                                           (ADJP-ADT (ADJ^N^SG drorag-drorag))))))

(OSHeliandC.2885.4895-4899)

In subordinate clauses, elements before a single that-complementizer are treated as in SpecCP, and elements between two that-complementizers are treated as involving a that-clause embedded within another that-clause, as in the below.

( (IP-MAT-SPE (CODE <R_587>)
              (NP-SBJ (PRO^N^3^SG Hie-he))
              (VBDI^3^SG quat-kwethan)
              (CP-THT-SPE (C that-that)
                          (CP-THT-SPE (PP (P an-an)
                                          (NP (D^D^SG them-the)
                                              (ADJ^D^SG selƀon-self)
                                              (N^D^SG dage-dag)
                                              (CODE <C>)
                                               ...
                                      (, ,-,)
                                      (CODE <R_589>)
                                      (IP-MAT-PRN-SPE (ADVP (ADV so-so))
                                                      (VBDI^3^SG quat-kwethan)
                                                      (NP-SBJ (PRO^N^3^SG he-he)))
                                      (C that-that)
                                      (IP-SUB-SPE (ADVP-DIR (ADV ostana-ostana))
                                                  (CODE <C>)
                                                  (NUMP-2 (NUM^N^SG en-en))
                                                  (MDDS^3^SG scoldi-skulan)
                                                  (VB scinan-skinan)
                                                  (CODE <R_590>)
                                                   ...
              (. .-.))
  (ID OSHeliandC.311.587-592))

Adverbial clauses (CP-ADV)

CP-ADV is the label used for adverbial clauses, which may have a variety of forms and functions.

In the HeliPaD, non-argumental that-clauses (e.g. with resultative meaning) are still labelled CP-THT and not CP-ADV.

Most subordinating conjunctions are treated as ADVPs in SpecCP along with a null C.

(CP-ADV (ADVP-TMP (ADV thuo-tho))
        (C 0)
        (IP-SUB (NP-SBJ (PRO^N^3^PL sia-he))
                (ADVP-TMP (ADV eft-eft))
                (PP (P te-te)
                    (NP (N^D^SG heƀanuuange-hevanwang)))
                (CODE <R_415>)
                (VBDI^3^PL uundun-windan)
                (PP (P thuru-thurh)
                    (NP (D^A^PL thiu-the)
                        (N^A^PL uuolkan-wolkan)))))

(OSHeliandC.227.414-415)

(CP-ADV (ADVP-LOC (ADV thar-thar))
        (C 0)
        (IP-SUB (NP-SBJ (PRO^N^3^SG hie-he))
                (PP (P an-an)
                    (NP (PRO$^D^3^SG is-is)
                        (N^D^SG rikie-riki)))
                (VBDI^3^SG sat-sittian)))

(OSHeliandC.384.715-718)

Such clauses are treated as relative (CP-REL) if, and only if, there is a coreferential adverb preceding it in the clause.

( (IP-MAT (CODE <R_1136>)
          (NP-SBJ *con*)
          (VBDI^3^SG fuor-faran)
          (NP-OB2-RFL (PRO^D^3^SG im-he))
          (PP (P te-te)
              (NP (D^D^PL them-the)
                  (N^D^PL friundun-friund)
                  (CODE <C>)
                  (CP-REL (WADVP-LOC-1 (ADV thar-thar))
                          (C 0)
                          (IP-SUB (IP-SUB-0 (ADVP-LOC *T*-1)
                                            (NP-SBJ (PRO^N^3^SG hie-he))
                                            (VN afuodid-afodian)
                                            (BEDI^3^SG uuas-wesan))
                                  (, ,-,)
                                  (CODE <R_1137>)
                                  (CONJP (IPX-SUB=0 (ADVP (ADV tirlico-tirliko))
                                                    (VN atogan-atiohan)))))))
          (. ,-,))
  (ID OSHeliandC.620.1136-1137))

A limited number of subordinators, mostly those that are not homophonous with adverbs, are treated as C elements (gif, hwand, ne (NEG+C), newa, so, than in comparatives, that, the, untat).

(CP-ADV (C unthat-untat)
        (IP-SUB (NP-SBJ (PRO^N^3^SG hie-he))
                (PP (P te-te)
                    (NP (N^D^SG middean-middia)))
                (VBDI^3^SG quam-kuman)
                (CODE <R_2241>)
                (NP-SBJ-PRN (N^N^SG uualdand-waldand))
                (PP (P mid-mid)
                    (NP (PRO$^I^3^SG is-is)
                        (N^I^SG uuerodu-werod)))))

(OSHeliandC.1168.2239-2241)

(CP-ADV (C huand-hwand)
        (IP-SUB (ADVP-LOC (ADV thar-thar))
                (BEDI^3^SG uuas-wesan)
                (NP-POS-1 (N^G^PL gumono-gumo))
                (GE+VN gitald-tellian)
                (CODE <R_2871>)
                (PP (P ano-ano)
                    (NP (NP (N^A^SG uuiƀ-wif))
                        (CONJP (CONJ endi-endi)
                               (NP (N^A^SG kind-kind)))))
                (CODE <C>)
                (NP-SBJ-PRN (N^N^PL uueros-wer))
                (ADVP (ADV atsamma-atsamne))
                (CODE <R_2872>)
                (NP-SBJ (NP-POS *ICH*-1)
                        (NUM^N^PL fif-fif)
                        (NUM^N^PL thusundig-thusundig))))

(OSHeliandC.1566.2869-2872)

As in the YCOE, so-clauses are treated as CP-ADV and not CP-CMP when there is no obvious extraction or comparison involved, particularly in the combinations reht so and al so - but this can be a tricky distinction to make.

( (IP-MAT-SPE (CODE <R_155>)
              (BEPI^3^PL sind-wesan)
              (NP-SBJ (PRO$^N^1^PL unca-unka)
                      (N^N^PL andbari-andbari))
              (CODE <C>)
              (ADJP-PRD (ADJR^N^PL odarlicron-otharlik))
              (, ,-,)
              (CODE <R_156>)
              (NP-SBJ-PRN (NP-SBJ-PRN (N^N^SG muod-mod))
                          (CONJP (CONJ endi-endi)
                                 (NP-SBJ-PRN (N^N^SG megincraft-meginkraft))))
              (, ,-,)
              (CODE <C>)
              (CP-ADV-SPE (C so-so)
                          (IP-SUB-SPE (NP-SBJ (PRO^N^1^DU uuit-wit))
                                      (ADVP-TMP (ADV iu-ju))
                                      (NP-TMP (QP (ADV so-so)
                                                  (Q^A^SG managan-manag))
                                              (N^A^SG dag-dag))
                                      (CODE <R_157>)
                                      (BEDI^1^PL uuarun-wesan)
                                      (PP (P an-an)
                                          (NP (D^D^SG thesaro-these)
                                              (CODE <MS_9a>)
                                              (NP-OB2 (N^D^SG uueroldi-werold))))))
              (. :-:))
  (ID OSHeliandC.74.155-157))

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (HVDI^3^SG habda-hebbian)
          (NP-SBJ (PRO^N^3^SG hie-he))
          (NP-OB1 (NP-POS (PRO$^G^1^SG usas-usa)
                          (N^G^SG uualdandes-waldand))
                  (CODE <R_191>)
                  (N^A^SG geld-geld))
          (GE+VN gilestid-lestian)
          (, ,-,)
          (CODE <C>)
          (CP-ADV (ADVP (ADV all-al))
                  (C so-so)
                  (IP-SUB (NP-SBJ (PRO$^N^3^SG is-is)
                                  (N^N^SG gigengi-gigengi))
                          (BEDI^3^SG uuas-wesan)
                          (CODE <R_192>)
                          (GE+VN gimarcod-markon)
                          (PP (P mid-mid)
                              (NP (N^D^PL mannon-man)))))
          (. .-.))
  (ID OSHeliandC.89.190-192))

Purpose clauses headed by a genitive demonstrative are not treated as CP-ADV but as CP-FRL-ADT.

V1 conditionals, usually unless-clauses, are CP-ADV with no C.

(CP-ADV (IP-SUB (NP-SBJ-X *exp*)
                (NEG ni-ne)
                (BEDS^3^SG uuari-wesan)
                (CP-THT-X (C that-that)
                          (IP-SUB (NP-SBJ (PRO^N^3^SG it-it))
                                  (NP-PRD (N^N^SG gibod-gibod)
                                          (NP-POS (NPR^G^SG godes-god)
                                                  (CODE <R_206>)
                                                  (ADJ^G^SG selƀes-self)))
                                  (BEDS^3^SG uuari-wesan)))))

(OSHeliandC.101.203-206)

That-clauses (CP-THT)

The HeliPaD is more liberal than the YCOE in its use of CP-THT, which is used broadly for any clause introduced by that, as well as some in the scope of negation introduced by ne, unless they are instances of CP-DEG. CP-THT is thus essentially a formal rather than functional label. CP-THT does not need to be the complement of a verb or adjective, and is usually unindexed at IP level.

(CP-THT (NEG+C ne-ne)
        (IP-SUB (NP-SBJ (PRO^N^3^SG siu-siu))
                (NP-OB1 (PRO^A^3^SG ina-he))
                (ADVP-TMP (ADV simla-simbla))
                (CODE <MS_16b>)
                (ADVP-DIR (ADV tharod-tharod))
                (CODE <R_457>)
                (PP (P te-te)
                    (NP (D^D^SG them-the)
                        (NP-POS (NPR^G^SG godes-god))
                        (N^D^SG uuihe-wih)))
                (CODE <C>)
                (VB forgeƀan-fargevan)
                (MDDI^3^SG scolda-skulan)))

(OSHeliandC.243.453-457)

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (GE+VBDI^1^SG gifragen-fregnan)
          (NP-SBJ (PRO^N^1^SG ik-ik))
          (CP-THT (C that-that)
                  (IP-SUB (NP-OB2 (PRO^D^3^SG iro-siu))
                          (ADVP-LOC (ADV thar-thar))
                          (NP-SBJ (N^N^SG sorga-sorga))
                          (GE+VBDI^3^SG gistuod-standan)))
          (, ,-,)
          (CODE <R_511>)
          (CP-THT-PRN (C that-that)
                      (IP-SUB (NP-OB1 (PRO^A^3^SG sia-siu))
                              (NP-SBJ (D^N^SG thiu-the)
                                      (ADJ^N^SG mikila-mikil)
                                      (N^N^SG maht-maht)
                                      (CODE <C>)
                                      (NP-POS (N^G^SG metođes-metod)))
                              (VBDI^3^SG tedelda-tedelian)
                              (, ,-,)
                              (CODE <R_512>)
                              (NP-SBJ-PRN (ADJ^N^PL uureth-wreth)
                                          (N^N^PL uurđigiscapu-wurdigiskap))))
          (. .-.))
  (ID OSHeliandC.271.510-512))

That-clauses may be subjects, as described in Expletive constructions.

That-clauses are quite often parenthetical to an overt demonstrative or object pronoun:

( (IP-MAT (CODE <C>)
          (BEDI^3^SG Uuas-wesan)
          (NP-SBJ (D^N^SG that-the)
                  (CP-THT-PRN *ICH*-1))
          (PP (PP (P an-an)
                  (NP (PRO$^D^3^PL is-is)
                      (N^D^PL uuordon-word)))
              (CONJP *ICH*-2))
          (ADJP-PRD (ADJ^N^SG scin-skin))
          (CODE <R_1208>)
          (CONJP-2 (CONJ iac-jak)
                   (PP (P an-an)
                       (NP (PRO$^D^3^PL is-is)
                           (N^D^PL dadeon-dad)))
                   (ADVP (ADV so-so)
                         (ADV samo-sama)))
          (CODE <C>)
          (CP-THT-PRN-1 (C that-that)
                        (IP-SUB (IP-SUB (NP-SBJ (PRO^N^3^SG hie-he))
                                        (NP-PRD (N^N^SG drohtin-drohtin))
                                        (BEDI^3^SG uuas-wesan)
                                        (, ,-,)
                                        (CODE <P_86>)
                                        (CODE <R_1209>)
                                        (NP-PRD-PRN (ADJ^N^SG himilisk-himilisk)
                                                    (N^N^SG herro-herro)))
                                (CODE <C>)
                                (CONJP (CONJ endi-endi)
                                       (IP-SUB (NP-SBJ *con*)
                                               (PP (P te-te)
                                                   (NP (N^D^SG helpu-helpa)))
                                               (VBDI^3^SG quam-kuman)
                                               (CODE <R_1210>)
                                               (PP (P an-an)
                                                   (NP (D^A^SG thesan-these)
                                                       (N^A^SG middilgard-middilgard)))
                                               (CODE <C>)
                                               (NP-OB2 (NP-POS (N^G^PL manno-man))
                                                       (N^D^PL barnon-barn))
                                               (, ,-,)
                                               (CODE <R_1211>)
                                               (NP-OB2-PRN (N^D^PL liudeon-liudi))
                                               (PP (P te-te)
                                                   (NP (D^D^SG theson-these)
                                                       (N^D^SG liohte-lioht)))))))
          (. .-.))
  (ID OSHeliandC.663.1207-1211))

( (IP-MAT (CODE <C>)
          (NP-SBJ (PRO^N^3^SG Hie-he))
          (GE+VBDI^3^SG gideda-don)
          (NP-SBJ-1 (PRO^A^3^SG it-it)
                    (CP-THT-PRN *ICH*-2))
          (ADVP-TMP (ADV san-san))
          (PP (P after-aftar)
              (NP (D^D^SG thiu-the)))
          (CODE <R_996>)
          (IP-SMC (NP-SBJ *ICH*-1)
                  (NP-OB2 (N^D^PL mannon-man))
                  (ADJP-PRD (ADJ^A^SG mari-mari)))
          (, ,-,)
          (CODE <C>)
          (CP-THT-PRN-2 (C that-that)
                        (IP-SUB (NP-SBJ (PRO^N^3^PL sia-he))
                                (ADVP-LOC (ADV thar-thar))
                                (NP-OB1 (ADJ^A^SG mahtina-mahtig)
                                        (CODE <R_997>)
                                        (N^A^SG herron-herro))
                                (HVDI^3^PL haƀdun-hebbian)))
          (. :-:))
  (ID OSHeliandC.547.995-997))

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (BEDI^3^PL uuarun-wesan)
          (NP-OB2-RFL (PRO^D^3^PL im-he))
          (ADVP-LOC (ADV thar-thar))
          (GE+VNI^A^PL gifarana-faran)
          (PP (P te-te)
              (NP (D^I^SG thiu-the)
                  (CODE <R_1229>)
                  (CP-THT-PRN (C that-that)
                              (IP-SUB (NP-SBJ (PRO^N^3^PL sia-he))
                                      (NP-OB1 (NP-POS (PRO$^G^1^SG uses-usa)
                                                      (N^G^SG drohtines-drohtin))
                                              (CODE <C>)
                                              (NP-OB1 (N^G^PL dadeo-dad))
                                              (NP-OB1 (CONJ endi-endi)
                                                      (NP (N^G^PL uuordo-word))))
                                      (CODE <R_1230>)
                                      (VB faran-faron)
                                      (MDDI^3^PL uuoldun-willian)))))
          (. ,-,))
  (ID OSHeliandC.668.1228-1230))

In some comparative contexts, the complementizer may be elided, in which case CPX-THT is used.

( (IP-MAT-SPE-0 (CODE <C>)
                (ADVP (ADV So-so))
                (MDPI^3^SG mag-mugan)
                (NP-SBJ (D^N^SG that-the)
                        (CP-THT-PRN-SPE *ICH*-1))
                (PP (P an-an)
                    (NP (PRO$^D^3^SG is-is)
                        (N^D^SG hugi-hugi)))
                (CODE <MS_49a>)
                (NP-PRD (QR^N^SG mera-mero)
                        (CP-CMP-SPE *ICH*-2))
                (CODE <P_122>)
                (CODE <R_1712>)
                (PP (P an-an)
                    (NP (D^D^SG thesaro-these)
                        (N^D^SG middilgard-middilgard)))
                (CODE <C>)
                (NP-OB2 (NP-POS (N^G^PL manno-man))
                        (GE+WADJ^D^SG gihuilicon-gihwilik))
                (CODE <R_1713>)
                (BE uuesan-wesan)
                (PP (P an-an)
                    (NP (D^D^SG thesaro-these)
                        (N^D^SG uueroldi-werold)))
                (CODE <C>)
                (CP-THT-PRN-SPE-1 (C that-that)
                                  (IP-SUB (NP-SBJ (PRO^N^3^SG hie-he))
                                          (ADVP-LOC (ADV hier-her))
                                          (NP-OB1 (N^G^SG uuammes-wam))
                                          (GE+VBPI^3^SG giduot-don)))
                (, ,-,)
                (CODE <R_1714>)
                (CP-CMP-SPE-2 (WADJP-PRD-3 0)
                              (C than-than)
                              (IPX-SUB-SPE=0 (ADJP-PRD *T*-3)
                                             (CPX-THT-SPE (IP-SUB-SPE ...
                (. .-.))
  (ID OSHeliandC.905.1711-1716))

Degree complements (CP-DEG)

Degree complements occur with the adverb so, and are traced to this constituent. In some instances they can be difficult to distinguish from purpose clauses.

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thann-than))
          (ADVP (ADV thoh-thoh))
          (GE+VBDI^3^SG gitruoda-truon)
          (ADVP (ADV so-so)
                (ADV uuel-wel)
                (CP-DEG *ICH*-1))
          (CODE <P_144>)
          (CODE <R_2029>)
          (PP (P an-an)
              (NP (PRO$^D^3^PL iru-iru)
                  (N^D^PL hugisceftion-hugiskaft)))
          (CODE <C>)
          (NP-SBJ (ADJ^N^SG helag-helag)
                  (N^N^SG thiorna-thiorna))
          (, ,-,)
          (CODE <R_2030>)
          (CP-DEG-1 (C that-that)
                    (IP-SUB (NP-OB1 (PRO^G^3^SG is-it))
                            (PP (P after-aftar)
                                (NP (D^D^PL them-the)
                                    (N^D^PL uuordon-word)))
                            (CODE <C>)
                            (NP-SBJ (NP-POS (N^G^SG uualdandes-waldand))
                                    (N^N^SG barn-barn))
                            (, ,-,)
                            (CODE <R_2031>)
                            (NP-SBJ-PRN (NP-POS (VGI^G^PL helandero-helian))
                                        (ADJS^N^SG best-bet))
                            (CODE <C>)
                            (VB helpan-helpan)
                            (MDDS^3^SG uueldi-willian)))
          (. .-.))
  (ID OSHeliandC.1050.2028-2031))

Comparative clauses (CP-CMP)

Comparative clauses are one of the most difficult things to annotate properly. Because of elision, the CPX and IPX ("incomplete") labels get a lot of use here.

The CPX-CMP label is used when the wh-operator and complementizer can be overtly represented in the annotation. This is usually because the category or position of the trace is not straightforwardly recoverable, or else because large-scale absence of structure makes reconstruction impossible.

( (IP-MAT-SPE (CODE <C>)
              (ADVP-TMP (ADV Nu-nu))
              (NP-SBJ (PRO^N^1^SG ik-ik))
              (NP-OB1 (PRO^A^2^PL iu-gi))
              (VB sendean-sendian)
              (MDPI^1^SG scal-skulan)
              (CODE <R_1874>)
              (PP (P after-aftar)
                  (NP (D^D^SG theson-these)
                      (N^D^SG landscepie-landskepi)))
              (CODE <C>)
              (CPX-CMP-SPE (C so-so)
                           (IPX-SUB-SPE (NP-SBJ (N^N^SG lamb-lamb))
                                        (PP (P under-undar)
                                            (NP (N^A^PL uuluos-wulf)))))
              (. :-:))
  (ID OSHeliandC.977.1873-1874))

CP-CMP dominates IP-SUB only if nothing other than the extracted category has been elided. In cases where more structure has been elided, IPX-SUB is used. Equal-sign coindexing (=0 with a corresponding -0) shows where the model for the elided structure is to be understood.

( (IP-MAT-SPE-0 (CODE <C>)
                (BEI^2^PL Uuesat-wesan)
                (NP-OB2-RFL (PRO^D^2^PL iu-gi))
                (ADJP-PRD (ADV so-so)
                          (ADJ^N^PL giuuara-giwar)
                          (CP-CMP-SPE *ICH*-1))
                (PP (P uuidar-withar)
                    (NP (D^I^SG thiu-the)))
                (, ,-,)
                (CODE <R_1883>)
                (PP (P uuid-with)
                    (NP (PRO$^D^3^PL iro-iru)
                        (N^D^PL fecnon-fekan)))
                (ADVP-LOC (ADV thar-thar))
                (CODE <C>)
                (CP-CMP-SPE-1 (WADJP-PRD-2 0)
                              (C so-so)
                              (IPX-SUB-SPE=0 (ADJP-PRD *T*-2)
                                             (NP-SBJ (MAN^N^3^SG man-man))
                                             (PP (P uuidar-withar)
                                                 (NP (N^D^PL fiondon-fiund)))
                                             (MDPI^3^SG scal-skulan)))
                (. .-.))
  (ID OSHeliandC.981.1882-1883))

The most usual types of CP(X)-CMP involve a head so, with an accompanying so or sulik X constituent (as in the above), than along with comparative morphology, or than X the (as in the below).

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (VBDI^3^PL quathun-kwethan)
          (CP-THT (C that-that)
                  (IP-SUB-0 (NP-OB2 (PRO^D^3^SG im-he))
                            (NP-SBJ (N^N^SG uualdand-waldand)
                                    (ADJ^N^SG self-self))
                            (, ,-,)
                            (CODE <R_2337>)
                            (NP-SBJ-PRN (NPR^N^SG god-god)
                                        (ADJ^N^SG alomahtig-alomahtig))
                            (CODE <C>)
                            (VN forgeƀan-fargevan)
                            (HVDS^3^SG habdi-hebbian)
                            (CODE <R_2338>)
                            (NP-OB1 (QR^A^PL merun-mero)
                                    (N^A^PL mahti-maht)
                                    (CODE <C>)
                                    (CP-CMP (WNP-OB1-1 0)
                                            (C than-than)
                                            (IPX-SUB=0 (NP-OB1 *T*-1)
                                                       (ADVP (ADV elcor-elkor))
                                                       (NP-OB2 (Q^D^SG enigon-enig)
                                                               (NP-POS (N^G^SG mannes-man))
                                                               (N^D^SG sunie-sunu)))))))
          (, ,-,)
          (CODE <R_2339>)
          (NP-OB1-PRN (NP-OB1-PRN (N^A^SG craft-kraft))
                      (CONJP (CONJ endi-endi)
                             (NP-OB1-PRN (N^A^PL cunsti-kunst))))
          (. ;-;))
  (ID OSHeliandC.1243.2336-2339))

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (GE+VBPI^3^SG gihugit-huggian)
          (CP-QUE (WNP-OB1-1 (WPRO^A huat-hwe)
                             (NP-POS *ICH*-2))
                  (C 0)
                  (IP-SUB (NP-OB1 *T*-1)
                          (NP-SBJ (PRO^N^3^SG hie-he)
                                  (ADJ^N^SG selƀo-self))
                          (GE+VBDI^3^SG gifrumida-frummian)
                          (CODE <R_3497>)
                          (NP-POS-2 (ADJ^G^SG grimmes-grim))
                          (ADVP-TMP (ADV thann-than)
                                    (ADV lang-lang)
                                    (CP-CMP (WADVP-TMP-3 0)
                                            (C the-the)
                                            (IP-SUB (ADVP-TMP *T*-3)
                                                    (NP-SBJ (PRO^N^3^SG hie-he))
                                                    (MDDI^3^SG muosta-motan)
                                                    (CODE <MS_98b>)
                                                    (NP-OB1 (PRO$^G^3^SG is-is)
                                                            (N^G^SG iuguthi-juguth))
                                                    (VB niotan-niotan))))))
          (. ;-;))
  (ID OSHeliandC.2000.3496-3497))

For instances where a subordinator may be elided, see the section on that-clauses.

Direct and indirect questions (CP-QUE)

Both direct and indirect questions are CP-QUE dominating an IP-SUB. The difference in annotation is that indirect questions contain a null C, whereas direct questions do not. When a wh-phrase is present, it is traced into the subordinate clause in which it belongs.

( (CP-QUE-SPE (CODE <C>)
              (' '-')
              (WADVP-1 (WADV Huo-hwo))
              (IP-SUB-SPE (ADVP *T*-1)
                          (MDPI^3^SG mag-mugan)
                          (NP-SBJ (D^N^SG that-the))
                          (GE+RD giuuerthan-werthan)
                          (ADVP (ADV so-so))
                          (' '-')
                          (IP-MAT-PRN (VBDI^3^SG quat$-kwethan)
                                      (NP-SBJ (PRO^N^3^SG $hie-he)))
                          (, ,-,)
                          (CODE <R_142>)
                          (' '-')
                          (ADVP-TMP (ADV after-aftar))
                          (PP (P an-an)
                              (NP (N^D^SG aldre-aldar))))
              (. ?-?))
  (ID OSHeliandC.65.141-142))

( (IP-MAT (CODE <R_210>)
          (NP-SBJ *con*)
          (VBDI^3^SG fragode-fragon)
          (ADVP (ADV niudlico-niudliko))
          (CODE <C>)
          (CP-QUE (WNP-PRD-1 (WPRO^N huat-hwe))
                  (C 0)
                  (IP-SUB (NP-PRD *T*-1)
                          (NP-SBJ (PRO$^N^3^SG is-is)
                                  (N^N^SG namo-namo))
                          (MDDS^3^SG scoldi-skulan)
                          (CODE <R_211>)
                          (BE uuesan-wesan)
                          (PP (P an-an)
                              (NP (D^D^SG thesaro-these)
                                  (N^D^SG uueroldi-werold)))))
          (. :-:))
  (ID OSHeliandC.105.210-211))

Yes-no questions are annotated as CP-QUE immediately dominating an IP-SUB but no C or wh-phrase.

( (CP-QUE-SPE (CODE <R_3813>)
              (IP-SUB-SPE (BEPI^3^SG is-wesan)
                          (NP-SBJ (PRO^N^3^SG it-it))
                          (NP-PRD (N^N^SG reht-reht))))
  (ID OSHeliandC.2196.3813))

When it introduces a yes-no question, hwethar is labelled WQ. These questions follow the same rules as others: C is present in indirect questions and absent in direct questions.

( (CP-QUE-SPE (CODE <C>)
              (' '-')
              (WQ Hueder-hwethar)
              (IP-SUB-SPE (VBPI^2^PL ledeat-ledian)
                          (NP-SBJ (PRO^N^2^PL gi-gi))
                          (NP-OB1 (VNI^A^SG uundan-windan)
                                  (N^A^SG gold-gold))
                          (CODE <R_555>)
                          (PP (P te-te)
                              (NP (N^D^SG geƀu-geva)))
                          (NP-OB2 (WADJ^D^SG huilicon-hwilik)
                                  (NP-POS (N^G^PL gumono-gumo))))
              (. ?-?))
  (ID OSHeliandC.291.554-555))

( (IP-MAT-SPE (CODE <C>)
              (NEG ne-ne)
              (VBI^2^PL ruokeat-rokian)
              (CP-QUE-SPE (WQ hueđer-hwethar)
                          (C 0)
                          (IP-SUB-SPE (NP-SBJ (PRO^N^2^PL gi-gi))
                                      (NP-OB1 (NP-POS (D^G^SG thes-the))
                                              (NP-OB1 (Q^A^SG enigan-enig)
                                                      (N^A^SG thanc-thank))
                                              (CONJP *ICH*-1))
                                      (VBPS^2^PL antfahan-andfahan)
                                      (CODE <R_1542>)
                                      (CONJP-1 (CONJ eftha-eftha)
                                               (NP-OB1 (N^A^SG lon-lon)))
                                      (PP (P an-an)
                                          (NP (D^D^SG thesaro-these)
                                              (ADJ^D^SG lehnun-lehni)
                                              (N^D^SG uueroldi-werold)))))
              (. ,-,))
  (ID OSHeliandC.818.1541-1542))

Hwethar may also serve as a wh-word meaning "which of two". In this role, it is tagged as WADJ.

( (IP-MAT (CODE <R_5409>)
          (ADVP-TMP (ADV Thuo-tho))
          (VBDI^3^SG bigan-biginnan)
          (NP-SBJ (D^N^SG thie-the)
                  (N^N^SG heritogo-heritogo))
          (CODE <C>)
          (NP-OB1 (D^A^SG thia-the)
                  (N^A^SG heri-heri)
                  (NP-POS (NPR^G^PL Iuđeono-Judeo)))
          (, ,-,)
          (CODE <R_5410>)
          (NP-OB1-PRN (D^A^SG that-the)
                      (N^A^SG folc-folk))
          (VB fragoian-fragon)
          (, ,-,)
          (CODE <C>)
          (CP-ADV (ADVP-LOC (ADV thar-thar))
                  (C 0)
                  (IP-SUB (NP-SBJ (PRO^N^3^PL sia-he))
                          (PP (NP (PRO^D^3^SG im-he))
                              (P fora-for))
                          (VBDI^3^PL stuodun-standan)))
          (, ,-,)
          (CODE <R_5411>)
          (CP-QUE (WNP-OB1-1 (WADJ^A^SG huederon-hwethar)
                             (NP-POS *ICH*-2))
                  (C 0)
                  (IP-SUB (IP-SUB-0 (NP-OB1 *T*-1)
                                    (NP-SBJ (PRO^N^3^PL sia-he))
                                    (NP-POS-2 (D^G^PL thero-the)
                                              (NUM^G^PL tueio-twene)
                                              (CP-REL-SPE *ICH*-3))
                                    (CODE <C>)
                                    (VB tuomian-tomian)
                                    (MDDS^3^PL uueldin-willian))
                          (, ,-,)
                          (CODE <R_5412>)
                          (CONJP (IPX-SUB=0 (NP-OB1 (N^G^SG ferahes-ferah))
                                            (VB biddian-biddian)))
                          (, :-:)
                           ...
  (ID OSHeliandC.3203.5409-5413))

When hwat has an adverbial role ("why" or "how"), it is treated as accusative and part of a WNP-ADT. See also Interjection Phrases.

( (CP-QUE-SPE (CODE <C>)
              (' '-')
              (WNP-ADT-1 (WPRO^A Huat-hwe))
              (IP-SUB-SPE (NP-ADT *T*-1)
                          (MDPI^2^PL uuelliat-willian)
                          (NP-SBJ (PRO^N^2^PL gi-gi))
                          (ADJP-2 (PRO$^G^1^PL minero-min))
                          (ADVP-LOC (ADV hier-her))
                          (' '-')
                          (IP-MAT-PRN (VBDI^3^SG quat$-kwethan)
                                      (NP-SBJ (PRO^N^3^SG $he-he)))
                          (, ,-,)
                          (CODE <R_3574>)
                          (' '-')
                          (NP-OB1 (ADJP *ICH*-2)
                                  (N^G^PL helpono-helpa))
                          (VB biddean-biddian))
              (. ?-?)
              (' '-'))
  (ID OSHeliandC.2061.3573-3574))

In some questions, wh-doubling may be found. In these cases, the matrix clause wh-element is treated as a dummy and coindexed to the lower, "real" version.

( (CP-QUE-SPE (CODE <C>)
              (WNP-1 (WPRO^N huat-hwe))
              (IP-SUB-SPE (VBPI^3^PL quethat-kwethan)
                          (NP-SBJ (D^N^PL thesa-these)
                                  (NP-POS (NPR^G^PL Iudeo-Judeo))
                                  (N^N^PL liudi-liudi))
                          (, ,-,)
                          (CODE <R_3040>)
                          (NP-SBJ-PRN (ADJ^N^SG mari-mari)
                                      (N^N^SG meginthioda-meginthioda))
                          (, ,-,)
                          (CODE <C>)
                          (CP-QUE-SPE (WNP-PRD-1 (WPRO^N huat-hwe)
                                                 (NP-POS *ICH*-2))
                                      (C 0)
                                      (IP-SUB-SPE (NP-PRD *T*-1)
                                                  (NP-SBJ (PRO^N^1^SG ik-ik))
                                                  (NP-POS-2 (N^G^PL manno-man))
                                                  (BEPS^1^SG si-wesan))))
              (. ?-?)
              (' '-'))
  (ID OSHeliandC.1683.3039-3040))

Relative clauses (CP-REL)

All relative clauses have a wh-position and a complementizer position. It is rare for neither to be overtly filled. Since thie and the are ambiguous between invariant C and demonstrative, the preference is to annotate it as a demonstrative if that is formally possible.

( (IP-MAT-SPE (CODE <R_1683>)
              (NP-SBJ (N^N^PL liudi-liudi)
                      (CP-REL-SPE *ICH*-1))
              (BEPI^3^PL sind-wesan)
              (NP-OB2 (PRO^D^3^SG im-he))
              (ADJP-PRD (ADJR^N^PL leoƀrun-liof))
              (ADJP-ADT (ADJ^I^SG mikilo-mikil))
              (CODE <C>)
              (CP-REL-SPE-1 (WNP-OB1-2 (D^A^PL thie-the))
                            (C 0)
                            (IP-SUB-SPE (NP-OB1 *T*-2)
                                        (NP-SBJ (PRO^N^3^SG hie-he))
                                        (NP-OB2-RFL (PRO^D^3^SG im-he))
                                        (PP (P an-an)
                                            (NP (D^D^SG them-the)
                                                (N^D^SG lande-land)))
                                        (GE+VBDI^3^SG giuuarahta-wirkian)
                                        (CODE <R_1684>)
                                        (NP-SBJ-PRN (N^N^SG uualdand-waldand))
                                        (PP (P an-an)
                                            (NP (N^D^SG uuilleon-willio)
                                                (PRO$^D^3^SG sinan-sin)))))
              (. .-.))
  (ID OSHeliandC.888.1683-1684))

( (IP-MAT (CODE <C>)
          (CONJ endi-endi)
          (NP-SBJ *con*)
          (NP-OB2 (D^D^PL them-the)
                  (N^D^PL mannon-man))
          (VBDI^3^SG sagda-seggian)
          (CODE <R_1296>)
          (NP-ADT (ADJ^D^PL spahon-spah)
                  (N^D^PL uuordon-word))
          (CODE <MS_38a>)
          (CODE <C>)
          (NP-OB2-PRN (D^D^PL them-the)
                      (CP-REL (WNP-OB1-1 0)
                              (C the-the)
                              (IP-SUB (NP-OB1 *T*-1)
                                      (NP-SBJ (PRO^N^3^SG hie-he))
                                      (PP (P te-te)
                                          (NP (D^D^SG thero-the)
                                              (N^D^SG spraco-spraka)))
                                      (ADVP-DIR (ADV tharod-tharod))
                                      (CODE <P_92>C>)
                                      (CODE <R_1297>C>)
                                      (NP-SBJ-PRN (NPR^N^SG Crist-Krist)
                                                  (N^N^SG alouualdo-alowaldo))
                                      (CODE <C>)
                                      (GE+VN gicoran-kiosan)
                                      (HVDI^3^SG haƀda-hebbian))))
           ...
  (ID OSHeliandC.710.1295-1299))

There is no evidence for that as an invariant complementizer in the HeliPaD (unlike e.g. in late OE).

Temporal and locative relative clauses behave like other relative clauses. There is usually an overt adverbial thar or than, but in some cases there may be only a complementizer the.

(ADVP-DIR (ADV thanan-thanan)
          (CP-REL-SPE (WADVP-1 0)
                      (C the-the)
                      (IP-SUB-SPE (ADVP *T*-1)
                                  (NP-SBJ (MAN^N man-man))
                                  (NP-OB1 (PRO^A^2^PL iu-gi))
                                  (VB antfahan-andfahan)
                                  (NEG ni-ne)
                                  (MDPI^3^SG uuili-willian))))

(OSHeliandC.1014.1946-1947)

(PP (P an-an)
    (NP (D^A^SG that-the)
        (ADJ^A^SG hoha-hoh)
        (N^A^SG hus-hus)
        (, ,-,)
        (CODE <C>)
        (CP-REL (WADVP-LOC-1 (ADV thar-thar))
                (C 0)
                (IP-SUB (ADVP-LOC *T*-1)
                        (NP-SBJ (D^N^SG thiu-the)
                                (N^N^SG heri-heri))
                        (VBDI^3^SG dranc-drinkan)
                        (, ,-,)
                        (CODE <R_2002>)
                        (NP-SBJ-PRN (D^N^PL thia-the)
                                    (NPR^N^PL Iudeon-Judeo))
                        (PP (P an-an)
                            (NP (D^D^SG them-the)
                                (N^D^SG gestseli-gastseli)))))))

(OSHeliandC.1030.1999-2002)

( (IP-MAT-SPE (CODE <R_2496>C>)
              (ADVP-TMP (ADV Than-than)
                        (CP-REL-SPE *ICH*-1))
              (VBPI^3^SG biginnit-biginnan)
              (NP-SBJ (PRO^D^3^SG im-he))
              (VB thunkean-thunkian)
              (, ,-,)
              (CODE <C>)
              (CP-REL-SPE-1 (WADVP-TMP-2 (ADV than-than))
                            (C 0)
                            (IP-SUB-SPE (IP-SUB-SPE (ADVP-TMP *T*-2)
                                                    (NP-SBJ (PRO^N^3^SG hie-he))
                                                    (PP (P under-undar)
                                                        (NP (D^D^SG thero-the)
                                                            (N^D^SG thieda-thiod)))
                                                    (VBPI^3^SG steđ-stan))
                                         ...
  (ID OSHeliandC.1336.2496-2498))

See also Case attraction.

Free relative clauses (CP-FRL)

Free relatives (CP-FRL) are treated essentially like nominals in terms of their external syntax: they take nominal extended labels and thus may be subjects, locatives, predicates, etc. In case of ambiguity between questions and free relatives, the question parse is chosen. Relative clauses with no possible antecedent that are headed by a demonstrative are treated as free relatives, as in the first example below.

( (IP-MAT-SPE (CODE <R_2582>C>)
              (NP-SBJ (PRO^N^1^SG ik-ik)
                      (ADJ^N^SG selƀo-self))
              (BEPI^1^SG biun-wesan)
              (CP-FRL-PRD-SPE (WNP-SBJ-1 (D^N^SG that-the))
                              (C 0)
                              (IP-SUB-SPE (NP-SBJ *T*-1)
                                          (ADVP-LOC (ADV thar-thar))
                                          (VBPI^1^SG saiu-saian)))
              (. ,-,))
  (ID OSHeliandC.1388.2582))

( (IP-MAT (CODE <C>)
          (NP-SBJ (D^N^PL thia-the)
                  (N^N^PL liudi-liudi)
                  (QP *ICH*-1))
          (RDDI^3^PL uurdun-werthan)
          (QP-1 (Q^N^PL alla-al))
          (CODE <R_2862>)
          (ADJP-PRD (ADJ^N^PL sada-sad))
          (NP-SBJ-PRN (ADJ^N^SG salig-salig)
                      (N^N^SG folc-folk))
          (, ,-,)
          (CODE <C>)
          (CP-FRL-SBJ-PRN (WNP-SBJ-2 (ADV so-so)
                                     (WPRO^N huat-hwe))
                          (C so-so)
                          (IP-SUB (NP-SBJ *T*-2)
                                  (ADVP-LOC (ADV thar-thar))
                                  (GE+VN gisamnod-samnon)
                                  (BEDI^3^SG uuas-wesan)
                                  (CODE <R_2863>)
                                  (PP (P fon-fan)
                                      (NP (Q^D^PL allon-al)
                                          (ADJ^D^PL uuidon-wid)
                                          (N^D^PL uuegon-weg)))))
          (. .-.))
  (ID OSHeliandC.1561.2861-2863))

( (IP-MAT (CODE <C>)
          (NP-SBJ (D^N^SG That-the)
                  (N^N^SG uuerod-werod)
                  (ADJ^N^SG ođer-othar))
          (VBDI^3^SG bed-bidan)
          (CODE <R_104>)
          (PP (P umbi-umbi)
              (NP (D^A^SG thena-the)
                  (N^A^SG alah-alah))
              (ADV utan-utan))
          (CODE <C>)
          (NP-SBJ-PRN (NP-POS (NPR^G^PL Hebreo-Ebreo))
                      (N^N^PL liudi-liudi))
          (, ,-,)
          (CODE <R_105>)
          (CP-FRL-TMP (WADVP-TMP-1 (WADV huan-hwan)
                                   (ADV er-er))
                      (C 0)
                      (IP-SUB (ADVP-TMP *T*-1)
                              (D^N^SG thie-the)
                              (NP-SBJ (ADJ^N^SG fruodo-frod)
                                      (N^N^SG man-man))
                              (CODE <C>)
                              (GE+VN gifrumid-frummian)
                              (HVDS^3^SG habdi-hebbian)
                              (CODE <P_12>)
                              (CODE <R_106>)
                              (NP-OB1 (NP-POS (N^G^SG uualdandes-waldand))
                                      (N^A^SG uuilleon-willio))))
          (. .-.))
  (ID OSHeliandC.41.103-106))

Free relatives with wh-elements are mostly of the so wh- so type. so wh- is in the wh-position, and the second so is in C.

( (IP-MAT (CODE <C>)
          (NP-SBJ (D^N^SG That-the)
                  (N^N^SG fri-fri))
          (NP-OB1 (Q^A^SG all-al))
          (VBDI^3^SG biheld-bihaldan)
          (CODE <R_436>)
          (PP (P an-an)
              (NP (PRO$^D^3^PL iro-iru)
                  (N^D^PL hugisceftion-hugiskaft)))
          (CODE <C>)
          (NP-SBJ-PRN (ADJ^N^SG helag-helag)
                      (N^N^SG thiorno-thiorna))
          (, ,-,)
          (CODE <R_437>)
          (NP-SBJ-PRN (D^N^SG thiu-the)
                      (N^N^SG magat-magath))
          (CODE <MS_16a>)
          (PP (P an-an)
              (NP (PRO$^D^3^SG iru-iru)
                  (N^D^SG muode-mod)))
          (, ,-,)
          (CODE <C>)
          (CP-FRL-OB1-PRN (WNP-OB1-1 (ADV so-so)
                                     (WPRO^A huat-hwe))
                          (C so-so)
                          (IP-SUB (NP-SBJ (PRO^N^3^SG siu-siu))
                                  (GE+VBDI^3^SG gihorda-horian)
                                  (IP-INF (NP-OB1 *T*-1)
                                          (NP-SBJ (D^A^PL thia-the)
                                                  (N^A^PL man-man))
                                          (VB sprecan-sprekan))))
          (. .-.))
  (ID OSHeliandC.236.435-437))

Infinitival clauses (IP-INF, IP-INF-NCO)

Infinitival clauses are normally the complements of verbs in the HeliPaD. Only with ECM verbs is a subject present in an infinitival clause.

(CP-FRL-OB1-PRN (WNP-OB1-1 (ADV so-so)
                           (WPRO^A huat-hwe))
                (C so-so)
                (IP-SUB (NP-SBJ (PRO^N^3^SG siu-siu))
                        (GE+VBDI^3^SG gihorda-horian)
                        (IP-INF (NP-OB1 *T*-1)
                                (NP-SBJ (D^A^PL thia-the)
                                        (N^A^PL man-man))
                                (VB sprecan-sprekan))))

(OSHeliandC.236.435-437)

( (IP-MAT-SPE (CODE <C>)
              (NP-SBJ (D^N^SG That-the)
                      (IP-INF-PRN-SPE *ICH*-1))
              (RDPI^3^SG uuirđit-werthan)
              (NP-OB2 (PRO^D^2^SG thi-thu))
              (NP-PRD (N^N^SG uuerk-werk)
                      (ADJ^N^SG mikil-mikil))
              (, ,-,)
              (CODE <P_38>)
              (CODE <R_502>)
              (IP-INF-PRN-SPE-1 (NP-OB1 (N^A^SG thrim-thrim))
                                (TO te-te)
                                (GE+VB githolonne-tholon))
              (. .-.)
              (' '-'))
  (ID OSHeliandC.264.501-502))

Exceptional case marking (ECM) verbs, also known as Accusative and infinitive (AcI) verbs, are lexically restricted to the following:

  • Perception verbs:
    • findan
    • horian
    • sehan
  • Verbs of permission and commanding:
    • biddian
    • hetan
    • latan
  • witan

They involve a bare infinitive and always include a subject.

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (GE+VBDI^3^PL gisahun-sehan)
          (IP-INF (NP-SBJ (PRO$^A^3^SG iro-iru)
                          (N^A^SG boggebon-boggevo))
                  (CODE <R_2739>)
                  (BE uuesan-wesan)
                  (PP (P an-an)
                      (NP (N^D^PL uunneon-wunnia))))
          (. .-.))
  (ID OSHeliandC.1476.2738-2739))

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (VBDI^3^SG hiet-hetan)
          (ADVP-TMP (ADV thuo-tho))
          (IP-INF (NP-SBJ (PRO$^A^3^SG is-is)
                          (N^A^SG uuapanberan-wapanberand))
                  (CODE <R_2780>)
                  (VB gangan-gangan)
                  (PP (P fan-fan)
                      (NP (D^D^SG them-the)
                          (N^D^SG gastseli-gastseli)))))
  (ID OSHeliandC.1503.2779-2780))

Arbitrary subjects of such clauses are given as the empty category *arb*.

In tough-movement constructions, the subject is traced to the infinitive, which is contained within the ADJP.

( (IP-MAT-SPE (CODE <C>)
              (NP-SBJ-1 *pro*)
              (ADJP-PRD (ADJ^N^SG othi-othi)
                        (IP-INF-SPE *ICH*-2))
              (BEPI^3^SG ist-wesan)
              (ADVP-DIR (ADV tharod-tharod))
              (IP-INF-SPE-2 (NP-OB1 *-1)
                            (TO te-te)
                            (VB faranne-faran))
              (CODE <R_1780>)
              (NP-OB2 (N^D^PL eldibarnon-eldibarn))
              (, ,-,)
              (CODE <C>)
              (CP-ADV-SPE (ADVP (ADV thoh-thoh))
                          (C 0)
                          (IP-SUB-SPE (NP-SBJ (PRO^N^3^SG it-it))
                                      (NP-OB2 (PRO^D^3^PL im-he))
                                      (PP (P at-at)
                                          (NP (D^D^SG them-the)
                                              (N^D^SG endie-endi)))
                                      (NEG ni-ne)
                                      (VBPS^3^SG dugi-dugan)))
              (. .-.))
  (ID OSHeliandC.935.1779-1780))

Non-complement infinitives (IP-INF-NCO) are those which are not the complements of verbs, nouns, or adjectives, and not the subject of the clause. Most usually, these co-occur with motion verbs.

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (VBDI^3^SG geng-gangan)
          (IP-INF-NCO (PP (P uuiđ-with)
                          (NP (PRO$^A^3^SG iro-iru)
                              (N^A^SG kind-kind)))
                      (VB sprekan-sprekan)
                      (CODE <R_2019>)
                      (PP (P uuiđ-with)
                          (NP (PRO$^A^3^SG iro-iru)
                              (N^A^SG suno-sunu)
                              (ADJ^A^SG selƀan-self))))
          (. ,-,))
  (ID OSHeliandC.1042.2018-2019))

Small clauses (IP-SMC)

Small clauses consist of a subject and a predicate. They are the only type of IP that does not contain a verbal element by default. The subjects of small clauses are almost always in accusative case.

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (HVDI^3^SG haƀda-hebbian)
          (ADVP-TMP (ADV eft-eft))
          (IP-SMC (NP-SBJ (PRO$^A^3^SG is-is)
                          (N^A^SG uuord-word))
                  (ADJP-PRD (ADJ^A^SG garo-garu)))
          (CODE <R_2024>)
          (NP-SBJ (ADJ^N^SG mahtig-mahtig)
                  (N^N^SG barn-barn)
                  (NP-POS (NPR^G^SG godes-god))))
  (ID OSHeliandC.1045.2023-2024))

( (IP-MAT-SPE (CODE <C>)
              (ADVP-TMP (ADV than-than))
              (VBPI^2^SG findis-findan)
              (NP-SBJ (PRO^N^2^SG thu-thu))
              (IP-SMC-SPE (ADJP-PRD (ADJ^A^SG gisundan-gisund))
                          (PP (P at-at)
                              (NP (N^D^SG hus-hus)))
                          (CODE <R_2151>)
                          (NP-SBJ (ADJ^A^SG maguiungan-magujung)
                                  (N^A^SG man-man)))
              (. :-:))
  (ID OSHeliandC.1105.2150-2151))

If the subject and predicate are not adjacent, the small clause is taken to be where the predicate is, and the subject is traced to that.

( (IP-MAT (CODE <C>)
          (NP-SBJ-1 (D^A^SG thena-the))
          (GE+VBDI^3^SG gideda-don)
          (NP-SBJ (D^N^SG thie-the)
                  (VGI^N^SG heland-helian)
                  (ADJ^N^SG self-self))
          (, ,-,)
          (CODE <R_2355>)
          (NP-SBJ-PRN (NPR^N^SG Crist-Krist))
          (PP (P thuru-thurh)
              (NP (PRO$^A^3^SG is-is)
                  (N^A^SG craft-kraft)
                  (ADJ^A^SG mikil-mikil)))
          (CODE <C>)
          (IP-SMC (NP-SBJ *ICH*-1)
                  (ADJP-PRD (ADJ^A^SG quican-kwik)))
          (PP (P aftar-aftar)
              (NP (N^D^SG dođe-doth)))
          (. ,-,))
  (ID OSHeliandC.1252.2354-2355))

Small clauses often occur with verbs of naming, with an accusative subject as usual.


Nominals

Noun Phrases (NP)

The head of a noun phrase is a nominal element, usually a noun (common or proper) or a pronoun. There is no explicit marking of any assumed null noun or pronoun in NPs.

(NP-SBJ (PRO^N^3^PL sea-he))

(NP-POS (NPR^G^SG Cristes-Krist))

(NP-OB1 (Q^A^SG all-al))

(OSHeliandC.9.32-42)

The word man, with its special tag MAN, heads an NP.

(NP-SBJ (MAN^N^3^SG man-man))

(OSHeliandC.42.106-113)

Noun-noun compounds are tagged as a single N when written together, as in the YCOE, but as separate nouns within the same NP when written separately, as in the case of proper names.

(NP-SBJ (NPR^N^SG Simon-Simon)
        (NPR^N^SG Petrus-Petrus))

(OSHeliandC.1691.3053-3054)

Titles are never labelled as separate appositive phrases in the HeliPaD, whether modified or not:

(NP-SBJ (NPR^N^SG Iohannes-Johannes)
        (CODE <C>)
        (N^N^SG iungro-jungaro)
        (NP-POS (NPR^G^SG Cristes-Krist)))

(OSHeliandC.3321.5610-5611)

(NP-SBJ (NPR^N^SG Iohannes-Johannes)
        (D^N^SG thie-the)
        (ADJ^N^SG guodo-god))

(OSHeliandC.1460.2710-2711)

See the section on Heads, modifiers and complements for information on NP-internal modification, and the section on Appositives and parentheticals for information on apposition.

NP-COM is not used, and nouns are assumed not to have nominal complements. NP-POS or clause-level adjuncts take over much of the work of this label in the HeliPaD.

Possessive pronouns are labelled PRO$, and like other inflected nominal categories in the HeliPaD are always given attributes, even in the case of the third person possessives which do not inflect. These items are also found as genitive pronouns in some cases, essentially when they are used alone in a phrase, or modified only by al, en or self.

( (IP-MAT (CODE <C>)
          (NP-SBJ (NUM^N^SG En-en)
                  (NP-POS *ICH*-1))
          (BEDI^3^SG uuas-wesan)
          (NP-POS-1 (PRO^G^3^PL iro-he))
          (ADVP-TMP (ADV thuo-tho))
          (ADVP-TMP (ADV noh-noh))
          (ADVP-TMP (ADV than-than))
          (CODE <R_47>)
          (NP-OB2 (NP-POS (N^G^PL firio-firihos))
                  (N^D^PL barnun-barn))
          (ADVP-LOC (ADV biforan-biforan))
          (. ,-,))
  (ID OSHeliandC.11.46-47))

( (IP-MAT-SPE (CODE <R_1632>)
              (CONJ ac-ak)
              (VBI^2^PL mithat-mithan)
              (NP-OB1 (PRO^G^3^SG is-it))
              (PP (P for-for)
                  (NP (ADJ^D^PL ođron-othar)
                      (N^D^PL mannon-man)))
              (. :-:))
  (ID OSHeliandC.864.1632))

( (IP-MAT-SPE (CODE <C>)
              (CONJ endi-endi)
              (NP-SBJ (D^N^SG thie-the)
                      (PRO$^N^2^SG iuuua-iuwa)
                      (N^N^SG fritho-frithu))
              (VBPI^3^SG huiriƀit-hwervan)
              (CODE <R_1944>)
              (ADVP-TMP (ADV eft-eft))
              (PP (P an-an)
                  (NP (NP-POS (PRO^G^2^PL iuuuer-gi)
                              (ADJ^G^PL selbaro-self))
                      (N^A^SG sith-sith)))
              (. ,-,))
  (ID OSHeliandC.1011.1943-1944))

NP-internal agreement is forced wherever possible, thus allowing ambiguous elements to be tagged (the HeliPaD's approach to attributes is "maximalist"). The main exception to this is with instrumental elements, which often co-occur with formally dative elements. This is not treated as a case clash, and instrumental is only preferred where unambiguous.

Noun phrase extended labels

In general, nominal extended labels for arguments in the HeliPaD work like they do in the PPCME2 and IcePaHC and not as they do in the YCOE. Hence, not every paragraph in this section will be flagged in green.

At clause level, NPs always either bear a functional label or an index to show where they have been extracted from. The possibilities for nominal extended labels are:

  • Arguments:
    • -SBJ: subject
    • -OB1: direct or only object
    • -OB2: indirect or second object
    • -PRD: predicate (also for ADJPs and PTPs)
  • Non-arguments:
    • -ADT: adjunct (also for ADJPs and PTPs)
    • -LOC: locative
    • -TMP: temporal
    • -VOC: vocative
    • -POS: possessor (only one immediately dominated by each NP)
  • -RFL: reflexive (not mutually exclusive with above)

These replace the YCOE's phrasal case labels.

The two broad classes of NPs in the corpus are arguments and adjuncts. Temporal (-TMP) and locative (-LOC) NPs are a subclass of adjuncts (-ADT), as are vocatives (-VOC) and left-dislocations (-LFD). Possessors (-POS) only occur NP-internally or coindexed to an NP-internal position.

Any NP that contains a pronoun coreferential with the subject of the clause is labelled -RFL (reflexive), no matter where it occurs in the clause.

Subjects

In general, nominal extended labels for arguments in the HeliPaD work like they do in the PPCME2 and IcePaHC and not as they do in the YCOE. Hence, not every paragraph in this section will be flagged in green.

All complete finite IPs must have a subject (-SBJ), other than the following:

  • Clauses with wita (equivalent to uton in the YCOE)
  • Imperative clauses (though they may have one)
  • The outer wrapper of conjoined clauses
( (IP-MAT-SPE (CODE <C>)
              (UTP Uuita-wita)
              (VB kiesan-kiosan)
              (NP-OB2 (PRO^D^3^SG im-he))
              (NP-OB1 (ADJ^A^SG ođerna-othar)
                      (CODE <R_224>)
                      (ADJ^A^SG niudsamana-niudsam)
                      (N^A^SG namon-namo))
              (. :-:))
  (ID OSHeliandC.113.223-224))

( (IP-MAT-SPE (CODE <C>)
              (VBI^2^SG Sagi-seggian)
              (NP-OB2 (PRO^D^1^PL us-we))
              (, ,-,)
              (CP-QUE-SPE (WPP-1 (P under-undar)
                                 (WNP (WADJ^D^SG huilicon-hwilik)
                                      (NP-POS *ICH*-2)))
                          (C 0)
                          (IP-SUB-SPE (PP *T*-1)
                                      (NP-SBJ (PRO^N^3^SG hie-he))
                                      (BEPS^3^SG si-wesan)
                                      (NP-POS-2 (D^G^PL thesaro-these)
                                                (N^G^PL cunnio-kunni))
                                      (VN afuodid-afodian)))
              (. .-.)
              (' '-'))
  (ID OSHeliandC.322.605))

( (IP-MAT-SPE (CODE <R_263>)
              (NEG ni-ne)
              (VBI^2^SG forohti-forhtian)
              (NP-SBJ (PRO^N^2^SG thu-thu))
              (NP-OB2 (PRO$^D^2^SG thinon-thin)
                      (N^D^SG ferahe-ferah))
              (. :-:))
  (ID OSHeliandC.140.263))

( (IP-MAT (CODE <R_994>)
          (IP-MAT-0 (NP-OB1 (D^A^SG That-the))
                    (MDDI^3^SG muosta-motan)
                    (NP-SBJ (NPR^N^SG Iohannes-Johannes))
                    (ADVP-TMP (ADV thuo-tho))
                    (, ,-,)
                    (CODE <C>)
                    (CP-ADV (ADVP (ADV all-al))
                            (C so-so)
                            (IP-SUB (NP-OB1 (PRO^A^3^SG it-it))
                                    (NP-SBJ (NPR^N^SG guod-god))
                                    (MDDI^3^SG uuelda-willian)))
                    (, ,-,)
                    (CODE <R_995>)
                    (GE+VB gisehan-sehan))
          (CONJP (CONJ endi-endi)
                 (IPX-MAT=0 (GE+VB gihorean-horian)))
          (. .-.))
  (ID OSHeliandC.546.994-995))

If no overt subject is present in such clauses, an empty subject is added.

Subjects are almost always nominative, though they are not required to be. A small selection of verbs may have non-nominative subjects for ease of retrieval. (This selection of verbs is likely to change in future editions of the corpus.)

  • thunkian (dat.)
  • likon (dat.)
  • lustian (acc.)
  • thurstian (acc.)
  • risan (dat.) (variable)
  • belgan (acc.) (variable)
  • spanan (acc.) (variable)
  • hreuwan (acc.) (variable)

In non-finite IPs there is no requirement to have a subject, on the whole. Overt subjects are marked -SBJ. ECM verbs have a subject; if it is absent, *arb* is used. Small clauses (-SMC) are required to have a subject and predicate.

( (IP-MAT (CODE <C>)
          (NP-SBJ-X *exp*)
          (BEDI^3^SG uuas-wesan)
          (NP-OB2 (PRO^D^3^PL im-he))
          (NP-PRD (NP-POS (PRO$^G^3^SG is-is)
                          (N^G^SG huldi-huldi))
                  (NP-POS-PRN (N^G^PL helpono-helpa))
                  (N^N^SG tharf-tharf))
          (CODE <R_1188>)
          (IP-INF-X (TO te-te)
                    (GE+VB githiononne-thionon))
          (. :-:))
  (ID OSHeliandC.650.1187-1188))

( (IP-MAT-SPE (CODE <C>)
              (NEG Ni-ne)
              (VBI^2^PL latat-latan)
              (IP-INF (NP-SBJ (PRO$^A^2^SG iuuan-iuwa)
                              (N^A^SG hugi-hugi))
                      (VB tuiflean-twiflian))
              (. !-!)
              (' '-'))
  (ID OSHeliandC.514.948))

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (HVDI^3^SG haƀda-hebbian)
          (ADVP-TMP (ADV eft-eft))
          (IP-SMC (NP-SBJ (PRO$^A^3^SG is-is)
                          (N^A^SG uuord-word))
                  (ADJP-PRD (ADJ^A^SG garo-garu)))
          (CODE <R_2024>)
          (NP-SBJ (ADJ^N^SG mahtig-mahtig)
                  (N^N^SG barn-barn)
                  (NP-POS (NPR^G^SG godes-god))))
  (ID OSHeliandC.1045.2023-2024))

As in the YCOE, and unlike in the PPCME2, presentational thar is not specially tagged, and is always part of an ADVP-LOC. Expletive it is also not specially marked, except as part of an expletive construction, when it is part of an NP-SBJ-X.

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (VBDI^3^SG quat-kwethan)
          (CP-THT (C that-that)
                  (IP-SUB (ADVP-LOC (ADV thar-thar))
                          (QP-1 (ADV so-so)
                                (Q^N^SG filo-filo))
                          (BEDS^3^SG uuari-wesan)
                          (CODE <R_2833>)
                          (NP-SBJ (QP *ICH*-1)
                                  (NP-POS (N^G^PL manno-man))
                                  (N^N^SG menigi-menigi))))
          (. :-:))
  (ID OSHeliandC.1537.2832-2833))

( (IP-MAT (CODE <C>)
          (NP-SBJ *con*)
          (VBDI^3^SG quat-kwethan)
          (CP-THT (C that-that)
                  (IP-SUB (NP-SBJ-X (PRO^N^3^SG it-it))
                          (NP-OB2 (NPR^D^SG gode-god))
                          (BEDS^3^SG uuari-wesan)
                          (CODE <R_2712>)
                          (NP-OB2-PRN (N^D^SG uualdande-waldand))
                          (ADJP-PRD (ADJ^N^SG uuidarmuod-witharmod))
                          (CODE <C>)
                          (CP-THT-X (C that-that)
                                     ...
  (ID OSHeliandC.1461.2711-2714))

Free relative subjects are labelled -SBJ. Extraposed clausal subjects are coindexed with *exp* as part of an expletive construction. There are no other clausal subjects in the HeliPaD.

( (IP-MAT (CODE <R_654>)
          (BEDI^3^PL uuarun-wesan)
          (NP-OB2 (PRO^D^3^PL im-he))
          (NP-PRD (ADJ^N^PL glauua-glau)
                  (N^N^PL guomon-gumo))
          (, ,-,)
          (CODE <C>)
          (CP-FRL-SBJ (WNP-SBJ-1 (D^N^PL the-the))
                      (C 0)
                      (IP-SUB (NP-SBJ *T*-1)
                              (NP-OB1 (D^A^PL thea-the)
                                      (N^A^PL gifa-geva))
                              (VBDI^3^PL leddun-ledian)))
          (. .-.))
  (ID OSHeliandC.349.654))

( (IP-MAT (CODE <C>)
          (NP-SBJ-X *exp*)
          (NP-OB2 (PRO^D^3^SG Im-he))
          (BEDI^3^SG uuas-wesan)
          (ADVP-TMP (ADV thuo-tho))
          (NP-PRD (N^N^SG uuilleo-willio)
                  (ADJ^N^SG mikil-mikil))
          (CODE <R_872>)
          (CP-THT-X (C that-that)
                    (IP-SUB (NP-SBJ (PRO^N^3^SG hie-he))
                            (PP (P fan-fan)
                                (NP (ADJ^D^PL solicon-sulik)
                                    (N^D^PL salđon-salitha)))
                            (CODE <C>)
                            (VB seggean-seggian)
                            (MDDS^3^SG muosti-motan)))
          (. .-.))
  (ID OSHeliandC.470.871-872))

Non-subject arguments

There are three types of non-subject argument: -OB1, -OB2, and -PRD.

The two types of object are used, broadly speaking, for accusative objects and for dative objects respectively. With certain verbs, genitive objects can also be either -OB1 or -OB2, depending on the case of the other object. Consult the treatment of individual words for details. With one verb, lerian, both objects can be accusative objects, with the people being taught as -OB2.

Single-object clauses:

( (IP-MAT (CODE <R_57>)
          (HVDI^3^PL habdun-hebbian)
          (PP (P fan-fan)
              (NP (NPR^D^SG Rumuburg-Rumuburg)))
          (CODE <MS_6b>)
          (CODE <C>)
          (NP-OB1 (N^A^SG riki-riki))
          (GE+VN giuunnan-winnan)
          (CODE <R_58>)
          (NP-SBJ (N^N^PL helmgitrosteon-helmgitrostio))
          (. :-:))
  (ID OSHeliandC.16.57-58))

( (IP-MAT-SPE (CODE <R_1612>)
              (CONJ ac-ak)
              (VBI^2^SG hilp-helpan)
              (NP-OB2 (PRO^D^1^PL us-we))
              (PP (P uuiđar-withar)
                  (NP (Q^D^PL allon-al)
                      (CODE <C>)
                      (ADJ^D^PL ubilon-uvil)
                      (N^D^PL dadeon-dad)))
              (. .-.)
              (" "-"))
  (ID OSHeliandC.857.1612))

Double-object clauses:

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Than-than))
          (HVDI^3^SG habda-hebbian)
          (ADVP-TMP (ADV thuo-tho))
          (NP-SBJ (N^N^SG drohtin-drohtin)
                  (NPR^N^SG god-god))
          (CODE <R_54>)
          (NP-OB2 (NP-POS (NPR^G^PL Romano-Roman))
                  (N^D^PL liudeon-liudi))
          (VN farliuuan-farlihan)
          (CODE <C>)
          (NP-OB1 (NP-POS (N^G^PL rikeo-riki))
                  (QS^A^SG mesta-mer))
          (. :-:))
  (ID OSHeliandC.14.53-54))

( (IP-MAT (CODE <R_5308>)
          (IP-MAT-0 (NP-SBJ *con*)
                    (MDDI^3^SG uuelda-willian)
                    (NP-OB1 (NP-POS (N^G^PL manno-man))
                            (N^A^PL barn-barn))
                    (CODE <C>)
                    (NP-OB2 (N^G^SG morthies-morth))
                    (VB atuomian-atomian))
          (, ,-,)
          (CODE <R_5309>)
          (CONJP (IPX-MAT=0 (VB nerian-nerian)
                            (PP (P af-af)
                                (NP (N^D^SG nodi-nod)))))
          (. .-.))
  (ID OSHeliandC.3143.5308-5309))

( (IP-MAT (CODE <R_2271>)
          (NP-SBJ *con*)
          (VBDI^3^SG lerda-lerian)
          (NP-OB2 (PRO^A^3^PL sia-he))
          (NP-OB1 (PRO$^A^3^SG iro-iru)
                  (N^A^SG giloƀon-gilovo)))
  (ID OSHeliandC.1201.2271))

Predicates, unlike other arguments, may be NPs or other phrases, most usually ADJP. They are usually nominative, but may also be genitive, or (in small clauses) accusative.)

( (IP-MAT (CODE <P_10>)
          (CODE <R_76>)
          (NP-PRD (NPR^N^SG Zacharias-Zacharias))
          (BEDI^3^SG uuas-wesan)
          (NP-SBJ (PRO^N^3^SG hie-he))
          (VN hetan-hetan)
          (. .-.))
  (ID OSHeliandC.24.76))

( (IP-MAT (CODE <R_82>)
          (NP-SBJ *con*)
          (BEDI^3^PL uuarun-wesan)
          (ADJP-PRD (ADV so-so)
                    (ADJ^N^PL gihoriga-gihorig))
          (CODE <C>)
          (NP-OB2 (N^D^SG heƀancuninge-hevankuning))
          (. ,-,))
  (ID OSHeliandC.31.82))

( (IP-MAT (CODE <R_347>)
          (VBDS^3^SG quami-kuman)
          (PP (P te-te)
              (NP (D^D^SG them-the)
                  (N^D^SG cnuosla-knosal)
                  (CP-REL *ICH*-1)))
          (NP-SBJ (GE+WPRO^N gihue-gihwe))
          (CODE <C>)
          (CP-REL-1 (WADVP-DIR-2 (ADV thanan-thanan))
                    (C 0)
                    (IP-SUB (ADVP-DIR *T*-2)
                            (NP-SBJ (PRO^N^3^SG hie-he))
                            (NP-PRD (N^G^SG cunneas-kunni))
                            (BEDI^3^SG uuas-wesan)
                            (, ,-,)
                            (CODE <R_348>)
                            (PTP-PRD-PRN (GE+VN giboran-beran)
                                         (PP (P fan-fan)
                                             (NP (D^D^PL them-the)
                                                 (N^D^PL burgion-burg))))))
          (. .-.))
  (ID OSHeliandC.194.347-348))

( (IP-MAT-SPE-0 (CODE <C>)
                (HVI^2^PL Hebbeat-hebbian)
                (IP-SMC-SPE (NP-SBJ (PRO$^A^2^SG iuuuan-iuwa)
                                    (N^A^SG muod-mod))
                            (PP (P uuider-withar)
                                (NP (D^D^PL them-the)))
                            (CODE <R_1877>)
                            (ADJP-PRD (ADV so-so)
                                      (ADJ^A^SG glauuan-glau))
                            (ADVP-DIR (ADV tegegnes-tegegnes)))
                (CODE <C>)
                 ...
  (ID OSHeliandC.979.1876-1880))

Verbs which take predicates include particularly wesan (BE) and weorthan (RD), but also hetan (in the passive) and a number of more minor constructions. Predicates are usually nominative, but can also be genitive ("they were of his kin").

Adjuncts

Adjuncts in the HeliPaD behave as they do in the YCOE. -TMP (temporal), -DIR (directional), -LOC (locative), -LFD (left-dislocation), and -VOC (vocative) should all be treated as subcases of adjuncts, which if not further specified are marked -ADT.

The division into arguments and adjuncts may not be watertight in all cases. Specialists in argument structure using the corpus are advised to check the annotation carefully and use their own judgement.

Case attraction

The determiners in some relative clauses may agree in case with the antecedent rather than bearing the case of the gap. If possible, these are treated as instances of apposition.

(NP-OB1-PRN (D^A^SG thena-the)
            (CP-REL-SPE (WNP-SBJ-3 (D^N^SG the-the))
                        (C 0)
                        (IP-SUB-SPE (NP-SBJ *T*-3)
                                    (NP-OB1 (Q^A^SG all-al)
                                            (N^A^SG reht-reht))
                                    (VBPI^3^SG bikann-bikunnan))))

(OSHeliandC.1017.1957-1961)

True case attraction is never annotated at phrase level (since the HeliPaD does not have phrase-level case labels), but there may be a mismatch between (gap) grammatical function and word-level case if a treatment as apposition would lead to a relative clause with both an empty complementizer and an empty operator.

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (NP-SBJ (D^N^SG thiu-the)
                  (N^N^SG magat-magath))
          (HVDI^3^SG habda-hebbian)
          (CODE <R_2767>)
          (GE+VN githionot-thionon)
          (PP (P ti-te)
              (NP (N^D^SG thanke-thank)))
          (CODE <C>)
          (NP-OB2 (NP-OB2 (N^D^SG thiodcuninge-thiodkuning))
                  (CODE <MS_77b>)
                  (CODE <R_2768>)
                  (CONJP (CONJ endi-endi)
                         (NP-OB2 (Q^D^SG allon-al)
                                 (D^D^SG them-the)
                                 (N^D^SG erlscipie-erlskepi)
                                 (CODE <C>)
                                 (CP-REL (WNP-SBJ-1 (D^D^SG them-the))
                                         (C 0)
                                         (IP-SUB (NP-SBJ *T*-1)
                                                 (PP (ADVP-LOC (ADV thar-thar))
                                                     (P inne-inne))
                                                 (BEDI^3^SG uuas-wesan)))
                                 (CODE <R_2769>)
                                 (NP-POS (ADJ^G^PL godaro-god)
                                         (N^G^PL gumono-gumo)))))
          (. :-:))
  (ID OSHeliandC.1494.2766-2769))

Adjective Phrases (ADJP)

ADJP is an endocentric phrasal category, headed by an adjective or a first- or second-person possessive pronoun (which can be thought of as a subclass of adjective). It is normally only projected when an adjective is modified (or takes a complement - though this is rare in the HeliPaD) or when the adjective is predicative, scrambled or extraposed and therefore not contained within an NP.

The most common modifiers for adjectives are adverbs or ADVPs. As usual for modifiers, a single adverb does not project a phrase, and adverbs only do so if they are themselves modified.

( (IP-MAT (CODE <C>)
          (NP-SBJ (D^N^SG That-the))
          (BEDI^3^SG uuas-wesan)
          (NP-PRD (ADJP (ADV so-so)
                        (ADJ^N^SG salig-salig))
                  (N^N^SG man-man))
          (, ,-,)
           ...
  (ID OSHeliandC.25.76-78))

( (IP-MAT (CODE <C>)
          (NEG ni-ne)
          (CODE <MS_7a>)
          (MDDI^3^SG muosta-motan)
          (NP-OB2 (PRO^D^3^PL im-he))
          (NP-SBJ (N^N^SG erƀiuuard-erviward))
          (CODE <R_80>)
          (PP (P an-an)
              (NP (PRO$^D^3^SG iro-iru)
                  (N^D^SG iuguthedi-juguthhed)))
          (CODE <C>)
          (ADJP-PRD (ADJ^N^SG giƀithig-gividig))
          (RD uuerthan-werthan)
          (. –-–))
  (ID OSHeliandC.28.79-80))

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (RDDI^3^SG uuarth-werthan)
          (NP-SBJ (D^N^SG that-the)
                  (CP-QUE-PRN *ICH*-1))
          (ADJP-PRD (ADVP (ADV so-so)
                          (ADV uuido-wido))
                    (ADJ^N^SG cuth-kuth))
           ...
          (. :-:))
  (ID OSHeliandC.1074.2071-2074))

Instrumental determiners and adjectives are taken as modifiers, as with ADVPs.

(CP-THT (C that-that)
        (IP-SUB (NP-SBJ (PRO^N^3^SG hie-he))
                (NEG ni-ne)
                (BEDS^3^SG uuari-wesan)
                (NP-OB2 (D^D^SG them-the)
                        (NP-POS (N^G^PL manno-man))
                        (N^D^SG folke-folk))
                (, ,-,)
                (CODE <R_5542>)
                (NP-OB2-PRN (D^D^SG them-the)
                            (N^D^SG uuerode-werod))
                (ADJP-PRD (D^I^SG thiu-the)
                          (ADJR^N^SG uurethra-wreth))))

(OSHeliandC.3278.5540-5542)

( (IP-MAT-SPE (CODE <R_1781>)
              (ADVP-TMP (ADV Thann-than))
              (VBPI^3^SG ligit-liggian)
              (ADVP-TMP (ADV eft-eft))
              (ADJP-1 (ADJ^N^SG oder-othar))
              (CODE <C>)
              (ADJP-2 (ADJR^N^SG engera-engi)
                      (ADJ^I^SG mikilu-mikil))
              (CODE <R_1782>)
              (NP-SBJ (ADJP *ICH*-1)
                      (ADJP *ICH*-2)
                      (N^N^SG uueg-weg))
              (PP (P an-an)
                  (NP (D^D^SG thesaro-these)
                      (N^D^SG uueroldi-werold)))
              (. ,-,))
  (ID OSHeliandC.936.1781-1782))

Adjective phrases may also contain an NP-POS, especially when superlative.

( (IP-MAT-SPE (CODE <R_2595>)
              (NP-SBJ (D^N^SG that-the))
              (BEPI^3^SG is-wesan)
              (ADJP-PRD (NP-POS (Q^G^PL allero-al)
                                (N^G^PL beuuo-beo))
                        (ADJS^N^SG bredost-bred))
              (. ;-;))
  (ID OSHeliandC.1397.2595))

As in the YCOE, ordinal numbers are labelled as adjectives.

( (IP-MAT (CODE <C>)
          (NP-SBJ (Q^N^SG sum-sum))
          (ADVP-LOC (ADV thar-thar))
          (ALSO oc-ok)
          (ADVP-TMP (ADV sithor-sithor))
          (VBDI^3^SG quam-kuman)
          (CODE <R_3422>)
          (PP (P an-an)
              (NP (D^A^SG thia-the)
                  (ADJ^A^SG elliftun-ellifto)
                  (N^A^SG tid-tid)))
          (. .-.))
  (ID OSHeliandC.1946.3421-3422))

Quantifier Phrases (QP)

Quantifier phrases are only found in modifying function in the HeliPaD, unlike ADJPs. Quantifiers and QPs can be either pre- or postnominal. The QP is only projected when the quantifier is modified or when it is floated, scrambled, or extraposed.

(NP-OB1 (QP (ADV so-so)
            (Q^A^SG manag-manag))
        (N^A^SG gibod-gibod)
        (NP-POS (NPR^G^SG godes-god)))

(OSHeliandC.4.12-15)

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (RDDI^3^SG uuarth-werthan)
          (NP-SBJ (NP-POS (D^G^SG thes-the)
                          (N^G^SG uuiƀes-wif))
                  (N^N^SG hugi-hugi)
                  (QP *ICH*-1))
          (CODE <R_282>)
          (PP (P after-aftar)
              (NP (D^D^SG them-the)
                  (N^D^SG arundie-arundi)))
          (CODE <C>)
          (QP-1 (Q^N^SG all-al))
          (GE+VN gihuorban-hwervan)
          (CODE <R_283>)
          (PP (P an-an)
              (NP (NP-POS (NPR^G^SG godes-god))
                  (N^A^SG uuilleon-willio)))
          (. .-.))
  (ID OSHeliandC.155.281-283))

Like other modifiers separated from their head, floated quantifiers are traced to the head, as in the above example.

Consult the treatment of individual words to find out which items are treated as quantifiers.

al may be treated as an adverb (ADV) in some cases, particularly when introducing al so-clauses.

( (IP-MAT (CODE <R_2656>)
          (NP-SBJ *con*)
          (VBDI^3^SG auuohs-awahsan)
          (ADVP (ADV all-al))
          (PP (P under-undar)
              (NP (D^D^SG theson-these)
                  (N^D^SG uueroda-werod)))
          (. :-:))
  (ID OSHeliandC.1428.2656))

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (HVDI^3^SG habda-hebbian)
          (NP-SBJ (PRO^N^3^SG hie-he))
          (NP-OB1 (NP-POS (PRO$^G^1^SG usas-usa)
                          (N^G^SG uualdandes-waldand))
                  (CODE <R_191>)
                  (N^A^SG geld-geld))
          (GE+VN gilestid-lestian)
          (, ,-,)
          (CODE <C>)
          (CP-ADV (ADVP (ADV all-al))
                  (C so-so)
                  (IP-SUB (NP-SBJ (PRO$^N^3^SG is-is)
                                  (N^N^SG gigengi-gigengi))
                          (BEDI^3^SG uuas-wesan)
                          (CODE <R_192>)
                          (GE+VN gimarcod-markon)
                          (PP (P mid-mid)
                              (NP (N^D^PL mannon-man)))))
          (. .-.))
  (ID OSHeliandC.89.190-192))

Number Phrases (NUMP)

In most respects, number phrases behave like quantifier phrases. They are endocentric, though they may contain multiple "heads" (in which case they are treated like compounds). NUMP is only used when there are multiple words in the same number, when the numbers are modified in some way, or when the numbers have been separated from their head. Complex numbers are not given any internal structure, except when involving conjunction.

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (BEDI^3^SG uuas-wesan)
          (NP-SBJ (PRO^N^3^SG siu-siu))
          (NP-PRD (N^N^SG uuidua-widowa))
          (PP (P after-aftar)
              (NP (D^D^SG thiu-the)))
          (CODE <R_513>)
          (PP (P at-at)
              (NP (D^D^SG them-the)
                  (N^D^SG friđuuuihe-frithuwih)))
          (CODE <C>)
          (NP-TMP (NUMP (NUMP (NUM^A^PL fiuuar-fiuwar))
                        (CONJP (CONJ endi-endi)
                               (NUMP (NUM^A^PL ahtoda-ahto))))
                  (CODE <R_514>)
                  (NP-POS (N^G^PL uuintro-wintar)))
           ...
  (ID OSHeliandC.272.512-516))

( (IP-MAT (CODE <C>)
           ...
          (CP-ADV (C antthat-untat)
                  (IP-SUB (ADVP-LOC (ADV thar-thar))
                          (NP-SBJ (N^N^PL uueros-wer))
                          (ADVP-DIR (ADV ostan-ostan))
                          (CODE <R_542>)
                          (NP-SBJ-PRN (ADJP (ADV suitho-switho)
                                            (ADJ^N^PL glauua-glau))
                                      (N^N^PL gumon-gumo)
                                      (NUMP *ICH*-1))
                          (CODE <C>)
                          (IP-INF-NCO (VB gangan-gangan))
                          (VBDI^3^PL quamun-kuman)
                          (CODE <R_543>)
                          (NUMP-1 (NUM^N^PL threa-thrie))
                          (PP (P te-te)
                              (NP (D^D^SG thero-the)
                                  (N^D^SG thiodo-thiod)))
                           ...
          (. :-:))
  (ID OSHeliandC.284.540-544))

The number en ("one") is always tagged as a number, even when it means "alone".

(NP-SBJ (NPR^N^SG god-god)
        (NUM^N^SG eno-en))

(OSHeliandC.1234.2322-2324)


Other constituents

Adverb Phrases (ADVP)

Adverb phrases are a straightforward endocentric phrasal category. They may take the extended labels -LOC (locative), -DIR (directional) or -TMP (temporal). They normally consist of an adverb on its own, potentially with another adverb modifying it (usually switho, so, than, or with temporal adverbs lango). Adverbs other than these are, as a rule, separated into different ADVPs even when adjacent and functionally equivalent.

( (IP-MAT (CODE <C>)
          (NP-SBJ (D^N^SG that-the))
          (BEPI^3^SG is-wesan)
          (ADVP-TMP (ADV noh-noh))
          (ADVP-TMP (ADV lango-lango))
          (ADJP-PRD (ADJ^N^SG scin-skin))
          (, ,-,)
          (CODE <MS_21b>)
          (CODE <R_648>)
          (PTP-PRD-PRN (GE+VN gicuthid-kwethan)
                       (NP-SBJ (N^N^SG craft-kraft)
                               (NP-POS (NPR^G^SG godes-god))))
          (. .-.))
  (ID OSHeliandC.342.647-648))

Instrumental determiners and adjectives are taken as modifiers, as with ADJPs.

( (IP-MAT (CODE <C>)
          (VBDI^3^PL truodun-truon)
          (NP-SBJ (PRO^N^3^PL sia-he))
          (ADVP-TMP (ADV sithor-sithor))
          (CODE <R_2070>)
          (ADVP (D^I^SG thiu-the)
                (ADVR mer-mer))
          (PP (P an-an)
              (NP (PRO$^A^3^SG is-is)
                  (N^A^SG mundburd-mundburd)))
           ...
          (. .-.))
  (ID OSHeliandC.1073.2069-2071))

( (IP-MAT (CODE <C>)
          (NP-SBJ (N^N^PL erlos-erl))
          (VBDI^3^PL thrungun-thringan)
          (CODE <R_182>)
          (ADVP-DIR (ADVR nahor-nah)
                    (ADJ^I^SG mikilu-mikil))
          (. :-:))
  (ID OSHeliandC.85.181-182))

Adverbs are not treated as taking complements in HeliPaD. Their apparent complements (e.g. the complement of nah and giliko below) are parsed as clausal adjuncts or arguments.

( (IP-MAT (CODE <R_2381>)
          (CONJ ac-ak)
          (VBDI^3^SG geng-gangan)
          (NP-OB2-RFL (PRO^D^3^SG im-he))
          (ADVP-TMP (ADV thuo-tho))
          (NP-SBJ (NP-SBJ (D^N^SG thie-the)
                          (ADJ^N^SG guodo-god))
                  (CODE <C>)
                  (CONJP (CONJ endi-endi)
                         (NP-SBJ (PRO$^N^3^PL is-is)
                                 (N^N^PL iungron-jungaro))))
          (PP (P mid-mid)
              (NP (PRO^D^3^SG im-he)))
          (, ,-,)
          (CODE <R_2382>)
          (NP-SBJ-PRN (N^N^SG friđubarn-frithubarn)
                      (NP-POS (NPR^G^SG godes-god)))
          (CODE <C>)
          (NP-LOC (D^D^SG them-the)
                  (N^D^SG fluode-flod))
          (ADVP-DIR (ADVR nahor-nah))
          (CODE <R_2383>)
          (PP (P an-an)
              (NP (NUM^A^SG en-en)
                  (N^A^SG scip-skip))
              (ADV innan-innan))
          (. ,-,))
  (ID OSHeliandC.1270.2381-2383))

( (IP-MAT-SPE-0 (CODE <R_2603>)
                (ADVP-LOC (ADV thar-thar))
                (MDPI^3^PL sculun-skulan)
                (NP-SBJ (PRO^N^3^PL sia-he))
                (PTP-ADT (GE+VNI^N^PL gibundana-bindan))
                (CODE <C>)
                (NP-OB1 (ADJ^A^SG bittra-bittar)
                        (N^A^SG logna-logna))
                (, ,-,)
                (CODE <R_2604>)
                (NP-OB1-PRN (N^A^SG thrauuerc-thrawerk))
                (VB tholon-tholon)
                (, ,-,)
                (CODE <C>)
                (IPX-MAT-PRN-SPE=0 (CONJ endi-endi)
                                   (CODE <MS_73a>)
                                   (NP-SBJ (D^N^PL thia-the)
                                           (ADJ^N^PL ođra-othar))
                                   (NP-OB1 (N^A^SG thioduuelon-thiodwelo))
                                   (CODE <R_2605>)
                                   (PP (P an-an)
                                       (NP (N^D^SG hebanrikie-hevanriki)))
                                   (, ,-,)
                                   (CODE <C>)
                                   (NP-OB2 (ADJ^D^SG huitero-hwit)
                                           (N^D^SG sunnun-sunno))
                                   (CODE <P_180>)
                                   (CODE <R_2606>)
                                   (VB luhtian-liuhtian)
                                   (ADVP (ADV gilico-giliko)))
                (. .-.))
  (ID OSHeliandC.1405.2603-2606))

Prepositional Phrases (PP)

Prepositional phrases in the HeliPaD are an endocentric category that take NPs (unmarked for function) as well as some types of CPs and ADVP-LOC as their complement.

(PP (P ti-te)
    (NP (D^D^SG them-the)
        (N^D^SG giuuirkie-giwerki)))

(OSHeliandC.7.19-20)

(PP (P neuan-newan)
    (CP-THT (C that-that)
            (IP-SUB (NP-SBJ (PRO^N^3^PL sia-he)
                            (NUM^N^PL fiori-fiuwar))
                    (PP (P te-te)
                        (NP (D^I^SG thio-the)))
                    (CODE <R_17>)
                    (PP (P thuru-thurh)
                        (NP (N^A^SG craft-kraft)
                            (NP-POS (NPR^G^SG godas-god))))
                    (CODE <C>)
                    (GE+VNI^N^PL gecorana-kiosan)
                    (RDDI^3^PL uurđun-werthan))))

(OSHeliandC.5.15-17)

In the HeliPaD, subordinating conjunctions are not treated as prepositions.

There are several multi-word prepositions in the HeliPad; these are RPs within PPs (preceding the preposition), or, more frequently, ADVs within PPs (usually locative or directional elements following the complement).

(PP (RP up-up)
    (P te-te)
    (NP (N^D^SG himile-himil)))

(OSHeliandC.797.1484-1491)

(PP (P an-an)
    (NP (N^A^SG felis-felis))
    (ADV uppan-uppan))

(OSHeliandC.948.1801-1810)

Where the latter class occurs without an NP complement, the ADVP-LOC is taken to be the complement itself.

(PP (P fan-fan)
    (ADVP-LOC (ADV ostan-ostan)))

(OSHeliandC.298.563-566)

Elements can occur before the P of the PP, in modifying function.

(PP (ADV suitho-switho)
    (P an-an)
    (NP (N^D^PL sorogon-sorga)))

(OSHeliandC.1514.2801-2802)

In several constructions, the NP or an ADVP-LOC complement can occur before the P. This is most common with thar or a personal pronoun.

( (IP-MAT (CODE <C>)
          (CONJ endi-endi)
          (NP-SBJ *con*)
          (NP-OB2 (PRO^D^3^SG im-he))
          (NP-OB1 (D^A^PL tha-the)
                  (N^A^PL geƀa-geva))
          (VBDI^3^PL druogun-dragan)
          (, ,-,)
          (CODE <P_50>)
          (CODE <R_674>)
          (NP-OB1-PRN (NP-OB1-PRN (N^A^SG gold-gold))
                      (CONJP (CONJ endi-endi)
                             (NP-OB1-PRN (N^A^SG uuihroc-wihrok)))
                      (CONJP *ICH*-1))
          (CODE <C>)
          (PP (P bi-bi)
              (NP (NP-POS (NPR^G^SG godes-god))
                  (N^D^PL teknon-tekan)))
          (CODE <R_675>)
          (CONJP-1 (CONJ endi-endi)
                   (NP-OB1-PRN (N^A^SG mirra-myrra)))
          (PP (ADVP-LOC (ADV thar-thar))
              (P midi-mid))
          (. .-.))
  (ID OSHeliandC.361.673-675))

( (IP-MAT (CODE <C>)
          (ADVP-TMP (ADV Thuo-tho))
          (VBDI^3^SG geng-gangan)
          (PP (NP (PRO^D^3^PL im-he))
              (P tuo-to))
          (NP-SBJ (D^N^SG thie-the)
                  (NP-POS (N^G^SG landes-land))
                  (N^N^SG uuard-ward))
          (. ,-,))
  (ID OSHeliandC.1764.3155))

Interjection Phrases (INTJP)

Interjections (INTJ) are dominated by an INTJP when modified or taking a complement. Otherwise they occur alone. Either way, INTJ(P) is itself at the highest phrasal level.

( (IP-MAT-SPE (CODE <R_1522>)
              (NP-SBJ *con*)
              (VBPS^3^SG queđe-kwethan)
              (INTJ ia-ja)
              (CP-ADV-SPE (C ef-ef)
                          (IP-SUB-SPE (NP-SBJ (PRO^N^3^SG it-it))
                                      (BEPS^3^SG si-wesan)))
              (. ,-,))
  (ID OSHeliandC.807.1522))

Interjections may take a question or that-clause as complement.

( (QTP (CODE <R_3024>)
       (' '-')
       (INTJP (INTJ Uuola-wola)
              (CP-THT-SPE (C that-that)
                          (IP-SUB-SPE (NP-SBJ (PRO^N^2^SG thu-thu))
                                      (NP-VOC (N^N^SG uuiƀ-wif))
                                      (HVPI^2^SG haƀis-hebbian)
                                      (CODE <C>)
                                      (NP-OB1 (N^A^SG uuilleon-willio)
                                              (ADJ^A^SG guodan-god)))))
       (. !-!))
  (ID OSHeliandC.1672.3024))

INTJP is usually an endocentric category, but can also dominate other elements - usually wh-"adverbial" hwat (which may well not really be an interjection here, and not merit the following punctuation, but is readily retrievable this way).

( (IP-MAT-SPE (CODE <C>)
              (INTJP (WADV Huat-hwe))
              (, ,-,)
              (NP-SBJ (PRO^N^1^PL uui-we)
                      (QP *ICH*-1))
              (NP-OB1 (D^A^SG thia-the))
              (ADVP-LOC (ADV hier-her))
              (VBPI^1^PL uuitun-witan)
              (QP-1 (Q^N^PL alla-al))
              (. ,-,))
  (ID OSHeliandC.1426.2654))

Participle Phrases (PTP)

Phrases headed by a participle that are not part of the main sequence of verbs (e.g. perfect or future periphrasis or modal construction) are labelled PTP. The participle can be either inflected (e.g. VNI, VGI) or uninflected (e.g. VN, VG). PTP is a formal, not a functional, label. On the whole, PTPs behave like ADJPs, and may be marked as adjuncts or predicates (the latter only when modified or in apposition). PPs that are adjacent to an NP they modify are contained within that NP.

(CP-ADV (ADVP-LOC (ADV thar-thar))
        (C 0)
        (IP-SUB (NP-SBJ (PRO^N^3^SG hie-he))
                (PP (P an-an)
                    (NP (PRO$^D^3^SG is-is)
                        (N^D^SG benki-bank)))
                (VBDI^3^SG sat-sittian)
                (CODE <R_2747>)
                (PTP-ADT (NP-ADT (N^I^SG uuinu-win))
                         (GE+VN giuulenkid-wlenkian))))

(OSHeliandC.1483.2745-2747)

( (IP-MAT (CODE <C>)
          (NP-SBJ (PRO^N^3^SG Hie-he))
          (NEG ni-ne)
          (BEDI^3^SG uuas-wesan)
          (ADVP (ADV thoh-thoh))
          (PP (P mid-mid)
              (NP (N^D^PL sibbeon-sibbia))
              (NP-PRN *ICH*-1)
              (NP-PRN *ICH*-2))
          (ADJP-PRD (ADJ^N^SG bifang-bilang))
          (CODE <R_65>)
          (NP-PRN-1 (N^D^PL auaron-avaro)
                    (NP-POS (NPR^G^SG Israheles-Israhel)))
          (CODE <C>)
          (NP-PRN-2 (N^D^SG eđiligiburdi-ethiligiburd))
          (, ,-,)
          (CODE <R_66>)
          (PTP-PRD-PRN (VN cuman-kuman)
                       (PP (P fon-fan)
                           (NP (PRO$^D^3^SG iro-iru)
                               (N^D^SG cnuosle-knosal))))
           ...
          (. .-.))
  (ID OSHeliandC.20.64-72))

PTPs have a sentential internal syntax, and may therefore take arguments, adjuncts, etc.

(PTP-ADT (NP-ADT (N^G^SG tionon-tiono))
         (VN atomid-atomian))

(OSHeliandC.555.1014-1018)

(PTP-PRD-PRN (GE+VN gicuthid-kwethan)
             (NP-SBJ (N^N^SG craft-kraft)
                     (NP-POS (NPR^G^SG godes-god))))

(OSHeliandC.342.647-648)

Foreign phrases (LATIN)

Foreign phrases, including isolated Latin titles, are labelled with the exocentric phrasal category LATIN and treated as a separate token. These phrases usually only include foreign words (FW).

( (LATIN (CODE <R_1600>)
         (FW Pater-pater)
         (FW noster-noster)
         (. :-:))
  (ID OSHeliandC.848.1600))

Wh-phrases (W*P)

There are in principle five types of wh-phrase, each corresponding to a non-wh-phrase: WNP, WQP, WPP, WADJP, WADVP. These phrases may be endocentric (i.e. headed by a wh-word), but need not be. See the POS manual and Treatment of individual words for details of how wh-words are tagged.

WQP would be used in instances of a quantifier modified by a wh-word (e.g. hwo manag), but there are no such examples in the Heliand.

In relative clauses, the head of a wh-phrase is a demonstrative or thar, not a wh-word. Free relatives allow both types. Wh-phrases may also be empty.

(WNP-SBJ-2 (D^N^PL thia-the))

(OSHeliandC.3.9-12)

(WADVP-LOC-1 (ADV thar-thar))

(OSHeliandC.129.249-252)

(WNP-PRD-1 (WPRO^N huat-hwe))

(OSHeliandC.105.210-211)

(WNP-OB1-5 0)

(OSHeliandC.9.32-42)

Morphological wh- elements need not project a W*P if they are not used as part of an extraction structure. This differs from their treatment in the YCOE, where they are treated as (for instance) plain quantifiers in this situation.

(NP-OB1 (NP-POS (ADJ^G^SG godcundeas-godkund))
        (WPRO^A huat-hwe))

(OSHeliandC.88.187-190)

(NP-SBJ (WADJ^N^SG huilik-hwilik)
        (ADJ^N^SG helag-helag)
        (N^N^SG man-man))

(OSHeliandC.283.537-540)