τὸ ἐγχειρίδιον

Annotation Manual for the Penn Parsed Corpora of Historical Greek

Beatrice Santorini

Beatrice Santorini did not have direct authorship, but much of the material in this manual is taken directly from her Annotation Manual for the Penn Parsed Corpora of Historical English. 

Table of Contents

I. Introduction
1. General Introduction
Current State of the Project
Philosophy and Goals
File Formats
Part-of-Speech (POS) Tagged Files (.pos)
Parsed Files (.psd)
Textual Markup
2. List of Tags and Empty Categories
Part-of-Speech (POS) Tags
Syntactic Tags
3. Splitting Words
Syntax for Splitting Words
Situations that always result in splitting
Situations that never result in splitting
Treatment of Individual Words
4. POS Annotation
Adjectives and Adverbs (ADJ, ADV)
Notes
Colors
Degree Words
Ordinal Numbers
Cardinal Numbers (NUM)
Notes
Clitics (CL)
Notes
Complementizers (C)
Notes
Coordinating Conjunctions (CONJ, CLTE)
Notes
Determiners and Demonstratives (D)
Notes
Foreign Words (FW)
Notes
Interjections (INTJ)
Notes
Negation (NEG)
Notes
Compound NEG Tags
Nominals (N, PRO, CLPRO)
Notes: Nominals
Notes: Pronominals
Ordinal Numbers
Particles (CLPRT, AN, KE, PRTQ, CLGE)
Notes: Modal Particles
Notes: CLGE
Notes: CLPRT
Notes: PRTQ
Prepositions (P)
Notes
Punctuation
Notes
Quantifiers (Q)
Notes
Subordinating Conjunctions
Verbs (VB*, BE*)
Notes
Wh- Words (W)
Notes
Compound Tags
II. Syntactic Annotation
5. Introduction
General Principles
Dash Tags
Structural Principles
Internal Structure of Phrases
Internal Structure of Clauses
IPs
Imperatives
Non-wh CPs
Wh- CPs
Fronting to pre-complementizer position
Adjunction to CP
Division into Sentence Tokens
6. Phrase-Level Constituents
Adjective Phrase
Complements of ADJ
Modifiers of ADJ
Adverb Phrase
Complements of ADV
Modifiers of ADV
Conjunction Phrase
Fragment
Interjection Phrase
Noun Phrase
Heads of NP
Complements of N
Modifiers of N
Number Phrase
Prepositional Phrase
Complements of P
Specifiers of P
Quantifier Phrase
Complements of Q
Modifiers of Q
Wh- Phrase
WADJP
WADVP
WNP
WPP
WQP
Non-Structural Labels
CODE
Foreign Language Passages
QTP
7. Conjunction
Word-Level Conjunction
Phrase-Level Conjunction
CONJPs with Empty Heads
Conjunction of Unlike Categories
Recursive Conjunction Structures
Shared Modifiers
Shared Pre-Modifiers
Shared Post-Modifiers
Conjunction and Negation
Extraposed CONJPs
Correlative Conjunction
Word-Level Correlative Conjunction
Phrase-Level Correlative Conjunction
Floated Conjunctions
Clausal Conjunction
Conjunction of Matrix IPs
Conjunction of Embedded IPs
Right-Node Raising
Gapping
Backwards Gapping
Conjoined Direct Speech
Conjunction of CPs
Clausal Conjunction and Shared Modifiers
8. Empty Categories
Empty Subjects
Subjects Elided under Conjunction
Arbitrary Subjects in ECM Infinitives
pro-Drop Subjects
Traces
Traces of A-Movement
Traces of Wh-Movement
Traces of Other A'-Movement
Other Empty Categories
Empty Objects
Empty Wh- Phrases and Complementizers
A Generic Empty Category
The Position of Empty Categories
The Position of the Null Copula
The Position of Empty Subjects
The Position of Traces
Resumptive Pronouns
9. Clause-Level Constituents
Adjective Phrase
Predicate Adjectives
Other Clause-Level ADJPs (ADJP-SPR)
Movement out of ADJP
Adverb Phrase
Noun Phrase
Subject
Object
Non-Argument (Adjunct) NP
Adverbial NP
Attributive (and/or Possessive) NP
Complement NP
Directional NP
Instrumental NP
Locative NP
Measure Phrase
Partitive NP
Predicate NP
Secondary Predicate NP
Temporal NP
Vocative
Bare NP
Prepositional Phrase
Quantifier Phrase
Words as Clausal Constituents
Adverbs
Floated Quantifiers
Negation
Participles
Extended Dash Tags
Left Dislocation
Appositive or Parenthetical
Resumptive
Direct Speech
Unknown
10. Subordinate Clauses
Adverbial Clauses
That-Clauses
Complement Clauses
Purpose Clauses
Degree Complements
Result Clauses
Comparative Clauses
Correlative Comparatives
Questions
Wh- in situ
Exclamative Clauses
Ordinary Relative Clauses
Reduced Relative Clauses
Reduced Relative Clauses versus Participial Clauses
Clause-Adjoined Relative Clauses
Free Relatives
ὡς Clauses with a Nominal Gap
Empty Operator Clauses
Infinitival Clauses
IP-INF-ADT
IP-INF-COM
IP-INF-DEG
IP-INF-PRP
IP-INF-SBJ
IP-INF-THT
Small Clauses
Participial Clauses
Absolute Clauses
Supplementary Participles
Participles of Indirect Discourse
11. Special Constructions
Prolepsis
Clausal Subjects
'If not' = 'except'
12. Discontinuous Phrases
Rightward Extraposition
Default Leftward Movement
Word-Level Fronting
Fronting of Non-Constituents
Phrase-Level Fronting
Clitic Displacement
Sentence-Position Particles
Clitic Pronouns and Verbs
13. Disfluencies
Break
False Start
Repetition
Tag
III. Planned Corpus Development and Known Issues
14. Known Issues
State of the Corpora
State of Each Corpus
New POS Tags (and Parsing Guidelines)
To Be Revised
Revisions to Herodotus
A. Appendix
Copular Verbs
Verbs taking Small Clauses
Verbs taking Secondary Predicates
Cognate Object Verbs

List of Tables

2.1. Legal POS Tags
2.2. Extended Verbal Tags
2.3. Basic Syntactic Tags
2.4. Dash Tags
2.5. Extended Dash Tags
2.6. Empty Categories
4.1. ADJ and ADV Tags
4.2. NUM Tags
4.3. CL Tags
4.4. C Tags
4.5. CONJ and CLTE Tags
4.6. D Tags
4.7. FW Tags
4.8. INTJ Tags
4.9. NEG Tags
4.10. Nominal Tags
4.11. Particle Tags
4.12. P Tags
4.13. Punctuation Tags
4.14. Q Tags
4.15. VB* and BE* Tags
4.16. Extended Verbal Dash Tags
4.17. W Tags
5.1. Common IP-Dash Tag Combinations
5.2. Common CP-Dash Tag Combinations

Part I. Introduction

Chapter 1. General Introduction

Current State of the Project

As of November 1, 2013, the Penn Parsed Corpora of Historical Greek are no longer being developed.

Philosophy and Goals

  • Our primary goal has been to create an annotation system that facilitates automated searches, not to give a correct linguistic analysis of each sentence. For instance, if a construction can be found unambiguously through a combination of properties of a bracketed sentence, our annotation may not contain all of the structure that a full phrase structure diagram of the sentence would have.

  • We have tried to plan our system so that at each stage of the annotation, information is added in a monotonic way. In particular, we want any future revisions of the bracketed structures always to add information, never to change it. This goal requires us to avoid subjective judgments since they are extremely error-prone. So, for example, we do not distinguish object control from E(xceptional) C(ase) M(arking), nor do we attempt to implement the argument-adjunct distinction (except where we have decided to make our annotation similar to the annotation of the PROIEL Project and/or the Perseus Ancient Greek and Latin Dependency Treebanks—namely, in the annotation of DP modifiers of DPs).

  • As many categories as possible should have clear meanings so that unclear cases can be relegated to a small number of categories of residual cases. The price of making most categories homogeneous is that these residual categories will not be. In future revisions of the corpus, it may be possible to divide some of these residual categories into homogeneous subcategories.

  • As much as possible, we have avoided making decisions that would be controversial, whether with regard to text interpretation or to linguistic theory. In doubtful cases, we either avoid specifying structure, or we use default rules to decide the case for search purposes. An example of the first strategy concerns VPs. These are not indicated in the corpus, since VP boundaries are normally indeterminate. This is clearly the case in Classical and New Testament Greek, which allow scrambling and where the internal structure of the VP is variable and changing. Even in modern English there are many cases in which it is not clear whether some phrase attaches as a daughter of VP or higher up in the tree. An example of the second strategy concerns PP attachment. Whenever it is unclear where a PP attaches, we attach it by default as high as possible.

File Formats

At present, the corpus is only available in the default .psd format. However, in the future we plan to release the corpus in three different formats:

  • part-of-speech (POS) tagged: .pos

  • parsed: .psd

  • Common Treebank XML: .xml

Part-of-Speech (POS) Tagged Files (.pos)

Part-of-speech (POS) tagged texts have the extension .pos\index{files!.pos files}. They contain the material in the text files with a POS tag added to each word. Editorial material including milestones (e.g., chapter and verse numbers in the format chapter:verse) is given the tag CODE. Text elements are separated from their POS tags by an underscore. The text is divided into tokens in the same way as in the text files.

Parsed Files (.psd)

Parsed files have the extension .psd. They contain a labeled bracketing of the text, with the first set of labeled parentheses around a word repeating the information from the POS-tagged files. The division into tokens in the parsed files is the same as in the POS files. Each token is enclosed with its ID in a set of unlabeled parentheses.

( (IP-MAT-SPE (CODE {VS:1_8})
              (CONJ0 ἀλλὰ)
              (NP-SBJ *pro*)
              (VBPP-FUT λήμψεσθε)
              (NP-OB1 (NA δύναμιν))
              (IP-ABS (VPR$-AOR ἐπελθόντος)
                      (NP-SBJ (D$ τοῦ) (ADJ$ ἁγίου) (N$ πνεύματος))
                      (PP (P ἐφ')
                          (NP (PROA ὑμᾶς))))
              (. ,))
  (ID GreekNT,Acts))
                

( (IP-MAT-SPE (CODE {VS:1_8})
              (CONJ0 καὶ)
              (NP-SBJ *con*)
              (VBPP-FUT ἔσεσθέ)
              (NP-PRD (NP-ATR (CLPRO$ μου))
                      (NS μάρτυρες))
              (PP (PP (P ἔν)
                      (CLTE τε)
                      (NP (ND Ἰερουσαλὴμ)))
                  (CONJP (CONJ καὶ)
                         (PP (P ἐν)
                             (NP (QD πάσῃ) (DD τῇ) (ADJD Ἰουδαίᾳ))
                                 (CONJP (CONJ καὶ)
                                        (NX (ND Σαμαρίᾳ))))))
                  (CONJP (CONJ καὶ)
                         (PP (P ἕως)
                             (NP (ADJ$ ἐσχάτου) (D$ τῆς) (N$ γῆς)))))
              (. .))
  (ID GreekNT,Acts))
                

Textual Markup

In general, it has not been possible to retain the TEI XML markup conventions of the Perseus Digital Library in their original form because of conflicts with the annotation system. The major changes made are as follows:

  • Brackets around words that represent editorial emendations have been removed.

  • All XML markup, including markup indicating quoted speech, has been removed.

Chapter 2. List of Tags and Empty Categories

Part-of-Speech (POS) Tags

Table 2.1. Legal POS Tags

Tag Use
, sentence-internal punctuation
. sentence-final punctuation
" quotation mark
-$ genitive case marker
-A accusative case marker
-D dative case marker
A
ADJ adjective, nominative
ADJ$ adjective, genitive
ADJA adjective, accusative
ADJD adjective, dative
ADJR adjective, comparative degree
ADJS adjective, superlative degree
ADV adverb
ADVR adverb, comparative degree
ADVS adverb, superlative degree
AN modal particle ἄν
B
BED εἰμί, secondary sequence
BEI εἰμί, imperative
BEN εἰμί, infinitive
BEO εἰμί, optative
BEP εἰμί, primary sequence
BES εἰμί, subjunctive
BPR εἰμί, participle
C
C complementizer
CLGE emphatic clitic particle γε
CLPRO clitic pronoun, nominative (Homer only)
CLPRO$ clitic pronoun, genitive
CLPROA clitic pronoun, accusative
CLPROD clitic pronoun, dative
CLPRT second-position particle
CLQ clitic indefinite pronoun/quantifier τις/τι, nominative
CLQ$ clitic indefinite pronoun/quantifier τις/τι, genitive
CLQA clitic indefinite pronoun/quantifier τις/τι, accusative
CLQD clitic indefinite pronoun/quantifier τις/τι, dative
CLTE conjunctive clitic particle τε
CODE non-textual material (e.g., milestones and parser’s comments)
CONJ coordinating conjunction
CONJ0 coordinating conjunction, clause-level and token-initial
D
D determiner, nominative singular
D$ determiner, genitive singular
DA determiner, accusative singular
DD determiner, dative singular
DS determiner, nominative plural
DS$ determiner, genitive plural
DSA determiner, accusative plural
DSD determiner, dative plural
F
FW foreign word (often non-inflecting)
I
INTJ interjection
K
KE modal particle κε(ν)/κα
N
N common noun, nominative singular
N$ common noun, genitive singular
NA common noun, accusative singular
ND common noun, dative singular
NEG negation
NPR proper noun, nominative singular
NPR$ proper noun, genitive singular
NPRA proper noun, accusative singular
NPRD proper noun, dative singular
NPRS proper noun, nominative plural
NPRS$ proper noun, genitive plural
NPRSA proper noun, accusative plural
NPRSD proper noun, dative plural
NS common noun, nominative plural
NS$ common noun, genitive plural
NSA common noun, accusative plural
NSD common noun, dative plural
NUM cardinal number
O
OTHER form of ἄλλος, nominative
OTHER$ form of ἄλλος, genitive
OTHERA form of ἄλλος, accusative
OTHERD form of ἄλλος, dative
P
P preposition
PRO personal pronoun, nominative
PRO$ personal pronoun, genitive
PROA personal pronoun, accusative
PROD personal pronoun, dative
PRTQ particle, interrogative (e.g., ἆρα)
Q
Q quantifier, nominative
Q$ quantifier, genitive
QA quantifier, accusative
QD quantifier, dative
QR quantifier, comparative
QS quantifier, superlative
QV quantificational adverb (e.g., πάντως)
V
VBD verb, active secondary sequence (imperfect, aorist, pluperfect)
VBDP verb, middle/passive secondary sequence (imperfect, aorist, pluperfect)
VBI verb, active imperative
VBIP verb, middle/passive imperative
VBN verb, active infinitive
VBNP verb, middle/passive infinitive
VBO verb, active optative
VBOP verb, middle/passive optative
VBP verb, active primary sequence (present, future, perfect)
VBPP verb, middle/passive primary sequence (present, future, perfect)
VBS verb, active subjunctive
VBSP verb, middle/passive subjunctive
VPR verb, active participle, nominative
VPR$ verb, active participle, genitive
VPRA verb, active participle, accusative
VPRD verb, active participle, dative
VPRP verb, middle/passive participle, nominative
VPRP$ verb, middle/passive participle, genitive
VPRPA verb, middle/passive participle, accusative
VPRPD verb, middle/passive participle, dative
W
WADJ wh-adjective, nominative
WADJ$ wh-adjective, genitive
WADJA wh-adjective, accusative
WADJD wh-adjective, dative
WADV wh-adverb
WD wh-determiner, nominative
WD$ wh-determiner, genitive
WDA wh-determiner, accusative
WDD wh-determiner, dative
WP preposition in a wh- phrase
WPRO wh-pronoun, nominative
WPRO$ wh-pronoun, genitive
WPROA wh-pronoun, accusative
WPROD wh-pronoun, dative
WQ εἰ in the meaning ‘whether’ introducing an indirect question
X
X unknown part of speech
Y
Y may be used temporarily to tag a lemma that is difficult to categorize

Table 2.2. Extended Verbal Tags

Tag Use
-AOR aorist (perfective aspect)
-FUT future
-IMPF imperfective aspect
-IND marks optatives replacing indicatives in secondary sequence
-KJV marks optatives replacing subjunctives in secondary sequence
-PASS syntactic passive (not middle!)
-PRF perfect aspect

Syntactic Tags

The contents of the following table(s) are not exhausative but rather give the most common combinations of basic, “dash,” and extended tags.

Table 2.3. Basic Syntactic Tags

A
ADJP adjective phrase
ADJP-PRD adjective phrase, predicate
ADJP-SPR adjective phrase, secondary predicate
ADJX adjectival constituent, ambiguous projection (ADJ, ADJ', or ADJP)
ADJY adjectival non-constituent
ADVP adverb phrase
ADVP-DIR adverb phrase, directional
ADVP-LOC adverb phrase, locative
ADVP-TMP adverb phrase, temporal
ADVX adverbial constituent, ambiguous projection (ADV, ADV', or ADVP)
C
CONJP conjunction phrase
CP-ADV adverbial clause
CP-CAR clause-adjoined relative clause
CP-COM complement clause (non-indicative)
CP-CMP comparative clause
CP-DEG degree complement clause
CP-EOP empty operator clause
CP-EXL exclamation (direct or indirect)
CP-FRL free relative clause
CP-PRP non-infinitival purpose clause
CP-QUE question (direct or indirect)
CP-REL relative clause
CP-RES result clause
CP-THT complement clause (indicative)
F
FRAG sentence fragment
I
INTJP interjection phrase
IP-ABS absolute clause, including genitive absolute
IP-IMP imperative clause
IP-INF infinitival clause
IP-INF-ABS absolute infinitive
IP-INF-COM infinitival complement clause
IP-INF-PRP infinitival purpose clause
IP-INF-SBJ infinitival subject (or associate in null-expletive construction)
IP-MAT matrix (= main) clause
IP-PPL participial clause
IP-PPL-COM supplementary participle (occurs with certain verbs)
IP-SMC small clause
IP-SUB subordinate clause
IY inflectional non-constituent
N
NP noun phrase
NP-ADT noun phrase, adjunct
NP-ADV noun phrase, adverbial
NP-AGT noun phrase, agent in a passive
NP-ATR noun phrase, attributive and/or possessive
NP-COM noun phrase, complement
NP-DIR noun phrase, directional
NP-INS noun phrase, instrumental
NP-LOC noun phrase, locative
NP-LFD noun phrase, left-dislocated
NP-MSR noun phrase, measure
NP-OB1 noun phrase, direct object
NP-OB2 noun phrase, indirect object
NP-OBP noun phrase, object in a case determined by prepositional verbal prefix
NP-OBQ noun phrase, object in a quirky case
NP-PAR noun phrase, partitive
NP-PRD noun phrase, predicate
NP-PRN noun phrase, parenthetical and/or appositive
NP-RSP noun phrase, resumptive
NP-SBJ noun phrase, subject
NP-SPR noun phrase, secondary predicate
NP-TMP noun phrase, temporal
NP-VOC noun phrase, vocative
NUMP number phrase
NX nominal constituent, ambiguous projection (N, N', or NP)
NY nominal non-constituent
P
PP prepositional phrase
Q
QP quantifier phrase
QTP quotation phrase
QX quantificational constituent, ambiguous projection (Q, Q', or QP)
QY quantificational non-constituent
R
RRC reduced relative clause (attributive participle)
V
VP verb phrase
W
WADJP wh- adjective phrase
WADVP wh- adverb phrase
WNP wh- noun phrase
WPP wh- prepositional phrase
WQP wh- quantifier phrase

Table 2.4. Dash Tags

A
-ABS absolute
-ADT adjunct
-ADV adverbial
-AGT agent (in a passive)
-AOR aorist (perfective aspect)
-ATR attributive and/or possessive
C
-CAR clause-adjoined relative
-CMP comparative
-COM complement
D
-DEG degree complement clause
-DIR directional
E
-EOP empty-operator construction
-EXL exclamative
F
-FRL free relative -FUT future
I
-IMP imperative
-IMPF imperfective aspect
-IND optative replacing indicative in secondary sequence
-INF infinitive
-INS instrumental
-INTRNS intransitive
K
-KJV optative replacing subjunctive in secondary sequence
L
-LFD left-dislocated
-LOC locative
M
-MSR measure (extent of time/space)
-MAT matrix (= main) clause
O
-OB1 direct object
-OB2 indirect object
-OBP object in case determined by prepositional verbal prefix
-OBQ object in quirky case
P
-PAR partitive
-PASS syntactic passive (not middle!)
-POS possessive adjective
-PPL participle
-PRD predicate
-PRF perfect aspect
-PRN appositive and/or parenthetical
-PRP purpose
Q
-QUE question
R
-REL relative clause
-RES result clause
-RSP resumptive
S
-SBJ subject
+SLF reflexive (on pronouns)
-SMC small clause
-SPE direct speech
-SPR secondary predicate
-SUB subordinate clause
T
-THT indicative ὅτι-type complement clause or infinitive of indirect discourse
-TMP temporal
-TRNS1 (direct) transitive
-TRNS2 (abstract) transitive
V
-VOC vocative

Table 2.5. Extended Dash Tags

Suffix Tag Definition Example
-LFD left-dislocated constituent NP-LFD left-dislocated noun phrase
-PRN appositive and/or parenthetical NP-PRN appositive and/or parenthetical noun phrase
-RSP resumptive constituent NP-SBJ-RSP resumptive subject
-SPE direct speech (only on CP, IP) IP-MAT-SPE matrix clause, direct speech

Table 2.6. Empty Categories

Symbol Definition
* trace of A-movement; also default empty category
0 null element (e.g., null complementizer, empty operator)
*CL* trace of clitic displacement
*con* subject elided under conjunction
*ICH* abbreviation mnemonic for “interpret constituent here”; trace of extraposition, scrambling, or other movement that does not fit neatly into the A/A' dichotomy
*pro* “small/baby pro” (i.e., pro-dropped subject)
*T* trace of A'-movement (i.e., wh- movement)

Chapter 3. Splitting Words

Syntax for Splitting Words

There are two syntaxes for splitting words. The first applies only in cases of crasis, while the second is the default applying in every other case.

  • Crasis:

    • Components are split according to the logical division between words, with the leftmost component word containing all breathing marks. The leftmost component word ends with @, and the rightmost component word starts with @. In the case that three (or more) words are combined due to crasis, the middle word(s) would have @ both prefixed and suffixed to the word.

    • The lemma of each (sub-)word component in the case of crasis is just the lemma that would have appeared had the word not appeared in a phonological environment resulting in crasis.

      (CP-ADV-SPE (ADV κἂ@-καί)
                  (C @@-εἰ#1)
                  (IP-SUB-SPE (AN @ν-ἄν)
                              ...))
                                      

    • Elsewhere:

      • Components are split according to the logical division between component parts. The leftmost component ends with @, and the rightmost component begins with @. In the case that three (or more) parts are combined, the middle part(s) would have @ both prefixed and suffixed.

      • The lemma of the first component part in every non-crasis situation is just the lemma representing the combination of parts. The second (as well as third, or any additional) component parts take the same lemma prefixed and suffixed with @ as their lemma.

        (ADVP-TMP (NEG+CONJ Οὐδέ@-οὐδέποτε) (CLPRT @ποτε-@οὐδέποτε@))                             
                                                

Situations that always result in splitting

In the following cases, a compound word should always be broken down into its components and the appropriate lemma applied to each component word.

  • Crasis.

  • Any combination of another word with ἄν.

  • Clusters of CLPRT, e.g., μέντοι.

  • Clusters of adverbials with CLPRT, e.g., ὥσπερ.

  • Most combinations of an element with CLGE, e.g., ἐγώγε

  • ὅστις when case is doubly-marked (e.g., ὅστις and ὅντινα) but not when case is only singly-marked (e.g., ὅτις and ὅτου). In ambiguous cases such as ἥτις, the assumption is that case is, rather than is not, doubly-marked.

Situations that never result in splitting

In the following cases, a compound word should never be broken down into its components.

  • Clusters of CONJ and CLPRT

  • Compound negatives of any type. This rule takes precedence over other rules, such that something like e.g., μήγε (NEG+CLGE) or οὐκοῦν (NEG+CLPRT) does not get split.

  • Compound reflexives.

  • ὅστις when case is singly-marked (e.g., ὅτις and ὅτου) but not when case is only singly-marked (e.g., ὅστις and ὅντινα).

Treatment of Individual Words

Αα

ἀγγέλλω

can select for a participle of indirect discourse, IP-PPL-THT

ἄγε, ἄγετε

tagged INTJ if not main verb imperative, VBI otherwise

αἰσθάνομαι

can select for a participle of indirect discourse, IP-PPL-THT

ἀκούω

can select for a participle of indirect discourse, IP-PPL-THT

can take a quirky genitive object, NP-OBQ

ἀλλά

tagged CONJ

ἄλλος

tagged OTHER

ἀμήν

tagged INTJ

ἀμφότεροι

tagged Q

ἄν

tagged AN

ἅπας

tagged Q

ἄρα

tagged CLPRT

ἆρα

tagged PRTQ

ἄρχω

can take a supplementary participle, IP-PPL-COM

αὐτός

  • When αὐτός behaves as a third-person singular pronoun, it is tagged PRO.

    ( (IP-MAT-SPE (CODE {VS:1_21})
                  (NP-SBJ (PRO αὐτὸς))
                  (CLPRT γὰρ)
                  (VBP-FUT σώσει)
                  (NP-OB1 (DA τὸν)
                          (NA λαὸν)
                          (NP-ATR (PRO$ αὐτοῦ)))
                  (PP (P ἀπὸ)
                      (NP (DS$ τῶν)
                          (NS$ ἁμαρτιῶν)
                          (NP-ATR (PRO$ αὐτῶν)))))
      (. .)
      (ID GreekNT,Matthew))                                    
                                    

  • When αὐτός appears in attributive position and means ‘same,’ it is tagged ADJ.

    ( (IP-MAT-SPE (CODE {VS:1_12})
                  (NP-SBJ (PRO σὺ))
                  (CLPRT δὲ)
                  (NP-PRD (D ὁ) (ADJ αὐτὸς))
                  (BEP-IMPF εἶ)
                  (. ,))
      (ID GreekNT,Hebrews))                                    
                                    

  • When αὐτός behaves as a non-argument reflexive, it is tagged PRO, and it is syntactically analyzed as an NP-PRN.

    ( (IP-MAT (CODE {VS:4_44})
              (NP-SBJ (NP-PRN (PRO αὐτὸς))
                      (CLPRT γὰρ)
                      (NPR Ἰησοῦς))
              (VBD-AOR ἐμαρτύρησεν)
              (CP-THT (C ὅτι)
                      (IP-SUB (NP-SBJ (N προφήτης))
                              (PP (P ἐν)
                                  (NP (DD τῇ) (ADJD ἰδίᾳ) (ND πατρίδι)))
                              (NP-OB1 (NA τιμὴν))
                              (NEG οὐκ)
                              (VBP-IMPF ἔχει)))
              (. .))
      (ID GreekNT,John))                                    
                                    

    When αὐτός behaves as a non-argument reflexive to a pro-dropped subject, the following structure (without an overt *pro* element) is used:

    ( (IP-MAT-SPE (CODE {VS:23_4})
                  (NP-SBJ (NP-PRN (DS αὐτοὶ-αὐτός)))
                  (CLPRT δὲ-δέ)
                  (NP-1 (DD τῷ-ὁ)
                        (ND δακτύλῳ-δάκτυλος)
                        (NP-ATR (PRO$ αὐτῶν-αὐτός)))
                  (NEG οὐ-οὐ)
                  (VBP-IMPF θέλουσιν-θέλω)
                  (IP-INF-COM (NP-INS *ICH*-1)
                              (VBN-AOR κινῆσαι-κινέω)
                              (NP-OB1 (PROA αὐτά-αὐτός)))
                  (. .))
      (ID GreekNT,Matthew))                                    
                                    

Γγ

γάρ

tagged CLPRT

γε

tagged CLGE

γιγνώσκω

can select for a participle of indirect discourse, IP-PPL-THT

Δδ

δέ

tagged CLPRT

δείκνυμι

can select for a participle of indirect discourse, IP-PPL-THT

δή

tagged CLPRT

δηλόω

can select for a participle of indirect discourse, IP-PPL-THT

δίς

tagged NUM

Εε

*ἑαυτός

tagged PRO*+SLF; no nominative

never split into component words; see splitting words for details

tagged C when translated 'if'

tagged WQ when translated 'whether' and occurs in an indirect question

tagged PRTQ when not translated in a direct question

given the lemma εἰ#1 when in combination with μή to mean 'except'

εἰμί

the verb ‘to be’; receives special set of verbal tags beginning with BE*

εἴτε

tagged CONJ

can be correlative: εἴτε...εἴτε

ἕκαστος

tagged Q

ἐκεῖνος

tagged D

ἔξω

Ηη

  • tagged CONJ, C, or P depending on the syntactic context

  • as CONJ when it means ‘or’

    ( (IP-IMP-SPE (CODE {VS:6_31})
                  (NEG μὴ)
                  (CLPRT οὖν)
                  (VBS-AOR μεριμνήσητε)
                  (IP-PPL (VPR-IMPF λέγοντες)
                          (CP-QUE-SPE (CP-QUE-SPE (WNP-1 (WPROA Τί))
                                                  (C 0)
                                                  (IP-SUB (NP-OB1 *T*-1)
                                                          (NP-SBJ *pro*)
                                                          (VBS-AOR φάγωμεν)))
                                      (. ;)
                                      (CONJP (CONJ ἤ)
                                             (CP-QUE-SPE (WNP-2 (WPROA Τί))
                                                         (C 0)
                                                         (IP-SUB (NP-OB1 *T*-2)
                                                                 (NP-SBJ *con*)
                                                                 (VBS-AOR πίωμεν))))
                                      (. ;)
                                      (CONJP (CONJ ἤ)
                                             (CP-QUE-SPE (WNP-3 (WPROA Τί))
                                                         (C 0)
                                                         (IP-SUB (NP-OB1 *T*-3)
                                                                 (NP-SBJ *con*)
                                                                 (VBSP-AOR περιβαλώμεθα))))))
                  (. ;))
      (ID GreekNT,Matthew))
                                
  • as CONJ when it occurs as a correlative conjunction: ἤ...ἤ

  • as C when it means ‘than’ and introduces a full comparative clause (see the section called “Comparative Clauses”)

    ( (IP-MAT (CODE {VS:7_9})
              (ADJP-PRD (ADJR κρεῖττον)
                        (CP-CMP *ICH*-1))
              (CLPRT γάρ)
              (BEP-IMPF ἐστιν)
              (IP-INF-SBJ (VBN-IMPF γαμεῖν))
              (CP-CMP-1 (WADJP-2 0)
                        (C ἢ)
                        (IP-SUB (ADJP-PRD *T*-2)
                                (BEP-IMPF *)
                                (IP-INF-SBJ (VBNP-IMPF πυροῦσθαι))))
              (. .))
      (ID GreekNT,1Corinthians))                                    
                                    

  • as P when it means ‘than’ and occurs in a comparative construction that does not appear to contain an elided clause (see also ὡς)

    ( (IP-MAT-SPE (CODE {VS:17_2})
                  (NP-SBJ *pro*)
                  (VBP-IMPF λυσιτελεῖ)
                  (NP-OBQ (PROD αὐτῷ))
                  (CP-ADV (C εἰ)
                          (IP-SUB (IP-SUB (NP-SBJ (N λίθος) (ADJ μυλικὸς))
                                          (VBPP-IMPF περίκειται)
                                          (PP (P περὶ)
                                              (NP (DA τὸν)
                                                  (NA τράχηλον)
                                                  (NP-ATR (PRO$ αὐτοῦ)))))
                                  (CONJP (CONJ καὶ)
                                         (IP-SUB (NP-SBJ *pro*)
                                                 (VBPP-PRF ἔρριπται)
                                                 (PP (P εἰς)
                                                     (NP (DA τὴν) (NA θάλασσαν)))))))
                  (PP (P ἢ)
                      (CP-ADV (C ἵνα)
                              (IP-SUB (NP-SBJ *pro*)
                                      (VBS-AOR σκανδαλίσῃ)
                                      (NP-OB1 (NP-PAR (DS$ τῶν) (ADJ$ μικρῶν) (DS$ τούτων))
                                              (NUM ἕνα)))))
                  (. .))
      (ID GreekNT,Luke))                                    
                                    

tagged CLPRT; also sometimes ἤ, as in ἤτοι

Ιι

ἰδού

tagged INTJ unless actually a full (= main) imperative

ἴθι

tagged INTJ if not main verb imperative, VBI otherwise

ἱκανός

  • tagged ADJR

  • can take an IP-INF-COM (epexegetical infinitive)

  • modifying a noun:

    ( (IP-MAT (CODE {VS:20_37})
              (NP-SBJ (ADJR ἱκανὸς)
                      (CLPRT δὲ)
                      (N κλαυθμὸς)
                      (NP-ATR *ICH*-2))
              (VBDP-AOR ἐγένετο)
              (NP-2 (Q πάντων))
              (. ,))
      (ID GreekNT,Acts))                                    
                                    

  • taking an IP-INF-COM:

    ( (IP-MAT (CODE {VS:3_16})
              (VBPP-IMPF ἔρχεται)
              (CLPRT δὲ)
              (NP-SBJ (D ὁ)
                      (ADJR ἰσχυρότερός)
                      (NP-CMP (CLPRO$ μου))
                      (, ,)
                      (CP-REL (WNP-1 (WPRO$ οὗ))
                              (C 0)
                              (IP-SUB (NEG οὐκ)
                                      (NP-SBJ *pro*)
                                      (BEP-IMPF εἰμὶ)
                                      (ADJP-PRD (ADJR ἱκανὸς)
                                                (IP-INF-COM (VBN-AOR λῦσαι)
                                                            (NP-OB1 (DA τὸν)
                                                                    (NA ἱμάντα)
                                                                    (NP-ATR (DS$ τῶν)
                                                                            (NS$ ὑποδημάτων)
                                                                            (NP-ATR-RSP-1 (PRO$ αὐτοῦ)))))))))
              (. .))
      (ID GreekNT,Luke))                                    
                                    

ἵνα

tagged C

ἵνατί

tagged C+CLQA; never split into component parts

Κκ

κα

tagged KE

κἀγώ

crasis for καὶ ἐγώ; split into two words (see splitting words for details) suffixed and prefixed with @:

(NODE (CONJ κἀ@-καί)
      (PRO @γώ-ἐγώ))
                    

κε(ν)

tagged KE

καί

  • καί is generally tagged CONJ

    (VBN-IMPF (VBN-IMPF ποιεῖν) (CLTE τε) (CONJ καὶ) (VBN-IMPF διδάσκειν))
                                    

    ...ὧν ἤρξατο Ἰησοῦς ποιεῖν τε καὶ διδάσκειν (Acts 1.1)

  • καί is also tagged CONJ when it occurs as part of the correlative conjunctionκαί...καί

  • καί is tagged ADV when it is used adverbially, at the phrase or sentence level

    • phrase-level:

      ( (IP-MAT-SPE (CODE {VS:2_26})
                    (ADVP-TMP (ADV ἔτι))
                    (CLPRT δὲ)
                    (NP-SBJ (ADV καὶ)
                            (D ἡ)
                            (N σάρξ)
                            (NP-ATR (CLPRO$ μου)))
                    (VBP-FUT κατασκηνώσει)
                    (PP (P ἐπ')
                        (NP (ND ἐλπίδι)))
                    (. .))
        (ID GreekNT,Acts))                                    
                                      
    • sentence-level (Acts 1.3):

      (CP-REL-6 (WNP-7 (WPROD οἷς))
                (C 0)
                (IP-SUB (NP-OBP *T*-7)
                        (ADVP (ADV καὶ))
                        (NP-SBJ *pro*)
                        (VBD-AOR παρέστησεν)
                        (IP-SMC (NP-SBJ (PROA+SLF ἑαυτὸν))
                                (VPRA-IMPF ζῶντα)
                                (PP (P μετὰ)
                                    (NP (D τὸ)
                                        (IP-INF (VBN-AOR παθεῖν)
                                                (NP-SBJ (PROA αὐτὸν))
                                                (PP (P ἐν)
                                                    (NP (QD πολλοῖς) (NSD τεκμηρίοις)))))                                            
                                              

καίτοι

tagged CONJ+CLPRT

Λλ

λαός

an example of a collective noun; collective nouns are tagged according to their morphological number, even if they induce plural agreement

λανθάνω

can take a supplementary participle, IP-PPL-COM

λίαν

tagged ADVR when introducing degree complement, otherwise ADV

Μμ

μάλιστα

tagged QS

μᾶλλον

tagged QR

μανθάνω

can select for a participle of indirect discourse, IP-PPL-THT

με

tagged CLPROA

μέν

tagged CLPRT

μή

usually tagged NEG, but tagged C when could be translated lest and introducing a subordinate clause—e.g., a fear clause

treated as the sister of εἰ when in a construction where εἰ + μή = 'except'

μηδαμός

tagged NEG+ADJ

μηδαμῶς

tagged NEG+ADV

μηδέ

tagged NEG+CONJ

μηδείς

tagged NEG+Q

μηδέτερος

tagged NEG+Q

μήν

tagged CLPRT

μήποτε

tagged NEG+CLPRT

μήτε

tagged NEG+CLTE

can be correlative: μήτε...μήτε

μήτι

tagged PRTQ

μιν

tagged CLPROA

μόνος

  • μόνος is most often tagged as an adjective

    (ADJP-PRD (ADJ μόνος))                        
                        

    καὶ οὐκ εἰμὶ μόνος, ὅτι ὁ πατὴρ μετ᾽ ἐμοῦ ἐστίν. (John 16.32)

  • μόνος is still tagged as an adjective when it is used for focus if it agrees with the noun in modifies

    (NP (ND ἄρτῳ) (ADV μόνῳ))                                    
                                    

    Οὐκ ἐπ᾽ ἄρτῳ μόνῳ ζήσεται ὁ ἄνθρωπος (Matthew 4.4)

  • when μόνος does not agree in case and/or number with the noun it modifies and it's used for focus, it is tagged ADV

    (NP-OB1 (DSA τοὺς-ὁ)
            (NSA ἀδελφοὺς-ἀδελφός)
            (NP-ATR (PRO$ ὑμῶν-ὑμεῖς))
            (ADV μόνον-μόνος))))                                    
                                    

    καὶ ἐὰν ἀσπάσησθε τοὺς ἀδελφοὺς ὑμῶν μόνον, τί περισσὸν ποιεῖτε; (Matthew 5.47)

  • when μόνος serves as an adverbial modifier of a verb, it is tagged ADV

    ( (IP-IMP-SPE (CODE {VS:8_8a})
                  (CONJ ἀλλὰ-ἀλλά)
                  (ADVP (ADV μόνον-μόνος))
                  (VBI-AOR εἰπὲ-λέγω)
                  (NP-ADV (ND λόγῳ-λόγος))
                  (, ,))
      (ID GreekNT,Matthew))                                    
                                    

μου

tagged CLPRO$

Νν

ναί

tagged INTJ

νυν

tagged CLPRT

Οο

tagged D

ὅδε

tagged D

ὀλίγος

tagged Q

ὅλος

usually tagged Q, but tagged ADJ when in attributive position

ὁμοιόω

can take an IP-SMC

ὁπότερος

tagged WPRO since ‘which of two’

ὁράω

can select for a participle of indirect discourse, IP-PPL-THT

ὅς

tagged WD or WPRO depending on whether it is modifying a noun or serving as the head of the WNP, respectively

ὅσος

tagged WADJ

ὅστις

tagged WPRO+CLQ

split into two words when case is doubly-marked but not when case is singly-marked; see splitting words for details

ὅταν

split into components and tagged C, AN respectively; see splitting words for details

ὅτε

  • when introducing adverbial clause without a gap, tagged C

    ( (IP-MAT (CODE {VS:27_31})
              (CONJ καὶ)
              (CP-ADV (C ὅτε)
                      (IP-SUB (NP-SBJ *pro*)
                              (VBD-AOR ἐνέπαιξαν)
                              (NP-OBP (PROD αὐτῷ))))
              (, ,)
              (VBD-AOR ἐξέδυσαν)
              (NP-OB2 (PROA αὐτὸν))
              (NP-OB1 (DA τὴν) (NA χλαμύδα)))
      (ID GreekNT,Matthew))                                    
                                    

  • when introducing a question or relative clause, tagged WADV

    ( (IP-MAT-SPE (CODE {VS:9_4})
                  (VBPP-IMPF ἔρχεται)
                  (NP-SBJ (N νὺξ)
                          (CP-REL (WADVP-1 (WADV ὅτε))
                                  (C 0)
                                  (IP-SUB (ADVP *T*-1)
                                          (NP-SBJ (NEG+Q οὐδεὶς))
                                          (VBPP-IMPF δύναται)
                                          (IP-INF-COM (VBNP-IMPF ἐργάζεσθαι)))))
                  (. .))
      (ID GreekNT,John))                                    
                                    

ὅτι

tagged C

οὐαί

tagged INTJ

οὐδαμός

tagged NEG+ADJ

οὐδέ

tagged NEG+CONJ

οὐδείς

tagged NEG+Q

οὐδέτερος

tagged NEG+Q

οὐκέτι

tagged NEG+ADV

οὐκοῦν

usually tagged NEG+CLPRT, but tagged as PRTQ in a question

οὖν

tagged CLPRT

οὔποτε

tagged NEG+CLPRT

οὔτε

tagged NEG+CLTE

can be correlative: οὔτε...οὔτε

οὗτος

tagged D

οὕτως

tagged QR

οὐχί

tagged PRTQ

Ππ

πάλιν

(ADVP (ADV πάλιν))                        
                    

does not get a -TMP function tag

παπαῖ

tagged INTJ

πάντως

tagged QV

πᾶς

usually tagged Q but tagged ADJ when in attributive position

παύω

can take a supplementary participle, IP-PPL-COM

περ

tagged CLPRT

πλεῖστος

tagged QS

πόθεν

tagged WADV

ποῖος

tagged WD or WPRO depending on whether it is modifying a noun or serving as the head of the WNP, respectively

πολύς

usually tagged Q but tagged ADJ when in attributive position

πόσος

tagged WADJ

ποτε

tagged CLPRT

πότε

tagged WADV

ποῦ

tagged WADV

πρῶτος

  • when an ordinal number, tagged ADJ

  • when adverbial (in the neuter singular or plural: πρῶτον, πρῶτα) tagged ADV

πω

tagged CLPRT

πῶς

tagged WADV

Σς

σε

tagged CLPROA

σου

tagged CLPRO$

σφεας

tagged CLPROA

σφι

tagged CLPROD

σύμπας

tagged Q

Ττ

τε

tagged CLTE

τίκτω

can take an NP-SPR

τις/τι

clitic indefinite pronoun quantifier; tagged CLQ*

τίς

tagged WD or WPRO depending on whether it is modifying a noun or serving as the head of the WNP, respectively

τοι

tagged CLPRT

τοίνυν

split into components and tagged CLPRT, CLPRT respectively; see splitting words for details

τοσοῦτος

tagged Q

τυγχάνω

can take a supplementary participle, IP-PPL-COM

Φφ

φέρε

tagged INTJ if not main verb imperative, VBI otherwise

φθάνω

can take a supplementary participle, IP-PPL-COM

Χχ

χαίρω

can take a supplementary participle, IP-PPL-COM

χάριν

usually tagged N, but tagged P when takes a dependent NP (may take a pre-head dependent NP)

χίλιοι

tagged ADJ, like all cardinal numbers that exist only in adjectival form (see the section called “Cardinal Numbers (NUM)”)

Ωω

tagged INTJ

ὡς

  • tagged C, WADV, ADV, or P depending on the context

  • C in its temporal use meaning ‘when’ or when used as a complementizer introducing any other type of subordinate clause such as a that-clause (see the section called “That-Clauses”) or a purpose clause (see the section called “Complement Clauses”)

    ( (IP-MAT (CODE {VS:10_28})
              (NP-SBJ *pro*)
              (VBD-IMPF ἔφη)
              (CLTE τε)
              (PP (P πρὸς)
                  (NP (PROA αὐτούς)))
              (IP-MAT-SPE (NP-SBJ (PRO Ὑμεῖς))
                          (VBPP-IMPF ἐπίστασθε)
                          (CP-THT (C ὡς)
                                  (IP-SUB (ADJP-PRD (ADJ ἀθέμιτόν))
                                          (BEP-IMPF ἐστιν)
                                          (NP-OB2 (ND ἀνδρὶ) (ADJD Ἰουδαίῳ))
                                          (IP-INF-SBJ (VBNP-IMPF (VBNP-IMPF κολλᾶσθαι) (CONJ ἢ) (VBNP-IMPF προσέρχεσθαι))
                                                      (NP-OBQ (ADJD ἀλλοφύλῳ))))))
              (. .))
      (ID GreekNT,Acts))                                    
                                    

  • WADV in adverbial and comparative clauses (see the section called “Adverbial Clauses” and the section called “Comparative Clauses” for details)

    ( (IP-MAT (CODE {VS:1_14})
              (CONJ καὶ)
              (NP-SBJ (DS οἱ)
                      (NS ὀφθαλμοὶ)
                      (NP-ATR (PRO$ αὐτοῦ)))
              (BEP-IMPF *)
              (ADVP (CP-CMP (WADVP-3 (WADV ὡς))
                            (C 0)
                            (IP-SUB (ADVP *T*-3)
                                    (BEP-IMPF *)
                                    (NP-SBJ (N φλὸξ)
                                            (NP-ATR (N$ πυρός))))))
              (. ,))
      (ID GreekNT,Revelation))                                    
                                    

  • ADV when it has the meaning of ‘so’ or ‘thus’ and does not introduce a subordinate clause

  • P in comparatives when there does not appear to be an elided clause (see as well)

    (NP-PRN (D ὁ)
            (RRC (VPR-IMPF ἔχων)
                 (IP-SMC (NP-SBJ (DSA τοὺς)
                                 (NSA ὀφθαλμοὺς)
                                 (NP-ATR (PRO$ αὐτοῦ)))
                         (PP (P ὡς)
                             (NP (NA φλόγα)
                                 (NP-ATR (N$ πυρός)))))))                                    
                                    

    (Revelation 2.18)

ὡσαννά

tagged INTJ

ὥστε

tagged C+CLTE; never split into component parts

Chapter 4. POS Annotation

Adjectives and Adverbs (ADJ, ADV)

Table 4.1. ADJ and ADV Tags

Tag Use
ADJ adjective, nominative
ADJ$ adjective, genitive
ADJA adjective, accusative
ADJD adjective, dative
ADJR adjective, comparative degree
ADJS adjective, superlative degree
ADV adverb
ADVR adverb, comparative degree
ADVS adverb, superlative degree
WADJ wh-adjective, nominative
WADJ$ wh-adjective, genitive
WADJA wh-adjective, accusative
WADJD wh-adjective, dative
WADV wh-adverb

Notes

  • No distinction is made between singular and plural adjectives in the positive degree.

  • In the comparative and superlative degrees, no distinctions are made with respect to the case or number of adjectives.

  • Possessive adjectives receive the “dash” tag -POS.

Colors

  • The basic tag for basic color terms is ADJ.

  • If a color term is also a substantive noun (e.g., orange in English), then it may be tagged either ADJ or N depending on the syntactic context.

  • Color terms that neither modify a noun nor function as the predicate of a clause are tagged N. For example, in English:

    • the white/N of an egg

    • a black/N like that of jet

      a deep red/N

Degree Words

Degree words (e.g., οὕτως, ἱκανός) are treated as comparatives.

Ordinal Numbers

  • Ordinal numbers are tagged ADJ.

  • πρῶτος also has an adverbial use. See πρῶτος for details.

  • Cases where an ordinal might be expected but without overt ordinal marking are treated as cardinal numbers (NUM).

Cardinal Numbers (NUM)

Table 4.2. NUM Tags

Tag Use
NUM cardinal number

Notes

  • All cardinal numbers are tagged NUM.

  • Cardinal numbers that only exist in adjectival form (e.g., χίλιοι) are tagged ADJ.

  • Greek equivalents of words such as double, treble, triple, etc., twice (see δίς, thrice, and once when analogous in meaning to twice, thrice are always tagged NUM.

    (ADVP (NUM δὶς))                            
                            

    (Mark 14:30)

  • Words such as double and score are treated as cardinal numbers, on a par with hundred and thousand, and tagged NUM. There may or may not be Greek parallels to this type of case.

Clitics (CL)

Table 4.3. CL Tags

Tag Use
AN modal particle ἄν
CLGE emphatic clitic particle γε
CLPRO clitic pronoun, nominative (Homer only)
CLPRO$ clitic pronoun, genitive
CLPROA clitic pronoun, accusative
CLPROD clitic pronoun, dative
CLPRT second-position sentential particle, other second position particles such as περ ‘indeed’
CLQ indefinite clitic pronoun/quantifier τις/τι, nominative
CLQ$ indefinite clitic pronoun/quantifier τις/τι, genitive
CLQA indefinite clitic pronoun/quantifier τις/τι, accusative
CLQD indefinite clitic pronoun/quantifier τις/τι, dative
CLTE conjunctive clitic particle τε
KE modal particle κε(ν)/κα

Notes

  • The clitic particles γε and τε both have their own tags: CLGE and CLTE, respectively.

  • The modal particles κε(ν)/κα and ἄν also have their own tags: KE and AN, respectively.

  • In the singular and third person plural of the oblique cases (not the genitive in the case of the third plural), the clitic forms of the personal pronouns—e.g., με, σε, μιν, σφι, σφεας, μου, σου (not a complete list!)—are tagged CLPRO\$, CLPROA, CLPROD as appropriate.

  • The second-position sentential particles receive the tag CLPRT. These are: δέ, δή, πω, οὖν, γάρ, μέν, νυν, τοι, ἄρα, , μήν. Other second-position (but not necessarily sentential) particles are: περ, ποτε

  • Particles in combination with other particles or other items receive multiple tags tags separated with a hyphen. The forms themselves are also hyphenated to indicate the segmentation. For example, καί-τοι receives the tag CONJ-CLPRT and τοί-νυν receives the tag CLPRT-CLPRT.

Complementizers (C)

Table 4.4. C Tags

Tag Use
C complementizer

Notes

  • ὅτι, ἵνα, ὅτε, etc. introducing any kind of subordinate clause are tagged C.

  • ὡς can be tagged C, WADV, ADV, or P depending on the syntactic context.

  • Likewise, can be tagged CONJ, C, or P depending on the syntactic context.

Coordinating Conjunctions (CONJ, CLTE)

Table 4.5. CONJ and CLTE Tags

Tag Use
CONJ conjunction, including correlative conjunctions
CONJ0 conjunction, clause-level and token-initial
NEG+CONJ compound negative conjunctions
CLTE clitic conjunctive particle τε

Notes

Determiners and Demonstratives (D)

Table 4.6. D Tags

Tag Use
D determiner or demonstrative, nominative singular
D$ determiner or demonstrative, genitive singular
DA determiner or demonstrative, accusative singular
DD determiner or demonstrative, dative singular
DS determiner or demonstrative, nominative plural
DS$ determiner or demonstrative, genitive plural
DSA determiner or demonstrative, accusative plural
DSD determiner or demonstrative, dative plural

Notes

  • The following lemmas are tagged with D: , οὗτος, ἐκεῖνος, ὅδε.

  • Demonstratives are always tagged D, regardless of whether they precede a noun. Note the difference between ordinary determiners and wh-words (see the section called “Wh- Words (W)” for details) in this regard.

    • demonstrative not preceding noun:

      ( (IP-MAT-SPE (CODE {VS:7_12})
                    (NP-SBJ (D οὗτος))
                    (CLPRT γάρ)
                    (BEP-IMPF ἐστιν)
                    (NP-PRD (NP (D ὁ) (N νόμος))
                            (CONJP (CONJ καὶ)
                                   (NP (DS οἱ) (NS προφῆται))))
                    (. .))
        (ID GreekNT,Matthew))                                    
                                      

    • wh-determiner preceding a noun:

      ( (IP-MAT (CODE {VS:4_2})
                (NP-SBJ *pro*)
                (VBP-PRF οἴδατε)
                (CLPRT γὰρ)
                (CP-QUE (WNP-1 (WDA τίνας) (NSA παραγγελίας))
                        (C 0)
                        (IP-SUB (NP-OB1 *T*-1)
                                (NP-SBJ *pro*)
                                (VBD-AOR ἐδώκαμεν)
                                (NP-OB2 (PROD ὑμῖν))
                                (PP (P διὰ)
                                    (NP (D$ τοῦ)
                                        (N$ κυρίου)
                                        (NP-PRN (NPR$ Ἰησοῦ))))))
                (. .))
        (ID GreekNT,1Thessalonians))                                    
                                      

    • wh-determiner not preceding a noun:

      ( (IP-IMP-SPE (CODE {VS:10_27})
                    (NP-OB1 (CP-FRL (WNP-1 (WPROA ὃ))
                                    (C 0)
                                    (IP-SUB (NP-OB1 *T*-1)
                                            (NP-SBJ *pro*)
                                            (VBP-IMPF λέγω)
                                            (NP-OB2 (PROD ὑμῖν))
                                            (PP (P ἐν)
                                                (NP (DD τῇ) (ND σκοτίᾳ))))))
                    (, ,)
                    (VBI-AOR εἴπατε)
                    (PP (P ἐν)
                        (NP (DD τῷ) (ND φωτί)))
                    (. .))
        (ID GreekNT,Matthew))                                    
                                      

Foreign Words (FW)

Table 4.7. FW Tags

Tag Use
FW foreign word

Notes

  • If a potential foreign word has an entry in the LSJ[1], it is not tagged as FW. This is true even for nouns that retain a foreign plural.

  • Foreign names are treated as nouns.

  • Titles (e.g., of literary works) in foreign languages are generally tagged as FW.

  • In foreign language sequences, everything (words, symbols, numbers, etc.) except punctuation is labeled FW.

Interjections (INTJ)

Table 4.8. INTJ Tags

Tag Use
INTJ interjection

Notes

  • INTJ is only used to tag words that are difficult or impossible to tag any other way, like the following: , παπαῖ, ναί, οὐαί, ἀμήν, ἰδού, ὡσαννά (not a complete list).

    ( (IP-MAT-SPE (CODE {VS:6_2})
                  (INTJ ἀμὴν)
                  (NP-SBJ *pro*)
                  (VBP-IMPF λέγω)
                  (NP-OB2 (PROD ὑμῖν))
                  (, ,)
                  (IP-MAT-SPE (NP-SBJ *pro*)
                              (VBP-IMPF ἀπέχουσιν)
                              (NP-OB1 (DA τὸν)
                                      (NA μισθὸν)
                                      (NP-ATR (PRO$ αὐτῶν))))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

  • When preceding another imperative verb, the following are also tagged as INTJ: ἄγε, ἄγετε, φέρε, ἴθι.

Negation (NEG)

Table 4.9. NEG Tags

Tag Use
NEG negation

Notes

  • The particles οὐ and μή are tagged NEG.

Compound NEG Tags

Nominals (N, PRO, CLPRO)

Table 4.10. Nominal Tags

Tag Use
CLPRO$ clitic pronoun, genitive
CLPROA clitic pronoun, accusative
CLPROD clitic pronoun, dative
CLQ indefinite clitic pronoun/quantifier τις/τι, nominative
CLQ$ indefinite clitic pronoun/quantifier τις/τι, genitive
CLQA indefinite clitic pronoun/quantifier τις/τι, accusative
CLQD indefinite clitic pronoun/quantifier τις/τι, dative
N common noun, nominative singular
N$ common noun, genitive singular
NA common noun, accusative singular
ND common noun, dative singular
NPR proper noun, nominative singular
NPR$ proper noun, genitive singular
NPRA proper noun, accusative singular
NPRD proper noun, dative singular
NPRS proper noun, nominative plural
NPRS$ proper noun, genitive plural
NPRSA proper noun, accusative plural
NPRSD proper noun, dative plural
NS common noun, nominative plural
NS$ common noun, genitive plural
NSA common noun, accusative plural
NSD common noun, dative plural
PRO personal pronoun, nominative
PRO$ personal pronoun, genitive
PROA personal pronoun, accusative
PROD personal pronoun, dative
WPRO wh-pronoun, nominative
WPRO$ wh-pronoun, genitive
WPROA wh-pronoun, accusative
WPROD wh-pronoun, dative

Notes: Nominals

  • Foreign names are tagged as nouns (N) rather than as foreign words (FW).

    εἰ/C δὲ/CLPRT ἐγὼ/PRO ἐν/P Βεεζεβοὺλ/ND ἐκβάλλω/VBP-IMPF τὰ/DSA δαιμόνια/NSA                            
                            

    (Luke 11.19)

  • Formally singular count nouns are tagged as singular (N), even when construed with a plural verb. Likewise, formally plural count nouns are tagged as plural (NS), even when construed with a singular verb (e.g., neuter plurals in Classical Greek).

  • Collective nouns such as λαός are tagged according to their morphological number (i.e., N for singular in this case), even if they are construed with a plural verb.

  • In general, forms with an overt adjectival suffix are tagged ADJ even if they are used as substantives.

    καὶ/CONJ ἀφοριοῦσιν/VBP-FUT τοὺς/DSA πονηροὺς/ADJA ἐκ/P μέσου/N$ τῶν/DS$ δικαίων/ADJ$                            
                            

    (Matthew 13.49)

  • Units of measure after numbers are tagged as singular or plural depending on overt number marking.

  • Common nouns standing in a genitive relationship with other nouns are tagged N$, NS$.

  • Words referring to groups of people (ethnic, ideological, or religious) are handled as follows: if the word has a morphologically adjectival form, it is tagged ADJ.

    • adjectival form:

                 τῶν/DS$ Ἰουδαίων/ADJ$                     
                                  
    • nominal form:

                 οἱ/DS Φαρισαῖοι/NS                         
                                      

Notes: Pronominals

  • All pronouns are tagged PRO except the clitic pronouns (CLPRO) and the indefinite clitic pronoun τις/τι (CLQ).

  • Reflexive pronouns are tagged with compound tags consisting of a PRO* tag followed by the dash tag -SLF.

  • Reciprocal pronouns will be identifiable by lemma in the final version of the corpus.

  • When αὐτός is used as a non-argument reflexive, it is still tagged PRO. See αὐτός for details.

Particles (CLPRT, AN, KE, PRTQ, CLGE)

Table 4.11. Particle Tags

Tag Use
AN modal particle ἄν
CLGE emphatic clitic particle γε
CLPRT second-position sentential particle or other second-position particle
KE modal particle κε(ν)/κα
PRTQ interrogative particle (e.g., ἆρα)

Notes: Modal Particles

  • The modal particle ἄν is tagged AN. Its dialectal variants κε(ν) and κα are tagged KE.

  • Words consisting of a complementizer plus the modal particle ἄν—e.g.,ὅταν—are tagged with the compound tag C+AN.

Notes: PRTQ

  • The interrogative particle ἆρα is always tagged PRTQ.

  • Other particles such as οὐχί, μήτι, and οὐκοῦν are only tagged PRTQ when they appear in questions and do not have any of their negative force.

  • When εἰ appears in a direct question (and doesn't get translated at all), it is tagged PRTQ instead of WQ.

Prepositions (P)

Table 4.12. P Tags

Tag Use
P preposition
WP preposition in a wh- prepositional phrase

Notes

  • Prepositions are tagged P.

  • Words traditionally analyzed as adverbs—e.g., ἔξω—that take a nominal complement (often in the genitive) are tagged P when they appear with such complements, ADV otherwise.

    (NP-SBJ *pro*)
    (VBD-AOR ἐξῆλθεν)
    (PP (P ἔξω)
        (NP (D$ τῆς) (N$ πόλεως)))
    (PP (P εἰς)
        (NP (NPRA Βηθανίαν)))                            
                            

    (Matthew 21.17)

  • Prepositions in a wh- prepositional phrase (WPP) are tagged WP to facilitate the creation of the correct type of trace for these in the emacs parser-mode.

Punctuation

Table 4.13. Punctuation Tags

Tag Use
, sentence-internal punctuation
. sentence-final punctuation
" quotation mark

Notes

  • Final punctuation indicates the end of a token. Note, however, that tokens can terminate without any punctuation.

  • Question marks (Greek ;) are tagged as sentence-final punctuation (usually) or sentence-internal punctuation depending on their relationships to the end of the token.

  • Periods preceding sentence-final quotes are tagged as sentence-final punctuation. The quotes are tagged as themselves.

  • All other sentence punctuation that does not coincide with the end of a token is tagged with a comma.

  • Interpuncts in the original text are represented by interpuncts in the annotated corpus. By default, they are tagged with ‘.’ but their tags should be changed to ‘,’ when they do not serve as sentence-final punctuation. (Note that an interpunct preceding a sentence final quote would count as sentence-final punctuation.)

  • Single and double quotation marks are tagged as double quotes. Directionality is not preserved.

  • Colons are replaced by periods by default, even when sentence-internal, due to interference with the colon in certain CODE elements, including chapter and verse markers.

Quantifiers (Q)

Table 4.14. Q Tags

Tag Use
Q quantifier, nominative
Q$ quantifier, genitive
QA quantifier, accusative
QD quantifier, dative
QR quantifier, comparative
QS quantifier, superlative
QV adverbial quantifier

Notes

Subordinating Conjunctions

See the section called “Complementizers (C)”.

Verbs (VB*, BE*)

Table 4.15. VB* and BE* Tags

Tag Use
BED εἰμί, secondary sequence
BEI εἰμί, imperative
BEN εἰμί, infinitive
BEO εἰμί, optative
BEP εἰμί, primary sequence
BES εἰμί, subjunctive
BPR εἰμί, participle
VBD verb, active secondary sequence (imperfect, aorist, pluperfect)
VBDP verb, middle/passive secondary sequence (imperfect, aorist, pluperfect)
VBI verb, active imperative
VBIP verb, middle/passive imperative
VBN verb, active infinitive
VBNP verb, middle/passive infinitive
VBO verb, active optative
VBOP verb, middle/passive optative
VBP verb, active primary sequence (present, future, perfect)
VBPP verb, middle/passive primary sequence (present, future, perfect)
VBS verb, active subjunctive
VBSP verb, middle/passive subjunctive
VPR verb, active participle, nominative
VPR$ verb, active participle, genitive
VPRA verb, active participle, accusative
VPRD verb, active participle, dative
VPRP verb, middle/passive participle, nominative
VPRP$ verb, middle/passive participle, genitive
VPRPA verb, middle/passive participle, accusative
VPRPD verb, middle/passive participle, dative

Table 4.16. Extended Verbal Dash Tags

Tag Use
-AOR aorist (perfective aspect)
-FUT future
-IMPF imperfective aspect
-IND marks optatives replacing indicatives in secondary sequence
INTRNS marks middle or passive verbs that exhibit intransitive syntax
-KJV marks optatives replacing subjunctives in secondary sequence
-PASS syntactic passive (not middle!)
-PRF perfect aspect
TRNS1 marks middle or (morphologically) passive verbs that exhibit transitive syntax, taking an NP-OB1, NP-OBP, or NP-OBQ
TRNS2 marks middle or (morphologically) passive verbs that exhibit a more abstract notion of transitive syntax, taking a complement infinitive (IP-INF-COM, IP-INF-THT), complementizer phrase (CP-COM, CP-THT), or small clause (IP-SMC).

Notes

  • The forms of the verb εἰμί are distinguished from all other verbs, but no distinction is made in the tag between auxiliary and main verb uses. In the syntactically annotated (parsed) files, auxiliary forms of εἰμί can be distinguished from main verb forms by the presence of a participle in the clause.

  • Modals are not at present distinguished from other verbs.

  • All participles get *PR(P) tags regardless of whether the participle is being used attributively (i.e., adjectivally) or not. Attributive versus circumstantial use is distinguished in the syntactic annotation: see the section called “Reduced Relative Clauses”, the section called “Participial Clauses”.

  • Middle and passive verbs both receive the *P tags. Even when a middle verb has active syntax—i.e., it takes a direct object in the accusative—it still receives the *P tag.

  • Syntactically passive verbs—that is, verbs that are passive with respect to their syntax, regardless of whether they are ambiguously middle/passive or unambiguously passive with respect to their morphology—receive the -PASS dash tag.

  • Syntactically active verbs that have middle or passive morphology receive either a -TRNS1 or a -TRNS2 dash tag, depending on what type of complement they take (see the table above).

    Note that taking a NP-OB2 argument regularly does not qualify a verb for receiving a -TRNS1 or -TRNS2 tag. See ἐτάξετο and ἀπεκρίθη in the examples below.

    ( (IP-MAT (CODE {VS:28_16})
              (NP-SBJ (DS Οἱ-ὁ) (CLPRT δὲ-δέ) (NUM ἕνδεκα-ἕνδεκα) (NS μαθηταὶ-μαθητής))
              (VBDP-AOR-INTRNS ἐπορεύθησαν-πορεύομαι)
              (PP (P εἰς-εἰς)
                  (NP (DA τὴν-ὁ) (NPRA Γαλιλαίαν-Γαλιλαία)))
              (PP (P εἰς-εἰς)
                  (NP (DA τὸ-ὁ)
                      (NA ὄρος-ὄρος)
              (CP-REL (WADVP-1 (WADV οὗ-οὗ))
                      (C 0)
                      (IP-SUB (ADVP-LOC *T*-1)
                              (VBDP-AOR-INTRNS ἐτάξατο-τάσσω)
                              (NP-OB2 (PROD αὐτοῖς-αὐτός))
                              (NP-SBJ (D ὁ-ὁ) (NPR Ἰησοῦς-Ἰησοῦς))))))
              (, ,))
      (ID GreekNT,Matthew))                            
                            

    ( (IP-MAT (CODE {VS:27_14})
              (CONJ καὶ-καί)
              (NEG οὐκ-οὐ)
              (NP-SBJ *pro*)
              (VBDP-AOR-INTRNS ἀπεκρίθη-ἀποκρίνομαι)
              (NP-OB2 (PROD αὐτῷ-αὐτός))
              (PP (P πρὸς-πρός)
                  (NP (NEG-ADV οὐδὲ-οὐδέ) (NUM ἓν-εἷς) (NA ῥῆμα-ῥῆμα)))
              (, ,)
              (CP-RES (C-CLTE ὥστε-ὥστε)
                      (IP-INF (VBN-IMPF θαυμάζειν-θαυμάζω)
                              (NP-SBJ (DA τὸν-ὁ) (NA ἡγεμόνα-ἡγεμών))
                              (ADVP (ADV λίαν-λίαν))))
              (. .))
      (ID GreekNT,Matthew))                            
                            

  • Verbs that have middle or passive morphology but do not have (obvious) active or passive syntax get the -INTRNS tag. This covers many situations:

    • Normally transitive verbs that have a special, intransitive meaning in the passive.

      For example, ὁράω 'see' can have the meaning 'appear' in the passive:

      ( (IP-MAT (CODE {VS:17_3})
                (CONJ καὶ-καί)
                (INTJ ἰδοὺ-ἰδοὺ)
                (VBDP-AOR-INTRNS ὤφθη-ὁράω)
                (NP-OB2 (PROD αὐτοῖς-αὐτός))
                (NP-SBJ (NPR Μωυσῆς-Μωϋσῆς) (CONJ καὶ-καί) (NPR Ἠλείας-Ἡλίας))
                (IP-PPL (VPR-IMPF συνλαλοῦντες-συλλαλέω)
                        (PP (P μετ'-μετά)
                            (NP (PRO$ αὐτοῦ-αὐτός))))
                (. .))
        (ID GreekNT,Matthew))                                    
                                      

      When the meaning of a passive isn't sufficiently different from the active, then -PASS should be used unless one of the other situations described below applies.

      In the following example, for instance, the meaning of the verb σείω 'shake' is not significantly different in the passive (although it is metaphorical), so it gets a -PASS tag.

      ( (IP-MAT (CODE {VS:21_10})
                (CONJ καὶ-καί)
                (IP-ABS (VPR$-AOR εἰσελθόντος-εἰσέρχομαι)
                        (NP-SBJ (PRO$ αὐτοῦ-αὐτός))
                        (PP (P εἰς-εἰς)
                            (NP (NPRA Ἰεροσόλυμα-Ἱεροσόλυμα))))
                (VBDP-AOR-PASS ἐσείσθη-σείω)
                (NP-SBJ (Q πᾶσα-πᾶς)
                        (D ἡ-ὁ)
                        (CODE {VS:21_11})
                        (N πόλις-πόλις))
                (IP-PPL (VPR-IMPF λέγουσα-λέγω)
                (CP-QUE-SPE (WNP-1 (WPRO Τίς-τίς))
                            (C 0)
                            (IP-SUB-SPE (NP-PRD *T*-1)
                                        (BEP-IMPF ἐστιν-εἰμί)
                                        (NP-SBJ (D οὗτος-οὗτος)))))
                (. ;))
        (ID GreekNT,Matthew))                                        
                                          

    • Somewhat by convention, we treat morphologically middle or passive verbs that depict a change-of-state (here: ψυγήσεται 'will grow cold') as intransitive and label them with -INTRNS.

      ( (IP-MAT-SPE (CODE {VS:24_12})
                    (CONJ καὶ-καί)
                    (PP (P διὰ-διά)
                        (NP (DA τὸ-ὁ)
                            (IP-INF (VBNP-AOR-INTRNS πληθυνθῆναι-πληθύνω)
                                    (NP-SBJ (DA τὴν-ὁ) (NA ἀνομίαν-ἀνομία)))))
                    (VBPP-FUT-INTRNS ψυγήσεται-ψύχω)
                    (NP-SBJ (D ἡ-ὁ)
                            (N ἀγάπη-ἀγάπη)
                            (NP-ATR (DS$ τῶν-ὁ) (Q$ πολλῶν-πολύς)))
                    (. .))
        (ID GreekNT,Matthew))                                        
                                          

      When it seems particularly unlikely that a change-of-state verb in the passive is actually a syntactic passive (because there doesn't seem to be an implicit agent), then -INTRNS is applied even if the verb is treated as passive elsewhere.

      ( (IP-MAT (CODE {VS:27_51a})
                (CONJ καὶ-καί)
                (NP-SBJ (D ἡ-ὁ) (N γῆ-γῆ))
                (VBDP-AOR-INTRNS ἐσείσθη-σείω)
                (, ,))
        (ID GreekNT,Matthew))                                        
                                          

    • Similarly, morphologically middle or passive verbs that depict an emotional change-of-state are labeled -INTRNS.

      ( (IP-MAT (CODE {VS:14_14a})
                (CONJ καὶ-καί)
                (NP-SBJ *con*)
                (VBDP-AOR-INTRNS ἐσπλαγχνίσθη-σπλαγχνίζομαι)
                (PP (P ἐπ'-ἐπί)
                    (NP (PROD αὐτοῖς-αὐτός))))
        (ID GreekNT,Matthew))                                        
                                          

    • (Likely) unergative verbs are also labeled -INTRNS when they have middle or passive morphology.

      ( (IP-IMP-SPE (CODE {VS:24_20})
                    (VBIP-IMPF-INTRNS προσεύχεσθε-προσεύχομαι)
                    (CLPRT δὲ-δέ)
                    (CP-ADV-SPE (C ἵνα-ἵνα)
                                (IP-SUB-SPE (NEG μὴ-μή)
                                            (VBSP-AOR-INTRNS γένηται-γίνομαι)
                                            (NP-SBJ (D ἡ-ὁ)
                                                    (N φυγὴ-φυγή)
                                                    (NP-ATR (PRO$ ὑμῶν-ὑμεῖς)))
                                            (NP-TMP (N$ χειμῶνος-χειμών) (NEG-CONJ μηδὲ-μηδέ) (ND σαββάτῳ-σάββατον))))
                    (. .))
        (ID GreekNT,Matthew))                                        
                                          

    • When a transitive verb has middle morphology but is being used intransitively (either with a pro-dropped object, or a simple intransitive use), then it also is labeled -INTRNS.

      ( (IP-MAT-SPE (CODE {VS:27_4a})
                    (NP-SBJ (PRO σὺ-σύ))
                    (VBPP-FUT-INTRNS ὄψῃ-ὁράω)
                    (. .))
        (ID GreekNT,Matthew))                                        
                                          

  • Future verbs receive the -FUT dash tag, which extends a VBP* base POS tag. A dash tag was chosen to represent future rather than an additional base POS tag (e.g., VBF/VBFP) because of the existence of future infinitives, future participles, and future optatives, which would be impossible to distinguish otherwise.

  • Note that the existence of future perfect passive verbs in Greek forces the possibility of a verb’s having up to three dash tags: VBPP-FUT-PRF-PASS. In cases of two or three verbal dash tags, the tags should be applied in the order tense-aspect-voice.

  • Subjunctives and optatives are distinguished from indicative verb forms. Optatives in 2◦ secondary representing original indicatives receive the -IND dash tag, while optatives in 2◦ sequence representing original subjunctives receive the -KJV dash tag.

  • Aspect is indicated on verbs with the dash tags -IMPF (= imperfective, includes the imperfect tense, present subjunctive, etc.), -AOR (= aorist/perfective, includes the aorist indicative, aorist subjunctive, etc.), and -PRF (= perfect, includes the present perfect, pluperfect, etc.).

  • Only voice, aspect, and tense, but not number or gender, are distinguished on participles.

Wh- Words (W)

Table 4.17. W Tags

Tag Use
WADJ wh-adjective, nominative
WADJ$ wh-adjective, genitive
WADJA wh-adjective, accusative
WADJD wh-adjective, dative
WADV wh-adverb
WD wh-determiner, nominative
WD$ wh-determiner, genitive
WDA wh-determiner, accusative
WDD wh-determiner, dative
WP preposition in a wh- phrase
WPRO wh-pronoun, nominative
WPRO$ wh-pronoun, genitive
WPROA wh-pronoun, accusative
WPROD wh-pronoun, dative
WQ εἰ in the meaning ‘whether’ introducing an indirect question

Notes

  • Wh- words that are used as interjections are tagged with the appropriate POS tag (not as INTJ). Their function is indicated at the phrasal level; see the section called “Interjections (INTJ)”.

  • Wh- adverbs (πῶς, πότε, ποῦ, πόθεν, etc. including ὅ-variants) are tagged WADV.

  • Compound wh- words receive compound tags. For example, ἵνατί receives the compound tag WADV+CLQA.

  • τίς, ὅς, and ποῖος are tagged WD when not acting as the head of a wh- noun phrase, and as WPRO otherwise. Note the difference in this regard between wh- words and ordinary determiners, which are tagged as determiners invariably. See the section called “Notes” for details and examples.

  • ὅστις is tagged WPRO+CLQ.

  • ὅτε is tagged C when introducing adverbial clauses and as WADV when introducing questions (including indirect questions) or relative clauses. Ambiguous cases are taken as the wh- word.

  • Quantificational interrogative/relative adjectives such as πόσος ‘how many’ are tagged WADJ.

Compound Tags

There are several instances requiring compound tags:



[1] Henry George Liddell and Robert Scott. A Greek-English Lexicon. Revised and augmented throughout by Sir Henry Stuart Jones with the assistance of Roderick McKenzie. Oxford. Clarendon Press. 1940.

Part II. Syntactic Annotation

Table of Contents

5. Introduction
General Principles
Dash Tags
Structural Principles
Internal Structure of Phrases
Internal Structure of Clauses
IPs
Imperatives
Non-wh CPs
Wh- CPs
Fronting to pre-complementizer position
Adjunction to CP
Division into Sentence Tokens
6. Phrase-Level Constituents
Adjective Phrase
Complements of ADJ
Modifiers of ADJ
Adverb Phrase
Complements of ADV
Modifiers of ADV
Conjunction Phrase
Fragment
Interjection Phrase
Noun Phrase
Heads of NP
Complements of N
Modifiers of N
Number Phrase
Prepositional Phrase
Complements of P
Specifiers of P
Quantifier Phrase
Complements of Q
Modifiers of Q
Wh- Phrase
WADJP
WADVP
WNP
WPP
WQP
Non-Structural Labels
CODE
Foreign Language Passages
QTP
7. Conjunction
Word-Level Conjunction
Phrase-Level Conjunction
CONJPs with Empty Heads
Conjunction of Unlike Categories
Recursive Conjunction Structures
Shared Modifiers
Shared Pre-Modifiers
Shared Post-Modifiers
Conjunction and Negation
Extraposed CONJPs
Correlative Conjunction
Word-Level Correlative Conjunction
Phrase-Level Correlative Conjunction
Floated Conjunctions
Clausal Conjunction
Conjunction of Matrix IPs
Conjunction of Embedded IPs
Right-Node Raising
Gapping
Backwards Gapping
Conjoined Direct Speech
Conjunction of CPs
Clausal Conjunction and Shared Modifiers
8. Empty Categories
Empty Subjects
Subjects Elided under Conjunction
Arbitrary Subjects in ECM Infinitives
pro-Drop Subjects
Traces
Traces of A-Movement
Traces of Wh-Movement
Traces of Other A'-Movement
Other Empty Categories
Empty Objects
Empty Wh- Phrases and Complementizers
A Generic Empty Category
The Position of Empty Categories
The Position of the Null Copula
The Position of Empty Subjects
The Position of Traces
Resumptive Pronouns
9. Clause-Level Constituents
Adjective Phrase
Predicate Adjectives
Other Clause-Level ADJPs (ADJP-SPR)
Movement out of ADJP
Adverb Phrase
Noun Phrase
Subject
Object
Non-Argument (Adjunct) NP
Adverbial NP
Attributive (and/or Possessive) NP
Complement NP
Directional NP
Instrumental NP
Locative NP
Measure Phrase
Partitive NP
Predicate NP
Secondary Predicate NP
Temporal NP
Vocative
Bare NP
Prepositional Phrase
Quantifier Phrase
Words as Clausal Constituents
Adverbs
Floated Quantifiers
Negation
Participles
Extended Dash Tags
Left Dislocation
Appositive or Parenthetical
Resumptive
Direct Speech
Unknown
10. Subordinate Clauses
Adverbial Clauses
That-Clauses
Complement Clauses
Purpose Clauses
Degree Complements
Result Clauses
Comparative Clauses
Correlative Comparatives
Questions
Wh- in situ
Exclamative Clauses
Ordinary Relative Clauses
Reduced Relative Clauses
Reduced Relative Clauses versus Participial Clauses
Clause-Adjoined Relative Clauses
Free Relatives
ὡς Clauses with a Nominal Gap
Empty Operator Clauses
Infinitival Clauses
IP-INF-ADT
IP-INF-COM
IP-INF-DEG
IP-INF-PRP
IP-INF-SBJ
IP-INF-THT
Small Clauses
Participial Clauses
Absolute Clauses
Supplementary Participles
Participles of Indirect Discourse
11. Special Constructions
Prolepsis
Clausal Subjects
'If not' = 'except'
12. Discontinuous Phrases
Rightward Extraposition
Default Leftward Movement
Word-Level Fronting
Fronting of Non-Constituents
Phrase-Level Fronting
Clitic Displacement
Sentence-Position Particles
Clitic Pronouns and Verbs
13. Disfluencies
Break
False Start
Repetition
Tag

Chapter 5. Introduction

The parsing scheme for the PPCHiG, which is also used by the Penn Parsed Corpora of Historical English, uses a limited tree representation in the form of labeled parentheses. All open parentheses have an associated label, either a phrase label (NP, ADJP, etc.) or a word label (N, ADJ, etc.), representing nodes in a tree. We use the terms ‘word label’ and ‘POS (part-of-speech) label’ interchangeably. A word label is associated with every word, but phrasal labels are not included in every case in which a fully labeled tree would require them. Intermediate projections in the sense of X' theory (N', ADJ', etc.) are not generally included in our representations. By comparison to trees in current syntactic theory, the trees in our corpora are therefore quite flat, and they are not required to be binary-branching.

The partial representation of phrase structure in our corpora is not intended to make a theoretical statement, but was adopted for practical reasons. Certain phrases are generally omitted in the annotation scheme because their boundaries are too difficult to define. The prime example is VP. The problematic character of VP is particularly obvious in early Middle English, where the order of the verb and its complements is in flux (at least on the surface). But even in Present-Day English, the attachment site of verbal adjuncts is systematically ambiguous between low attachment to VP and high attachment at the clause level. Other categories, such as DP, were omitted because the cost of including them outweighs their usefulness. Intermediate projections are omitted for both reasons. In no case should the lack of any particular phrase label be taken to imply that earlier forms of Greek failed to include the corresponding syntactic category. The trees in the corpora are simply underspecified.

The examples in this section of the manual are constructed in Modern English so as to be maximally accessible. The remainder of the manual contains examples from the corpora. The examples are mostly from New Testament Greek (NTG); examples from Herodotus and Homer will be added later.

General Principles

As just mentioned, the structures in our corpora generally include neither a VP nor intermediate projections like I'. As a result, IP immediately dominates all verbs (to be understood in a broad sense, including modals and auxiliaries) and sentence-level constituents. A typical parse structure is the following:

(IP-MAT (NP-SBJ (NPR Mary))
        (HVP has)
        (BEN been)
        (VAG meaning)
        (IP-INF (TO to)
                (VB go))
        (PP (P for)
            (NP (D a) (N week)))))                
            

Dash Tags

  • Dash tags indicate syntactic subcategory or function. In general, phrase labels indicate a constituent’s category (NP, PP, ADJP, etc.). Additional labels (so-called dash tags), which are separated from the category label by a hyphen, indicate subcategory or function:

    • NP-SBJ (subject noun phrase)

    • ADJP-SPR (secondary predicate adjective phrase)

    • PP-PRN (parenthetical prepositional phrase)

  • All clause labels carry a dash tag. Not all phrases are labeled with a dash tag, and more than one such tag is possible (IP-INF-PRP = purpose infinitive, NP-SBJ-RSP = resumptive subject noun phrase, etc.). See the section called “Internal Structure of Clauses” for further basic information and Chapter 10, Subordinate Clauses for detailed discussion.

  • All sentence-level NPs carry a dash tag. Bare NPs are either complements of a nonverbal head (e.g., a preposition), or part of a conjunction structure.

    (IP-MAT (NP-TMP (N Yesterday))          ← temporal NP
            (NP-SBJ (NPR Mary))             ← subject
            (VBD told)
            (NP-OB2 (NPR Jane))             ← second object
            (CP-THT (C that)
                    (IP-SUB (NP-SBJ (PRO she))
                            (VBD studied)
                            (NP-MSR (QP (ADVR too) (Q much)))   ← measure NP
                            (PP (P during)
                                (NP (D the) (N+N weekend))))))  ← no dash tag    
                            

  • ADVPs are marked for function as directional (ADVP-DIR), locative (ADVP-LOC), or temporal (ADVP-TMP). Other functions are unmarked.

    (IP-MAT (NP-SBJ (NPR Mary))
            (ADVP (ADV happily))
            (VBD put)
            (NP-OB1 (D the) (N book))
            (ADVP-LOC (ADV there))
            (ADVP-TMP (ADV+WARD afterward)))                            
                            

  • PPs are not marked for function. However, parenthetical PPs are possible.

    (IP-MAT (NP-SBJ (NPR Mary))
            (VBD put)
            (NP-OB1 (D the) (N book))
            (PP (P on)
                (NP (D the) (N table)))
            (PP (P on)
                (NP (NPR Saturday))))                            
                            

Structural Principles

  • Sentence-level constituents bear phrasal labels.

    [Caution] Caution

    Exception: IP can immediately dominate the following word-level constituents:

    • verbal material (auxiliaries, modals, verbs, participles)

    • sentential conjunctions (CONJ0)

    • single-word interjections (INTJ)

    • sentential negation (NEG)

    • floated quantifiers (Q*)

    • particles (AN, CLGE, CLPRT, KE, PRTQ)

    • word-level constituents indexed with an *ICH*-trace

  • In general, no intermediate projections. Intermediate projections (N', ADJ', etc.) are represented only in certain conjunction structures and situations involving movement. The absence of intermediate projections has several consequences:

    • Within phrasal constituents, the phrasal node generally immediately dominates the head.

      (NP (D the) (N girl))
      
      (PP (P in)
          (NP (D the) (N spring)))
      
      (ADJP (ADV very) (ADJ big))                                        
                                          

    • Specifiers are not distinguished from pre-head modifiers.

    • Adjuncts are never distinguished from complements in structural terms. They may be distinguished by the presence of dash tags.

  • Single words that are not sentence-level constituents do not project phrasal nodes if the phrasal node would share the same category as the head.

    (NP (D a) (ADJ big) (N cat))            ← single-word case
    
    (NP (D a)
        (ADJP (ADV very) (ADJ big))         ← multi-word case
        (N cat))                            
                            

  • [Caution] Caution

    Exception: In the PPCHiG, post-head modifiers do not project structure beyond the word level unless they consist of more than one word; this marks a departure from the guidelines for the English corpora. Like pre-head modifiers, they are represented as sisters to the head, except for attributive participles, which are always given the RRC label.

    (NP (ADJ various) (ADJ black) (NS cats))
    
    (NP (ADJ black) (NS cats) (ADJ galore))
    
    (NP (D a)
        (RRC (VAN forsaken))
        (N castle))
    
    (NP (D an) (N castle)
        (RRC (VAN forsaken))                ← RRC = reduced relative clause                                    
                                    

  • Default high attachment. In cases where the attachment level of constituents is ambiguous or difficult to determine, the default is to attach as high as possible.

    (IP-MAT (NP-SBJ (NPR Mary))
            (VBD saw)
            (NP-OB1 (D the) (N man))
            (PP (P with)                    ← high by default
                (NP (D the) (N telescope))))
    
    (NP (D the)
        (N story)
        (CP-REL (WNP-1 0)
                (C that)
                (IP-SUB (NP-ACC *T*-1)
                        (NP-SBJ (PRO they))
                        (VBP tell)))
        (PP (P about)                       ← high by default
            (NP (PRO the)
                (N king))))
    (NP (D the)
        (N story)
        (RRC (BAG being)
             (VAN told))
        (PP (P about)                       ← high by default
            (NP (PRO the)
                (N king))))                            
                            

Internal Structure of Phrases

The internal structure of all non-clausal phrases is fundamentally similar.

  • Heads project a phrasal node.

    [Caution] Caution

    Exception 1: Certain heads, such as determiners (D) and particles (CLPRT, PRT), never project a phrase. Verbs do not project a VP.

    [Caution] Caution

    Exception 2: Single-word pre-head and post-head modifiers do not project a phrasal node if the phrasal node would share the same category as the head.

    (NP (D the) (NS girls)
        (PP (P on)
            (NP (D the) (N beach))))
    
    (NP (Q many) (ADJ happy) (NS girls)     ← two single-word pre-head modifiers
        (PP (P on)
            (NP (D the) (VAN overcrowded) (N beach))))
    
    (NP (ADJP (ADJ happy) (CONJ and) (VAN excited))   ← multi-word pre-head modifier
        (NS girls))
    
    (NP (Q many)                            ← single-word pre-head modifier
        (ADJP (ADV very) (ADJ happy))       ← multi-word pre-head modifier
        (NS girls)
        (PP (P on)
            (NP (D the) (VAN overcrowded) (N beach))))
    
    (NP (QP (ADV very) (Q many))            ← multi-word pre-head modifier
        (ADJP (ADV very) (ADJ happy))       ← multi-word pre-head modifier
        (NS girls)
        (PP (P on)                          ← post-head modifier
            (NP (D the) (N beach)
                (PP (P with)
                    (NP (D the) (ADJ big) (NS dunes))))))
    
    (ADJP (ADV very) (ADJ happy))	        ← single-word pre-head modifier
    
    (PP (ADV especially)		        ← single-word pre-head modifier
        (P on)
        (NP (NPRS Saturdays)))
    
    (PP (ADV right)			        ← single-word pre-head modifier
        (P up)
        (NP (D the) (N street)))
    
    (ADVP (ADV very) (ADV slowly))          ← single-word pre-head modifier
    
    (NP-PRD (NS γενεαὶ) (NUM δεκατέσσαρες)) ← single-word post-head modifier                            
                            

  • In general, the head of a phrase is overt and matches the category of the phrase, but in certain cases, there is no matching head. Sometimes, this is simply an artifact of our annotation scheme and due to the fact that some words receive a more specific label than others. For instance, pronouns are labeled PRO, not D or N. In other instances, the absence of an overt head reflects the possibility of some deletion process. The most common type of categorial mismatch between heads and phrases is nominal elements not labeled N, NS, etc.: PRO, CLPRO, CLQ(?), D(?), Q(?). As indicated by the question marks in the above list, it is not always clear whether a particular word modifies an elided head or is itself the head of the phrase. It is mainly for this reason that we do not attempt to make a systematic distinction between the two cases. The most common elision case is of NPs containing only a determiner and an adjective (or sometimes only an adjective).

    (IP-MAT (IP-MAT-1 (NP-SBJ (NPR Mary))
                      (VBD gave)
                      (NP-OB2 (NPR Jane))
                      (NP-OB1 (D a) (ADJ red) (N ribbon)))
            (CONJP (CONJ and)
                   (IP-MAT=1 (NP-OB2 (NPR Lucy))
                             (NP-OB1 (D a) (ADJ blue)))))   ← elided head noun                            
                            

Internal Structure of Clauses

Clauses are labeled either CP or IP. CPs contain either a complementizer or a wh- position (or both). IPs contain neither. All IPs and CPs carry dash tags indicating their subtype, as follows.

Table 5.1. Common IP-Dash Tag Combinations

Phrase Level Use
IP-IMP imperative
IP-INF infinitive
IP-INF-COM complementary infinitive (control, ECM)
IP-INF-IMP infinitive used as imperative
IP-INF-PRP purpose infinitive
IP-INF-SBJ subject infinitive
IP-INF-THT infinitive of indirect discourse
IP-MAT declarative matrix IP
IP-PPL participial clause
IP-ABS absolute clause, including genitive absolute
IP-PPL-COM supplementary participle (with certain verbs)
IP-PPL-THT participial clause of indirect discourse
IP-SMC small clause
IP-SUB subordinate clause

Table 5.2. Common CP-Dash Tag Combinations

Phrase Level Use
CP-ADV adverbial clause
CP-CAR clause-adjoined relative
CP-COM non-indicative complement clause
CP-CMP comparative clause
CP-DEG degree complement clause
CP-EOP empty operator CP
CP-EXL exclamative clause
CP-FRL free relative clause
CP-QUE direct or indirect question
CP-REL relative clause
CP-THT indicative complement clause

IPs

All IPs except subjectless imperatives and subjectless infinitives have a subject in our annotation scheme. If the subject is not overt, an empty subject is added. Clauses do not contain VP. As a rule, daughters of IP are phrasal (but see the section called “Internal Structure of Phrases” for exceptions).

(IP-MAT (CONJ0 But)                      ← sentential conjunction
        (INTJ alas)                     ← single-word interjection
        (, ,)
        (NP-SBJ (PRO we))
        (MD will)                       ← modal
        (NEG not)                       ← negation
        (Q all)                         ← floated quantifier
        (VB end)                        ← verb
        (RP up))                        ← particle
        (PP (P with)
            (NP-OB1 (PRO$ our) (N favorite)))
        (. .))                    
                

Although it's not common, occasionally IPs can carry additional function tags such as -SBJ or -PRD. In these cases, -SBJ or -PRD function tags follow the -SPE tag indicating direct speech, if applicable.

  • IP-MAT serving as a subject:

    ( (IP-MAT (CODE {VS:4_4})
              (NP-SBJ (D ὁ-ὁ))
              (CLPRT δὲ-δέ)
              (IP-PPL (VPRP-AOR ἀποκριθεὶς-ἀποκρίνομαι))
              (VBD-AOR εἶπεν-λέγω)
              (IP-MAT-SPE (VBPP-PRF-PASS Γέγραπται-γράφω)
                          (IP-MAT-SPE-SBJ (PP (PP (NEG Οὐκ-οὐ)
                                                  (P ἐπ'-ἐπί)
                                                  (NP (ND ἄρτῳ-ἄρτος) (ADJD μόνῳ-μόνος)))
                                              (CONJP *ICH*-1))
                                          (VBPP-FUT ζήσεται-ζάω)
                                          (NP-SBJ (D ὁ-ὁ) (N ἄνθρωπος-ἄνθρωπος))
                                          (, ,)
                                          (CONJP-1 (CONJ ἀλλ'-ἀλλά)
                                                   (PP (P ἐπὶ-ἐπί)
                                                       (NP (QD παντὶ-πᾶς)
                                                           (ND ῥήματι-ῥῆμα)
                                                           (RRC (VPRPD-IMPF ἐκπορευομένῳ-ἐκπορεύομαι)
                                                   (PP (P διὰ-διά)
                                                       (NP (N$ στόματος-στόμα)
                                                           (NP-ATR (N$ θεοῦ-θεός))))))))))
              (. .))
      (ID GreekNT,Matthew))                            
                            

  • IP-MAT serving as a predicate:

    ( (IP-IMP-SPE (CODE {VS:9_13})
                  (IP-PPL (VPRP-AOR πορευθέντες-πορεύομαι))
                  (CLPRT δὲ-δέ)
                  (VBI-AOR μάθετε-μανθάνω)
                  (CP-QUE-SPE (WNP-1 (WPRO τί-τίς))
                              (C 0)
                              (IP-SUB-SPE (NP-SBJ *T*-1)
                                          (BEP-IMPF ἐστιν-εἰμί)
                                          (IP-MAT-PRD (NP-OB1 (NP (NA Ἔλεος-ἔλεος))
                                                      (CONJP *ICH*-2))
                                                      (NP-SBJ *pro*)
                                                      (VBP-IMPF θέλω-θέλω)
                                                      (CONJP-2 (CONJ καὶ-καί)
                                                               (NP (NEG οὐ-οὐ) (NA θυσίαν-θυσία))))))
                  (. .))
      (ID GreekNT,Matthew))                                
                                

Imperatives

Imperatives are labeled IP-IMP. Only overt subjects are included in the annotation.

  • No overt subject:

    (IP-IMP (VBI-IMPF ὕπαγε)
            (PP (P εἰς)
                (NP (DA τὸν)
                    (NA οἶκόν)
                    (NP-ATR (CLPRO$ σου)))))                                
                                

    (Matthew 9.6)

  • Overt subject:

    (IP-IMP (VBI-AOR ἐλθάτω)
            (NP-SBJ (D ἡ)                   ← overt subject
                    (N βασιλεία)
                    (NP-ATR (CLPRO$ σου)))))                                
                                

    (Matthew 6.10)

Non-wh CPs

Indicative complement clauses (CP-THT), non-indicative complement clauses (CP-COM), degree complements (CP-DEG), and certain adverbial clauses (CP-ADV) have the following basic structure:

(CP (C {word, 0})
    (IP ...))                    
                

The complementizer position is always included; when not filled by an overt complementizer, it contains 0 (zero).

  • Overt complementizer:

    (CP-THT (C ὅτι)                         ← overt complementizer
            (IP-SUB (NP-SBJ (NPR Ἰωάνης))
                    (VBDP-AOR-PASS παρεδόθη))))                                
                                

    (Matthew 4.12)

  • No overt complementizer:

    (IP-SUB (NP-SBJ *pro*)
            (VBP-PRF οἴδατε)
            (CP-THT (C 0)                   ← null complementizer
                    (IP-SUB (NP-SBJ *T*-1)
                            (VBD-AOR ἐστερέωσεν)))) (Acts 3.16)                                
                                

Wh- CPs

A number of clause types, listed below, contain both a wh- position and a complementizer position. This is to allow for the case in which both positions are filled. Empty wh- positions and empty complementizers are both indicated by 0 (zero). The wh- operator is coindexed to a trace of the same category. See the section called “Traces of Wh-Movement” and the section called “The Position of Traces” for details.

  • certain adverbial clauses (CP-ADV)

  • clause-adjoined relatives (CP-CAR)

  • comparative clauses (CP-CMP)

  • exclamatives (CP-EXL)

  • free relatives (CP-FRL)

  • questions (CP-QUE)

  • relative clauses (CP-REL)

  • Both wh- and complementizer positions filled (possibly restricted to NTG; 42 examples in the GNT):

    (CP-REL (WNP-1 (WPRO ὅσα))
            (C+AN ἐὰν)
            (IP-SUB (NP-SBJ *pro*)
                    (VBS-IMPF θέλητε)
                    (CP-COM (C ἵνα)
                            (IP-SUB (NP-OB1 *T*-1)
                                    (VBS-IMPF ποιῶσιν)
                                    (NP-OB2 (PROD ὑμῖν))
                                    (NP-SBJ (DS οἱ) (NS ἄνθρωποι))))))                                
                                

    (Matthew 7.12)

  • Only wh- position filled:

    (CP-QUE (WADVP-1 (WADV Τί))
            (C 0)
            (IP-SUB (ADVP *T*-1)
                    (NP-SBJ *pro*)
                    (VBPX-IMPF (VBPP-IMPF θορυβεῖσθε) (CONJ καὶ) (VBP-IMPF κλαίετε))))                                
                                

    (Mark 5.39)

  • Only complementizer position filled (rare in Greek; 1 example in the GNT):

    (CP-QUE-SPE (WNP-1 (WPRO Τίς))
                (CLPRT ἄρα)
                (C 0)
                (IP-SUB (NP-PRD *T*-1)
                        (NP-SBJ-LFD (D οὗτός)
                                    (CP-REL *ICH*-2))
                        (BEP-IMPF ἐστιν)
                        (CP-REL-2 (WNP-3 0)
                                  (C ὅτι)
                                  (IP-SUB (NP-SBJ (ADV καὶ)
                                                  (NP (D ὁ) (N ἄνεμος))
                                                  (CONJP (CONJ καὶ)
                                                         (NP (D ἡ) (N θάλασσα))))
                                          (VBP-IMPF ὑπακούει)
                                          (NP-OBP-RSP-3 (PROD αὐτῷ))))))                                
                                

    (Mark 4.41)

Fronting to pre-complementizer position

Fronted elements can occupy Spec(CP), the position immediately preceding the complementizer. Since the specifier position is not explicitly indicated in our annotation system for any phrasal category, these elements simply appear within the CP in pre-head position. Such fronted constituents are coindexed with an *ICH* trace or with a resumptive (-RSP) phrase.

(CP-ADV (NP-LFD (PRO σὺ))
        (CLPRT οὖν)
        (C+AN ἐὰν)
        (IP-SUB (NP-SBJ-RSP *pro*)
                (VBS-AOR προσκυνήσῃς)
                (PP (P ἐνώπιον)
                    (NP (PRO$ ἐμοῦ)))))                    
                

(Luke 4.7)

Adjunction to CP

Material appearing before a wh- element must be adjoined to CP rather than occupying Spec(CP) (since the wh- element occupies that position), but our annotation does not explicitly express the distinction between the two types of positions.

( (CP-QUE-SPE (CODE {VS:21_25})
              (NP-LFD (D τὸ)
                      (N βάπτισμα)
                      (NP-PRN (D τὸ)
                              (NP-ATR (NPR$ Ἰωάνου)))
              (WADVP-1 (WADV πόθεν))
              (C 0)
              (IP-SUB (ADVP-DIR *T*-1)
                      (NP-SBJ-RSP *pro*)
                      (BED-IMPF ἦν))
              (. ;))
  (ID GreekNT,Matthew))                    
                

Division into Sentence Tokens

The conventions that govern the division of the text into sentence tokens in connection with conjunction are discussed in more detail in the section called “Clausal Conjunction”. What follows is a list of illustrative examples. Blank lines indicate a break between sentence tokens.

  • No token break with conjoined intransitive verbs.

    ( (IP-MAT (NP-SBJ (PRO He))
              (VBD (VBD came) (, ,) (VBD saw) (, ,) (CONJ and) (VBD conquered))
              (. .)))                                
                                

  • Ordinary VP conjunction triggers token break.

    ( (IP-MAT (NP-SBJ (PRO He))
              (VBD sang)
              (. ,)))
    
    ( (IP-MAT (CONJ and)
              (NP-SBJ *con*)
              (VBD danced)
              (NP-OB1 (D the) (N polka))
              (. .)))                                
                                

  • [Caution] Caution

    Proper treatment depends on meaning (not on punctuation).

    He came ,_, and left fifty years later ._.	← both events in the same year
    
    ( (IP-MAT (NP-SBJ (PRO He))
              (VBD (VBD came) (, ,) (CONJ and) (VBD left))
              (ADVP-TMP (NP-MSR (NUM 50) (NS years))
                        (ADVR later))
              (. .)))
    
    He came ,_.
    
    and left fifty years later ._.			← both events separated by 50 years
    
    ( (IP-MAT (NP-SBJ (PRO He))
              (VBD came)
              (. ,)))
    
    ( (IP-MAT (CONJ and)
              (NP-SBJ *con*)
              (VBD left)
              (ADVP-TMP (NP-MSR (NUM 50) (NS years))
                        (ADVR later))
              (. .)))                                    
                                    

  • No token break with first direct speech clause. See the section called “Conjoined Direct Speech”.

    ( (IP-MAT (NP-SBJ (PRO He))
              (VBD said)
              (, ,)
              (" ")			← default high attachment (contrary to sense)
              (IP-MAT-SPE (NP-SBJ (PRO I))
                          (VBD conquered))
              (" ")			← default high attachment (contrary to sense)
              (. .)))                                
                                

  • Further direct speech clauses become separate tokens. See the section called “Conjoined Direct Speech”.

    He said ,_, " I came ,_.
    
    I saw ,_.
    
    I conquered ._. "_"
    
    ( (IP-MAT (NP-SBJ (PRO He))
              (VBD said)
              (, ,)
              (" ")			← default high attachment (contrary to sense)
              (IP-MAT-SPE (NP-SBJ (PRO I))
                          (VBD came))
              (. ,)))
    
    ( (IP-MAT-SPE (NP-SBJ (PRO I))
                  (VBD saw)
                  (. ,)))
    
    ( (IP-MAT-SPE (NP-SBJ (PRO I))
                  (VBD conquered)
                  (. .)
                  (" ")))                                
                                

  • Clause-adjoined relatives (generally no token break, regardless of capitalization).

    She baked brownies ,_, whereupon joy reigned unconfined ._.
    
    She baked brownies ,_, Whereupon joy reigned unconfined ._.                                
                                

  • Ordinary combination of direct speech and clause-adjoined relative (no token break).

    She said ,_, " I will bake brownies ,_, "_" Whereupon joy reigned unconfined ._.                                
                                

  • Separate direct speech token forces clause-adjoined relative to become separate token.

    She said ,_, " I will fix the errors ._,
    
    and I will do so now ._. "_"
    
    Whereupon joy reigned unconfined ._.                                
                                

  • Where possible, interjections and vocatives are treated as part of a preceding or following sentence token.

Chapter 6. Phrase-Level Constituents

Adjective Phrase

Adjective phrases (ADJP) consist of an adjectival head, possibly accompanied by modifiers and/or complements. Participles, though tagged as VPR(P)*, can function as heads of ADJP. Quantifiers may also function as heads of ADJP when they have adjectival morphology. At the phrasal level, comparative adjectives accompanied by a determiner are labeled either ADJP or NP, depending on whether they are functioning as predicates or referring expressions. Such phrases are labeled NP if they pick out a particular entity. Superlative adjectives are usually embedded within or projecting an NP.

Complements of ADJ

Adjectives can take NP, PP, or clausal complements.

  • NP complements of adjectives are labeled NP-COM.

    (ADJP-PRD (NP-COM (CLPRO$ μου))
              (ADJ ἄξιος))                                
                                

    (Matthew 10.37)

  • Genitive NPs following comparative adjectives, however, are labeled NP-CMP.

    (NP-SBJ (D ὁ)
            (CLPRT δὲ)
            (RRC (PP (P ὀπίσω)
                     (NP (CLPRO$ μου)))
                 (VPRP-IMPF ἐρχόμενος)))
    (ADJP-PRD (ADJR ἰσχυρότερός)
              (NP-CMP (CLPRO$ μου)))
    (BEP-IMPF ἐστίν)                                
                                

    (Matthew 3.11)

  • PP complements of ADJ:

    (ADJP-PRD (ADJ μέγας)
              (PP (P ἐνώπιον)
                  (NP (N$ Κυρίου))))                                
                                

    (Luke 1.15)

  • CP complements of ADJ:

    • Infinitival complements with or without τοῦ (IP-INF-COM or CP-COM and IP-INF).

      ( (IP-MAT-SPE (CODE {VS:23_15})
                    (NP-SBJ (PRO ἡμεῖς))
                    (CLPRT δὲ)
                    (PP (P πρὸ)
                        (NP (D$ τοῦ)
                            (IP-INF (VBN-AOR ἐγγίσαι)
                                    (NP-SBJ (PROA αὐτὸν)))))
                    (ADJP-PRD (ADJ ἕτοιμοί)
                              (CP-COM *ICH*-1))
                    (BEP-IMPF ἐσμεν)
                    (CP-COM-1 (C τοῦ)
                              (IP-INF (VBN-AOR ἀνελεῖν)
                                      (NP-OB1 (PROA αὐτόν))))
                    (. .))
        (ID GreekNT,Acts))                                            
                                              

    • That complements (CP-THT). There are no examples of this in the GNT, but it is a conceivable possibility.

    • Comparative complements (CP-CMP).

      (ADJP-PRD (ADJR μείζων)
                (CP-CMP *ICH*-1))
      (CLPRT δὲ)
      (BEP-IMPF *)
      (NP-SBJ (D ὁ)
              (RRC (VPR-IMPF προφητεύων)))
      (CP-CMP-1 (WADJP-2 0)
                (C ἢ)
                (IP-SUB (ADJP-PRD *T*-2)
                        (BEP-IMPF *)
                        (NP-SBJ (D ὁ)
                                (RRC (VPR-IMPF λαλῶν)
                                     (NP-ADV (NSD γλώσσαις))))))                                        
                                          

      (1Corinthians 14.5)

    • Degree complements (CP-DEG).

      (ADJP (ADJ χαλεποὶ)
            (ADVR λίαν)
            (CP-DEG (C ὥστε)
                    (IP-INF (NEG μὴ)
                            (VBN-IMPF ἰσχύειν)
                            (NP-SBJ (PROA τινὰ))
                            (IP-INF-COM (VBN-AOR παρελθεῖν)
                                        (PP (P διὰ)
                                            (NP (D$ τῆς) (N$ ὁδοῦ) (D$ ἐκείνης)))))))                                        
                                          

      (Matthew 8.28)

Modifiers of ADJ

Modifiers of adjectives are adverbs and adverb phrases.

  • adverbs:

    (ADJP (ADJA μεγάλην) (ADV σφόδρα))                                
                                

    (Matthew 1.17)

  • adverb phrases:

    (ADJP-PRD (ADVP (ADVR περισσότερον) (ADV ἔτι))
              (ADJ κατάδηλόν))                                
                                

    (Hebrews 7.15)

Adverb Phrase

Adverb phrases (ADVP) consist of an adverb, quantifier, or number head, possibly accompanied by modifiers and/or complements. The different types of adverb phrase that we distinguish (directional, locative, temporal, etc.) are discussed in the section called “Adverb Phrase”

Complements of ADV

Complements of adverbs include PPs, comparatives, and degree clauses.

  • PP complements of ADV:

    (ADVP-LOC (NEG οὐ)
              (ADV μακρὰν)
              (PP (P ἀπὸ)
                  (NP (NUM ἑνὸς)
                      (Q$ ἑκάστου)
                      (NP-PAR (PRO$ ἡμῶν)))))                                
                                

    (Acts 17.27)

  • Comparative complements of ADV (CP-CMP).

    (ADVP (QR μᾶλλον)
          (CP-CMP (WADVP-2 0)
                  (C ἢ)
                  (IP-SUB=1 (ADVP *T*-2)
                            (PP (P ἐπὶ)
                                (NP (DSD τοῖς)
                                    (NUMP (NUM ἐνενήκοντα)
                                          (NUM ἐννέα))
                                    (NP-PRN (DSD τοῖς)
                                            (RRC (NEG μὴ) (VPRPD-PRF πεπλανημένοις))))))))                                
                                

    (Matthew 18.13)

  • Degree complements of ADV (CP-DEG).

    (ADVP (ADVR οὕτως)
          (CP-DEG (C ὥστε)
                  (IP-INF (VBN-AOR πιστεῦσαι)
                          (NP-SBJ (NP-PAR (ADJ$ Ἰουδαίων) (CLTE τε) (CONJ καὶ) (NPRS$ Ἑλλήνων))
                                  (ADJA πολὺ)
                                  (NA πλῆθος)))))                                
                                

    (Acts 14.1)

Modifiers of ADV

Modifiers of adverbs are other adverbs, adverb phrases, quantifiers, quantifier phrases.

  • adverbs and adverb phrases:

    (ADVP (ADV καὶ) (ADV πάλιν))                                
                                

    (Acts 17.32)

  • quantifiers and quantifier phrases:

    (ADVP (QD πολλῷ) (QR μᾶλλον))                                
                                

    (Mark 10.48)

Conjunction Phrase

See Chapter 7, Conjunction

Fragment

Fragments are grammatical utterances consisting of at least two constituents, for which there is not enough material to construct an IP. In Greek, these are often stand-alone participial clauses, as in the second example below. Fragments that are direct speech are labeled QTP rather than FRAG.

( (FRAG (CODE {VS:1_1})
        (NP-SBJ (N ΑΡΧΗ)
                (NP-ATR (D$ τοῦ)
                        (N$ εὐαγγελίου)
                        (NP-COM (NPR$ Ἰησοῦ)
                                (NP-PRN (NPR$ Χριστοῦ)))))
        (. .))
  (ID GreekNT,Mark))                
            

( (FRAG (CODE {VS:24_26})
        (IP-PPL (ADVP-TMP (ADV ἅμα))
                (ADVP (ADV καὶ))
                (VPR-IMPF ἐλπίζων)
                (CP-THT (C ὅτι)
                        (IP-SUB (NP-SBJ (NS χρήματα))
                                (VBPP-FUT-PASS δοθήσεται)
                                (NP-OB2 (PROD αὐτῷ))
                                (PP (P ὑπὸ)
                                    (NP-AGT (D$ τοῦ) (NPR$ Παύλου))))))
        (. .))
  (ID GreekNT,Acts))                
            

( (FRAG (CODE {VS:15_31})
        (PP (P νὴ)
            (NP (DA τὴν)
                (ADJA ὑμετέραν)
                (NA καύχησιν)
                (CP-REL *ICH*-1)))
        (, ,)
        (NP-VOC (NS ἀδελφοί))
        (, ,)
        (CP-REL-1 (WNP-2 (WPRO ἣν))
                  (C 0)
                  (IP-SUB (NP-OB1 *T*-2)
                          (NP-SBJ *pro*)
                          (VBP-IMPF ἔχω)
                          (PP (P ἐν)
                              (NP (NPRD Χριστῷ)
                                  (NP-PRN (NP (NPRD Ἰησοῦ))
                                          (CONJP (NP (DD τῷ)
                                                     (ND κυρίῳ)
                                                     (NP-ATR (PRO$ ἡμῶν)))))))))
        (. .))
  (ID GreekNT,1Corinthians))                
            

( (IP-MAT (CODE {VS:13_51})
          (NP-SBJ *pro*)
          (VBP-IMPF λέγουσιν)
          (NP-OB2 (PROD αὐτῷ))
          (QTP (INTJ Ναί))
          (. .))
  (ID GreekNT,Matthew))                
            

Interjection Phrase

Single words tagged as INTJ do not project an INTJP, even when they are constituents at the clausal level, except as noted below.

( (IP-IMP-SPE (CODE {VS:5_24})
              (CONJ0 καὶ)
              (INTJ ὕπαγε)
              (ADVP-TMP (ADV πρῶτον))
              (VBIP-AOR-PASS διαλλάγηθι)
              (NP-OB2 (DD τῷ)
                      (ND ἀδελφῷ)
                      (NP-ATR (CLPRO$ σου)))
              (, ,))
  (ID GreekNT,Matthew))                
            

The label INTJP is used in the following cases:

  • When the interjection makes up a separate token, as is commonly the case with the word ἀμήν in Paul’s letters in the PPCNTG, etc.

    ( (INTJP (CODE {VS:1_25})
             (INTJ ἀμήν)
             (. .))
      (ID GreekNT,Romans))                            
                            

  • When the interjection takes a complement, which can be (in theory) an NP, PP, or clause; only the first of these options seems to exist in the GNT.

    ( (QTP (CODE {VS:18_7})
           (INTJP (INTJ Οὐαὶ)
                  (NP-COM (DD τῷ) (ND κόσμῳ)))
           (PP (P ἀπὸ)
               (NP (DS$ τῶν) (NS$ σκανδάλων)))
           (. .))
      (ID GreekNT,Matthew))                            
                            

  • When the interjection is itself a complement. There are no examples of this in the GNT, but it seems possible. Thus an example from the annotation manual for the Penn Parsed Corpora of Historical English:

    (PP (PP (P by)
            (INTJP (INTJ yea)))
        (CONJP (CONJ and)
               (PP (P by)
                   (INTJP (INTJ nay)))))                            
                            

  • In the case of vocatives preceded by ὦ, the accompanying interjection ὦ is treated as a clause-level constituent rather than as a constituent of the vocative NP.

    (INTJ ὦ)
    (NP-VOC (N ἄνθρωπε))                            
                            

    (Romans 9.20)

  • When a word that is not tagged INTJ at the word level (see the section called “Interjections (INTJ)”) is used as an interjection, either alone or as part of a phrase.

Noun Phrase

Heads of NP

  • Noun phrases are canonically headed by a nominal element, most commonly a noun or a pronoun.

  • An NP without an overt nominal element (N, CLPRO, PRO) is assumed to have an empty nominal head. The empty head is not explicitly included in the annotation, but is implicit in the NP that it projects. An NP may therefore consist entirely of a demonstrative, adjective, number, quantifier, or possessive expression, or of phrases containing these elements, as illustrated by the following examples and as discussed in more detail below.

  • demonstrative:

    (NP-OB1 (DSA Ταῦτα))                            
                            

    (Matthew 1.20)

  • adjective:

    (NP (DD τῇ) (ADJD ἐρήμῳ))                            
                            

    (Matthew 3.3)

  • number:

    (NP (NUM ἑνὸς)
        (Q$ ἑκάστου)
        (NP-PAR (PRO$ ἡμῶν)))                            
                            

    (Acts 17.27)

  • quantifier:

    (NP-SBJ (Q πάντα))                            
                            

    (Matthew 5.18)

  • possessive expression:

    (NP-ATR (CLPRO$ μου))                            
                            

    (Mark 16.17)

  • At the clausal level, comparative adjectives accompanied by a determiner are labeled as either NP or ADJP, depending on whether they are functioning as referring expressions or as predicates. They are labeled NP when the phrase in question functions as a noun phrase (subject, object of verb or preposition, etc.). This includes cases where the phrase clearly picks out a particular entity, even if the phrase is the complement of a copular verb. Otherwise, comparative adjectives project an ADJP.

    ( (IP-MAT-SPE (CODE {VS:16_10})
                  (NP-SBJ (D ὁ)
                          (ADJP (ADJ πιστὸς)
                                (PP (P ἐν)
                                    (NP (ADJS ἐλαχίστῳ)))))
                  (ADJP-PRD (ADV καὶ)
                            (PP (P ἐν)
                                (NP (ADJD πολλῷ)))
                            (ADJ πιστός))
                  (BEP-IMPF ἐστιν)
                  (, ,))
      (ID GreekNT,Luke))                            
                            

  • At the phrasal level, superlative adjectives are generally embedded within or projecting an NP.

    (NP-VOC (ADJS κράτιστε) (NPR Φῆλιξ))                            
                            

    (Acts 24.2)

  • A superlative adjective projects an ADJP if it meets all of the following conditions:

    • It occurs alone (without a determiner, possessive, or modifier).

    • It is the complement of a copular verb.

    • It functions as a predicate rather than as a referring expression (that is, it does not pick out a particular entity).

    ( (IP-MAT-SPE (CODE {VS:2_6})
                  (CONJ0 Καὶ)
                  (NP-SBJ (PRO σύ)
                          (, ,)
                          (NP-PRN (NP (NPR Βηθλεὲμ))
                                  (CONJP (NP (N γῆ)
                                             (NP-ATR (NPR$ Ἰούδα)))))
                          (, ,))
                  (ADVP (NEG+ADV οὐδαμῶς))
                  (ADJP-PRD (ADJS ἐλαχίστη))
                  (BEP-IMPF εἶ)
                  (PP (P ἐν)
                      (NP (DSD τοῖς)
                          (NSD ἡγεμόσιν)
                          (NP-ATR (NPR$ Ἰούδα))))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

Complements of N

  • NP complements of N. NP complements of nouns are labeled NP-COM. In the traditional terms of Greek grammar[2], these NP complements, most often in the genitive case, are objective genitives.

    (NP (DA τὴν)
        (NA μετοικεσίαν)
        (NP-COM (NPR$ Βαβυλῶνος)))                            
                            

  • CP complements of N. Nouns can take clausal complements (CP-THT, IP-INF, etc.) (see Chapter 10, Subordinate Clauses for details).

Modifiers of N

NPs are modified by adjectives and adjective phrases (ADJ, ADJP), genitives, including possessives, (NP-ATR, NP-PAR), measure noun phrases (NP-MSR), numbers and number phrases (NUM, NUMP), quantifiers and quantifier phrases (Q, QP), prepositional phrases (PP), ordinary and reduced relative clauses (CP- REL, RRC), and by appositives and parentheticals (NP-PRN).

  • adjectives and adjective phrases:

    (NP (N$ πνεύματος) (ADJ$ ἁγίου))                                
                                

    (Matthew 1.18)

    (NP-OB1 (NA χαρὰν)
            (ADJP (ADJA μεγάλην) (ADV σφόδρα)))                                
                                

    (Matthew 2.10)

  • genitives:

    • NP-ATR: In the traditional terms of Greek grammar, the following types of genitives are tagged NP-ATR: genitive of possession, subjective genitive, genitive of material, appositional genitive, and genitive of value.

      (NP-OB2 (DD τῷ)
              (ND ἀδελφῷ)
              (NP-ATR (CLPRO$ σου)))                                            
                                              

      (Matthew 5.24)

    • NP-PAR: Partitive genitives are labeled NP-PAR.

      (NP (NUM ἑνὸς)
          (Q$ ἑκάστου)
          (NP-PAR (PRO$ ἡμῶν)))                                        
                                          

      (Acts 17.27)

  • measure noun phrases (rare modifying nouns in Greek, more often modifying verbs, as also occurs in the example below, corresponds with the genitive of measure in the traditional terms of Greek grammar):

    ( (IP-MAT (CODE {VS:2_44})
              (IP-PPL (VPR-AOR νομίσαντες)
                      (CLPRT δὲ)
                      (IP-INF-THT (NP-SBJ (PROA αὐτὸν))
                                  (BEN-IMPF εἶναι)
                                  (PP (P ἐν)
                                      (NP (DD τῇ) (ND συνοδίᾳ)))))
              (NP-SBJ *pro*)
              (VBD-AOR ἦλθον)
              (NP-MSR (NP-MSR (N$ ἡμέρας))
                      (NA ὁδὸν)))
      (ID GreekNT,Luke))                                
                                

  • numbers and number phrases:

    (NP-SBJ (NUM μίαν) (NA τρίχα))                                
                                

    (Matthew 5.36)

    (NP (NSD ψυχαῖς)
        (NUMP (NUM ἑβδομήκοντα) (NUM πέντε)))                                
                                

    (Acts 7.14)

  • quantifiers and quantifier phrases:

    (NP-SBJ (Q Πᾶσαι)
            (CLPRT οὖν)
            (DS αἱ)
            (NS γενεαὶ)
            (PP (P ἀπὸ)
                (NP (NPR$ Ἀβραὰμ)))
            (PP (P ἕως)
                (NP (NPR$ Δαυεὶδ))))                                
                                

    (Matthew 1.17)

    (NP-OB1 (QP (NEG οὐκ) (QA ὀλίγην))
            (NA ἐργασίαν))                                
                                

    (Acts 19.24)

  • prepositional phrases:

    (NP-OB1 (DSA τοὺς)
            (NSA προφήτας)
            (NP-PRN (DS τοὺς)
                    (PP (P πρὸ)
                        (NP (PRO$ ὑμῶν)))))                                
                                

    (Matthew 5.12)

  • ordinary and reduced relative clauses (This section focuses on the relation of relative clauses to the head they modify. The internal structure of relative clauses is discussed in the section called “Ordinary Relative Clauses” for full relative clauses and the section called “Reduced Relative Clauses” for reduced relative clauses.):

    (NP-SBJ (D ὁ)
            (N ἀστὴρ)
            (CP-REL (WNP-1 (WPROA ὃν))
                    (C 0)
                    (IP-SUB (NP-OB1 *T*-1)
                            (NP-SBJ *pro*)
                            (VBD-AOR εἶδον)
                            (PP (P ἐν)
                                (NP (DD τῇ) (ND ἀνατολῇ))))))                                
                                

    (Matthew 2.9)

    ( (IP-MAT-SPE (CODE {VS:3_10})
                  (NP-SBJ (Q πᾶν)
                          (CLPRT οὖν)
                          (N δένδρον)
                          (RRC (NEG μὴ)
                               (VPR-IMPF ποιοῦν)
                               (NP-OB1 (NA καρπὸν) (ADJA καλὸν))))
                  (VBPP-IMPF ἐκκόπτεται))
      (ID GreekNT,Matthew))                                
                                

  • appositives and parentheticals:

    (NP (NPRA Ἀνδρέαν)
        (NP-PRN (DA τὸν)
                (NA ἀδελφὸν)
                (NP-ATR (NPR$ Σίμωνος))))                                
                                

    (Mark 1.16)

Number Phrase

The following discussion focuses on the internal structure of NUMP. For the integration of number phrases in the larger clausal structure, see the section called “Heads of NP” and the section called “Measure Phrase”.

  • In the ordinary case, a number phrase consists of either a multi-word number, or a number modified in some way. When single numbers are written out in words, they are given flat structures.

    (NUMP (NUM ἑκατὸν) (NUM εἴκοσι'))                            
                            

    (Acts 1.15)

  • When coordinating conjunctions or determiners are part of a number, they are labeled with their own POS tag (CONJ, D), but otherwise treated as if tagged NUM—i.e., the structure is still flat.

    (NP (NSD ἔτεσι)
        (NUMP (ADJD τετρακοσίοις) (CONJ καὶ) (NUM πεντήκοντα)))                            
                            

    (Acts 13.20)

  • As usual, single-word modifiers are not enclosed in phrasal brackets, but multi-word modifiers are. When directly modifying a number and meaning ‘about, approximately’, ὡς is tagged ADV.

    (NUMP (ADV ὡς) (NUM ἑβδομήκοντα) (NUM ἕξ))                            
                            

    (Acts 27.37)

  • NUMP can surround syntactically complex phrases that function as number phrases. Such number phrases can be headed by categories that can head NP or QP (for instance, ADJ or Q).

    (NUMP (QP (Q no) (QR less))
                (PP (P than)
                    (NP (NUM 280))))                            
                            

  • Other heads (for instance, P) project a maximal projection in the way that they usually would, which is then enclosed by NUMP to encode the phrase’s function. Some of the following examples include sentence-level NPs for clarity.

    (NP-OB1 (NUMP (PP (P between)
                      (NP (NUMP (NUM thirty) (CONJ and) (NUM forty)))))
                          (NS persons))                            
                            

  • Finally, NUMP is used rather loosely in some instances to contain numerical constituents of unclear structure.

    (NUMP (NUM five)
          (PP (P times)
              (NP (NUM ten))))                                
                                

Prepositional Phrase

PPs are headed by prepositions and take complements of various categories. The most common are NP and CP, but ADJP, ADVP, IP, PP, etc. are possible as well.

Complements of P

  • NP complements of P.

    (PP (P ἐν)
        (NP (DD τῇ) (ADJD ἐρήμῳ)))                            
                            

    (Mark 1.3)

  • Pre-head NP complements of P. Some PPs or words that are treated as PPs when they take a dependent NP (such as χάριν) take a pre-head complement. This is rare in the GNT even with items that regularly or exclusively took pre-head complements in Classical Greek.

    (WPP-1 (NP (WPRO$ τίνος))
           (P ἕνεκα))                            
                            

    (Acts 19.32)

  • ADVP Complements of P.

    (PP (P ἕως)
        (ADVP (ADV ἄρτι)))                            
                            

    (Matthew 11.12)

  • CP complements of P. The most common of these are comparative clauses and adverbial clauses. For details on the structures of these types of subordinate clauses, see the section called “Comparative Clauses” and the section called “Adverbial Clauses”.

    (PP (P πρὶν)
        (CP-ADV (C ἢ)
                (IP-INF (ADVP (NUM δὶς))
                        (NP-SBJ (NA ἀλέκτορα))
                        (VBN-AOR φωνῆσαι))))                            
                            

    (Mark 14.30)

    (PP (P ὡς)
        (CP-ADV (C+AN ὅταν)
                (IP-SUB (NP-SBJ (D ὁ) (N λύχνος))
                        (NP-ADV (DD τῇ) (ND ἀστραπῇ))
                        (VBS-IMPF φωτίζῃ)
                        (NP-OB1 (CLPROA σε)))))                            
                            

    (Luke 11.36)

  • IP complements of P.

    (PP (NEG μὴ)
        (P εἰς)
        (IP-INF (NP-ADV (ADJA κενὸν))
                (NP-OB1 (DA τὴν)
                        (NA χάριν)
                        (NP-ATR (D$ τοῦ) (N$ θεοῦ)))
                (VBNP-AOR δέξασθαι)
                (NP-SBJ (PROA ὑμᾶς))))                            
                            

    (2Corinthians 6.1)

  • PP complements of P.

    (PP (P ἕως)
        (PP (P ἔξω)
            (NP (D$ τῆς) (N$ πόλεως))))                            
                            

    (Acts 21.5)

Specifiers of P

Items immediately preceding P, such as adverbs and constituent negation are treated as specifiers of PP. As with all specifiers in our annotation scheme, they are represented as sisters of the head.

  • adverbs:

    (PP (ADV μακρὰν)
        (P ἀπ')
        (NP (PRO$ αὐτῶν)))                                
                                

    (Matthew 8.30)

  • negation:

    (CP-FRL (WNP-1 (WPRO ὃς))
            (C 0)
            (IP-SUB (IP-SUB-2 (NP-SBJ *T*-1)
                              (AN ἂν)
                              (VBS-AOR ἀπολύσῃ)
                              (NP-OB1 (DA τὴν)
                                      (NA γυναῖκα)
                                      (NP-ATR (PRO$ αὐτοῦ)))
                              (PP (NEG μὴ)
                                  (P ἐπὶ)
                                  (NP (ND πορνείᾳ))))                                
                                

    (Matthew 19.9)

Quantifier Phrase

The following discussion focuses on the internal structure of QP. For the integration of Q(P) into the larger clausal structure, see the section called “Words as Clausal Constituents”, the section called “Heads of NP”, and the section called “Measure Phrase”.

Complements of Q

Quantifiers often take genitive NP complements. These are labeled as is appropriate given their semantics—usually NP-PAR.

(NP-OB1 (QA πολλοὺς)
        (NP-PAR (DS$ τῶν)
                (NS$ υἱῶν)
                (NP-ATR (NPR$ Ἰσραὴλ))))                    
                

(Luke 1.16)

Modifiers of Q

For examples and discussion of modifiers of Q that are labeled as NP-MSR, see the section called “Measure Phrase”.

The most common modifiers of Q are degree modifiers—e.g., ὡς—all of which are tagged ADVR in this con- text.

ὡς/ADVR πλεῖστα/Q                    
                

Other (non-comparative) modifiers are tagged ADV. Non-adverbial modifiers include other quantifiers, numbers, negation, and demonstratives. Quantifiers can also be modified by phrases. Multi-word modifiers of quantifiers that are themselves headed by Q are labeled QP.

(NP-MSR (QP (NEG οὐ)
            (QA πολλὰς))
        (NSA ἡμέρας))                    
                

(John 2.12)

Wh- Phrase

In general, the internal structure of wh- phrases is identical to that of their non-wh counterparts. However, the content of a wh- phrase can generally (except for WPP) be 0 (zero), indicating an empty operator. In pied-piping contexts, embedded wh- constituents recursively project a wh- feature all the way up to the phrase in Spec(CP). See the section called “Traces of Wh-Movement” for further discussion and examples. In general, wh- phrases are not annotated with dash tags indicating information concerning grammatical function, subcategory, and the like. The relevant information is associated with the associated wh- trace. Exceptions to the general rule occur in pied-piping contexts where the wh- phrase is not associated with a trace (as sometimes occurs). In such cases, we add a dash to the wh- phrase’s label. Again, see the section called “Traces of Wh-Movement” for examples.

WADJP

A WADJP may consist of an empty operator (WADJP 0) (only in comparatives) or an adjective modified by how

  • WADJP 0:

    (IP-MAT-SPE-1 (ADJP-PRD (ADJR ἀνεκτότερον)
                            (CP-CMP *ICH*-2))
                  (NP-SBJ *pro*)
                  (BEP-FUT ἔσται)
                  (NP-OB2 (ND γῇ)
                          (NP-ATR (NPRS$ Σοδόμων) (CONJ καὶ) (NPRS$ Γομόρρων)))
                  (PP (P ἐν)
                      (NP (ND ἡμέρᾳ)
                          (NP-ATR (N$ κρίσεως))))
                  (CP-CMP-2 (WADJP-3 0)
                            (C ἢ)
                            (IP-SUB=1 (ADJP-PRD *T*-3)
                                      (NP-OB2 (DD τῇ) (ND πόλει) (DD ἐκείνῃ)))))                                
                                

    (Matthew 10.15)

  • adjective modified by how:

    (NODE (CP-QUE (WADJP-3 (WADV hou) (ADJ seruisable))     ← ADJ modified by HOW
                  (C 0)
                  (IP-SUB (ADJP *T*-3)
                          (NP-SBJ (PRO he))
                          (BED was)
                          (PP (P to)
                              (NP (PRO$ his) (N moder)))))
          (ID CMAELR3,41.440))                                
                                

WADJPs that modify nouns are contained within a WNP, as are WADJPs functioning as measure phrases at the clausal level.

(NODE (CP-QUE (WNP-3 (WADJP (WADV how) (ADJ longe)))
              (C 0)
              (IP-SUB (NP-MSR *T*-3)
                      (NP-SBJ (PRO thou))
                      (HVP hast)
                      (VBN continued)
                      (PP (P in)
                          (NP (N synne)))))
      (ID CMCTPARS,322.C2.1496))                    
                

WADVP

A WADVP consists either of an empty operator (WADVP 0) or is headed by a wh- adverb.

  • empty operator:

  • headed by a wh- adverb:

    (CP-REL-1 (WADVP-2 (WADV ὅθεν))
              (C 0)
              (IP-SUB (ADVP-DIR *T*-2)
                      (NP-SBJ *pro*)
                      (VBD-AOR ἐξῆλθον))))                                
                                

    (Matthew 12.44)

WNP

A WNP may consist of an empty operator (WNP 0), the wh- pronouns (WPRO), or a noun modified by the wh- determiners (WD).

  • empty operator (in a ὡς clause with a nominal gap, see the section called “ὡς Clauses with a Nominal Gap”):

    ( (IP-MAT (CODE {VS:1_24})
              (VPRP-AOR-1 Ἐγερθεὶς)
              (CLPRT δὲ)
              (NP-SBJ (D ὁ) (NPR Ἰωσὴφ))
              (IP-PPL (VPRP-AOR *ICH*-1)
                      (PP-1 (P ἀπὸ)
                            (NP (D$ τοῦ) (N$ ὕπνου)))
              (VBD-AOR ἐποίησεν)
              (NP-OB1 (CP-FRL (WNP-2 0)
                              (C ὡς)
                              (IP-SUB (NP-OB1 *T*-2)
                                      (VBD-AOR προσέταξεν)
                                      (NP-OB2 (PROD αὐτῷ))
                                      (NP-SBJ (D ὁ)
                                              (N ἄγγελος)
                                              (NP-ATR (N$ Κυρίου)))))))
      (ID GreekNT,Matthew))                                
                                

  • empty operator (with later pronoun resumption):

    (CP-REL-2 (WNP-3 0)
              (C ὅτι)
              (IP-SUB (NP-SBJ (ADV καὶ)
                              (NP (D ὁ) (N ἄνεμος))
                              (CONJP (CONJ καὶ)
                                     (NP (D ἡ) (N θάλασσα))))
                      (VBP-IMPF ὑπακούει)
                      (NP-OBP-RSP-3 (PROD αὐτῷ))))                                
                                

    (Mark 4.41)

  • WPRO:

    (NP-SBJ (D ὁ)
            (N ἀστὴρ)
            (CP-REL (WNP-1 (WPROA ὃν))
                    (C 0)
                    (IP-SUB (NP-OB1 *T*-1)
                            (NP-SBJ *pro*)
                            (VBD-AOR εἶδον)
                            (PP (P ἐν)
                                (NP (DD τῇ) (ND ἀνατολῇ))))))                                
                                

    (Matthew 2.9)

  • WD:

    (CP-QUE (WNP-1 (WD ποίᾳ) (ND ἡμέρᾳ))
            (C 0)
            (IP-SUB (NP-TMP *T*-1)
                    (NP-SBJ (D ὁ)
                            (N κύριος)
                            (NP-ATR (PRO$ ὑμῶν)))
                    (VBPP-IMPF ἔρχεται)))                                
                                

    (Matthew 24.42)

WPP

WPP is used for pied-piping. There are no empty WPPs in the current corpora. (See documentation for the Penn Parsed Corpora of Historical English for why empty WPPs were annotated in the PPCME1.)

(CP-REL (WPP-1 (P ἐν)
               (NP (WPROD ᾧ)))
        (C 0)
        (IP-SUB (PP *T*-1)
                (NP-SBJ *pro*)
                (VBD-AOR εὐδόκησα)))                    
                

(Matthew 3.17)

Prepositions immediately dominated by WPP are tagged WP to facilitate the proper creation of traces for these under certain kinds of movement.

(NP-LFD (CP-FRL (WPP (WP-1 εἰς)
                     (WNP (WDA-2 ἣν)))
			    (CLPRT δ')
                (C 0)
                (IP-SUB (AN ἂν)
                        (PP (P *T*-1)
                            (NP (DA *T*-2)
                                (NA (NA πόλιν) (CONJ ἢ) (NA κώμην))))
                      (NP-SBJ *pro*)
                      (VBS-AOR εἰσέλθητε))))             
                

(Matthew 10.11)

WQP

In the Penn Parsed Corpora of Historical English, a WQP consists of a quantifier modified by how or an interrogative quantifier. It is always contained within some other wh- phrase, in the same way as a measure QP is always contained within some other phrase.

In the Greek corpora, a WQP can also consist of an interrogative quantifier that is extracted from within the NP that it modifies.

  • WQP extracted from NP:

    (CP-QUE-SPE (WQP-1 (WADJA Πόσους))
                (C 0)
                (IP-SUB (NP-SBJ *pro*)
                        (VBP-IMPF ἔχετε)
                        (NP-OB1 (QP *T*-1)
                                (NSA ἄρτους))))                    
                    

    (Mark 8.5)

  • WQP contained within another wh- phrase:

    ( (CP-QUE-SPE (CODE {VS:8_20})
                  (CP-ADV (C ὅτε)
                          (IP-SUB (NP-OB1 (DSA τοὺς) (NUM ἑπτὰ))
                                  (PP (P εἰς)
                                      (NP (DSA τοὺς) (ADJA τετρακισχιλίους)))))
                  (, ,)
                  (WNP-1 (WQP (WADJ$ πόσων) (NS$ σφυρίδων))
                         (NSA πληρώματα)
                         (NP-PAR (NS$ κλασμάτων)))
                  (C 0)
                  (IP-SUB (NP-OB1 *T*-1)
                          (NP-SBJ *pro*)
                          (VBD-AOR ἤρατε))
                  (. ;))
      (ID GreekNT,Mark))                                
                                

Non-Structural Labels

CODE

The label CODE is used for milestones, textual markup information, and similar information such as page numbers. Like punctuation, it is attached as high as possible. See Chapter 1, General Introduction for examples and discussion.

Foreign Language Passages

Foreign language passages longer than a single word are labeled with the name of the language. Although the internal syntax of foreign language passages is treated as a black box, their syntactic function in the Greek sentence is indicated in the ordinary way. Quotations are labeled QTP.

QTP

The label QTP surrounds non-sentential and foreign language quotations. It is also used to distinguish “mention” from “use”. It is a black box that indicates that the interior structure of a string is not relevant to the clause. Sentence fragments that are direct speech are labeled QTP rather than FRAG. For full clauses that are direct speech, see the section called “Direct Speech”.



[2] See the Appendix (pp. 694-703) in Hansen and Quinn’s Greek: An Intensive Course for descriptions of these terms.

Chapter 7. Conjunction

Word-Level Conjunction

Constituents that are labeled only at the word level (i.e., verbs) conjoin at the word level. When an active verb is conjoined with a middle-passive verb, the appropriate verbal POS tag with an -X appended is used to show that the voice of the conjoined verb is unknown. It is not possible to conjoin—at the word level—two verbs that are marked differently for tense (or aspect). See the section called “Conjunction of Unlike Categories” for what to do in this situation.

  • word-level verbal conjunction:

    (VBP-IMPF (VBP-IMPF ἀκούετε) (CONJ καὶ) (VBP-IMPF βλέπετε))                            
                            

    (Matthew 11.4)

  • word-level verbal conjunction of unlike-voiced verbs:

    (VBOX-AOR (VBOP-AOR κύσαιτο) (CONJ καὶ) (VBO-AOR τέκοι))                            
                            

    (ID Gortyn, LawCode.61)

For simplicity, categories that project phrases also conjoin at the word level, provided that all conjuncts consist of a single word. Since the constituenthood of the conjunction structure is already indicated by a phrasal tag, no additional word-level tag surrounds the structure, unlike in the case of verbs above.

  • (NP (NSA ἡγεμόνας) (CLPRT δὲ) (CONJ καὶ) (NSA βασιλεῖς))                        
                        

    (Matthew 10.18)

  • (ADVP-TMP (ADV σήμερον) (CONJ καὶ) (ADV αὔριον))                            
                            

    (Luke 13.32)

Phrase-Level Conjunction

When any of the conjuncts consists of more than a single word, the second and following conjuncts are labeled as CONJPs, headed by the conjunction, and adjoined to the first conjunct at the phrasal level, as illustrated in the following schema:

(XP (XP (X first-conjunct))
    (CONJP (CONJ conjunction)
           (YP (Y second-conjunct)))
    (CONJP (CONJ conjunction)
           (ZP (Z third-conjunct) (Z more-of-third-conjunct))))                
            

All phrasal categories are conjoined in the same way. Conjunction of clauses, which follows essentially the same pattern but presents special difficulties, is discussed separately below.

  • (NP-OB1 (NP (DA τὸ) (NA παιδίον))
            (CONJP (CONJ καὶ)
                   (NP (DA τὴν)
                       (NA μητέρα)
                       (NP-ATR (PRO$ αὐτοῦ)))))                            
                            

    (Matthew 2.13)

  • (PP (PP (P ἐν)
            (NP (DSD ταῖς) (NSD συναγωγαῖς)))
        (CONJP (CONJ καὶ)
               (PP (P ἐν)
                   (NP (DSD ταῖς) (NSD ῥύμαις)))))                            
                            

    (Matthew 6.2)

CONJPs with Empty Heads

Instances of word-level conjunction generally contain at least one overt conjunction. In the small number of cases where this is not the case, CONJPs are added to make the structure clear and distinguish these cases from noun compounds.

(NODE (NP-SBJ (NS abbeyes)
              (, .)
              (CONJP (NS priories))
              (, .)
              (CONJP (NS bernes)))
      (ID CMAYENBI,30.488))                
            

All clear cases of phrasal conjunction, including ones without an overt conjunction, contain CONJP. This is most common in cases with more than two conjuncts, where an overt conjunction appears only before the final conjunct. But there are also cases without any explicit conjunction at all.

  • more than two conjuncts:

    (NP-OB1 (NP (ADJA χωλούς))
            (, ,)
            (CONJP (NP (ADJA κυλλούς)))
            (, ,)
            (CONJP (NP (ADJA τυφλούς)))
            (, ,)
            (CONJP (NP (ADJA κωφούς)))
            (, ,)
            (CONJP (CONJ καὶ)
                   (NP (ADJA ἑτέρους) (QA πολλούς))))                            
                            

    (Matthew 15.30)

  • no explicit conjunction:

    (NP-SBJ (NP (NUM μία) (N ποίμνη))
            (, ,)
            (CONJP (NP (NUM εἷς) (N ποιμήν))))                            
                            

    (John 10.16)

In general, sets of constituents (NPs, PPs, etc.) acting as a unit are treated either as appositive or conjoined, and this is indicated either by the presence of the extended tag -PRN (see the section called “Appositive or Parenthetical”), or by the presence of CONJP on the second (and following) constituents. The immediately following structure is therefore illegal:

(NODE (WNP-2 (WNP (WD what) (N mede))     ← illegal structure, no -PRN or CONJP
             (, ,)
             (WNP (WD what) (N blysse))
             (, ,)
             (WNP (WD what) (N coroune))))

(NODE (WNP-2 (WNP (WD what) (N mede))     ← corrected structure, with CONJPs
             (, ,)
             (CONJP (WNP (WD what) (N blysse)))
             (, ,)
             (CONJP (WNP (WD what) (N coroune))))
      (ID CMAELR3,26.14))                
            

Conjunction of Unlike Categories

Conjuncts are generally of the same category, but not always. In the case of unlike categories, the category enclosing the entire conjunction structure is, by default, that of the first conjunct.

(NP-OB2 (NP (DD τῷ)
            (RRC (VPRPD-IMPF δαιμονιζομένῳ)))
        (CONJP (CONJ καὶ)
               (PP (P περὶ)
                   (NP (DS$ τῶν) (NS$ χοίρων)))))                
            

(Mark 5.16)

When two words that conjoin at the word level have different labels (unless they are verbs that differ only with respect to voice—see the section called “Word-Level Conjunction” above), a CONJP is added. In most cases of single-word conjunction, the difference between conjoining like and unlike categories is an artifact of our system of labeling rather than reflecting true conjunction of unlike categories.

(NODE (DON (DON done)
           (CONJP (CONJ and)
                  (VBN caused)))
      (ID CMMALORY,65.2185))                
            

And a Greek example:

( (IP-MAT-SPE (CODE {VS:12_44a})
              (CONJ καὶ-καί)
              (IP-PPL (VPR-AOR ἐλθὸν-ἔρχομαι))
              (NP-SBJ *pro*)
              (VBP-IMPF εὑρίσκει-εὑρίσκω)
              (IP-SMC (NP-SBJ *)
                      (VPRA-IMPF (VPRA-IMPF σχολάζοντα-σχολάζω)
                                 (CONJP (CONJ καὶ-καί)
                                        (VPRPA-PRF-PASS σεσαρωμένον-σαρόω))
                                 (CONJP (CONJ καὶ-καί)
                                        (VPRPA-PRF-PASS κεκοσμημένον-κοσμέω))))
              (. .))
  (ID GreekNT,Matthew))                
            

Cases involving the conjunction of unlike categories that result in morphological mismatches are treated differently; see the section called “Backwards Gapping”.

Recursive Conjunction Structures

Conjunction structures can be recursive. We attempt to represent the (perceived) intent of the text, using various clues such as the distribution of conjunctions, punctuation, syntactic parallelism, and alliteration. In the absence of evidence for grouping, the default is to give a non-recursive structure.

(NP-OB1 (NP (NEG μὴ)
            (DSA τοὺς)
            (NSA πόδας)
            (NP-ATR (CLPRO$ μου))
            (ADV μόνον))
        (CONJP (CONJ ἀλλὰ)
               (ADV καὶ)
               (NP (NP (DSA τὰς) (NSA χεῖρας))
                   (CONJP (CONJ καὶ)
                          (NP (DA τὴν) (NA κεφαλήν))))))                
            

(John 13.9)

Shared Modifiers

Shared Pre-Modifiers

See also the section called “Clausal Conjunction and Shared Modifiers”. When the first conjunct includes pre-head modifiers, it is possible in principle for conjunction to apply at the word level or at the phrase level. In order to indicate the (potential) attachment ambiguity, the second and following conjuncts are labeled (W)ADJX, (W)ADVX, or (W)NX.

[Important] Important

(W)ADJX, (W)ADVX, and (W)NX are used even when the attachment ambiguity is resolved by the context.

Since the second conjunct in these cases is never taken to be a full phrase on its own, the first conjunct is never surrounded by phrasal brackets of its own.

(NP-SBJ (Q πολλοὶ)
        (NS τελῶναι)
        (CONJP (CONJ καὶ)
               (NX (ADJ ἁμαρτωλοὶ))))                    
                

(Mark 2.15)

This convention applies only to ADJX, ADVX, NX, and their wh- counterparts. It does not extend to PP.

(NODE (PP (PP (FP bote)
              (P in)
              (NP (ONE oon)))
          (, ,)
          (CONJP (PP (P at)
                     (NP (ONE oon))))
          (CONJP (CONJ and)
                 (PP (P wit)
                     (NP (ONE oon)))))
      (ID CMAELR3,35.268))                    
                

Shared Post-Modifiers

See also the section called “Clausal Conjunction and Shared Modifiers”. In keeping with our general principles, shared post-modifiers are attached as high as possible.

(NP-SBJ (NP (NS σημεῖα))
        (CONJP (CONJ καὶ)
               (NP (NS τέρατα)))
        (Q πολλὰ))                    
                

(Acts 5.12)

Conjunction and Negation

[Important] Important

In both of the following constructions, not is bracketed as low as possible, reflecting the fact that it forms a constituent with the phrase that it negates. In other words, not is never treated as a correlative conjunction or as part of one.

(XP (XP (X first-conjunct))
    (CONJP (YP (NEG not) (Y second-conjunct))))     ← X NOT Y

(XP (XP (X first-conjunct))
    (CONJP (CONJ conjunction)
           (YP (NEG not) (Y second-conjunct))))     ← X AND/BUT NOT Y            
        

  • X not Y, X AND not Y, X BUT not Y:

    (NP (NP (ND καινότητι)
            (NP-ATR (N$ πνεύματος)))
        (CONJP (CONJ καὶ)
               (NP (NEG οὐ)
                   (ND παλαιότητι)
                   (NP-ATR (N$ γράμματος)))))                            
                            

    (Romans 7.6)

  • not (only) X BUT (also) Y

    [Important] Important

    Also, unlike not (only), attaches as a sister to the conjunction.

    :

    (NP-OB1 (NP (NEG μὴ)
                (DSA τοὺς)
                (NSA πόδας)
                (NP-ATR (CLPRO$ μου))
                (ADV μόνον))
            (CONJP (CONJ ἀλλὰ)
                   (ADV καὶ)
                   (NP (NP (DSA τὰς) (NSA χεῖρας))
                       (CONJP (CONJ καὶ)
                              (NP (DA τὴν) (NA κεφαλήν))))))                            
                            

    (John 13.9)

Extraposed CONJPs

Not infrequently, the second (or following) conjunct of a conjoined phrase is extraposed. In such cases, the extraposed conjunct is coindexed with an *ICH* trace. This is one of the few cases of rightward movement allowed in the corpora. See the section called “Rightward Extraposition” for extensive discussion.

( (IP-MAT (CODE {VS:21_12})
          (CONJ0 καὶ)
          (NP-OB1 (NP (DSA τὰς)
                      (NSA τραπέζας)
                      (NP-ATR (DS$ τῶν) (NS$ κολλυβιστῶν)))
                  (CONJP *ICH*-1))
          (NP-SBJ *con*)
          (VBD κατέστρεψεν)
          (CONJP-1 (CONJ καὶ)
                   (NP (DSA τὰς)
                       (NSA καθέδρας)
                       (NP-ATR (DS$ τῶν)
                               (RRC (VPR$-IMPF πωλούντων)
                                    (NP-OB1 (DSA τὰς) (NSA περιστεράς))))))
          (, ,))
  (ID GreekNT,Matthew))                
            

Extraposition of a second (or following) conjunct from a phrase consisting of conjoined single words is treated in the same way. Although for simplicity’s sake, these cases normally do not contain a CONJP, when extraposition is involved, a CONJP is introduced (motivated by the fact of extraposition itself). Apart from the CONJP, no additional phrasal nodes are introduced. As with the phrasal case just discussed, the extraposed CONJP is coindexed with an *ICH* trace.

(IP-INF-COM (NP-OBQ (ND θεῷ)
                    (CONJP *ICH*-1))
            (VBN-IMPF δουλεύειν)
            (CONJP-1 (CONJ καὶ)
                     (ND μαμωνᾷ)))                
            

(Matthew 6.24)

Correlative Conjunction

Word-Level Correlative Conjunction

Single-word cases are bracketed flat, as in the general case.

(NP-OB1 (CONJ καὶ) (NA ψυχὴν) (CONJ καὶ) (NA σῶμα))                   
                

(Matthew 10.28)

Phrase-Level Correlative Conjunction

In phrase-level correlative conjunction, the first conjunction is treated as a sister of the first conjunct:

(NP-OB1 (CONJ καὶ)
        (NP (DSA τὰς) (NSA χεῖρας))
        (CONJP (CONJ καὶ)
               (NP (DA τὴν) (NA πλευρὰν))))                    
                

(John 20.20)

Floated Conjunctions

Floated conjunctions are attached as high as possible, usually at the IP level, but lower if necessary. Where possible, they are coindexed with an *ICH* trace in the position where they are interpreted.

(NODE (IP-SUB (NP-SBJ (D +te) (N kyng))
              (MD schulde)
              (CONJ-1 no+ter)
              (VB bere)
              (NP-OB1 (CONJ *ICH*-1) (N swerd) (CONJ ne) (N knyf)))
      (ID CMPOLYCH,VI,445.3266))                    
                

Clausal Conjunction

Conjunction of Matrix IPs

See also the section called “Conjoined Direct Speech”. Main clauses are separated even when conjoined. The clause-initial CONJ in the second and subsequent clause(s) is treated as a constituent of the clause, and an empty subject (NP-SBJ *con*) is included in the second and subsequent clauses. An (NP-SBJ *con*) should only be used in Greek when a true conjunction is involved—e.g., καί—and not when sentential particles (δέ, οὖν) are involved.

( (IP-MAT (CODE {VS:21_12})
          (CONJ0 Καὶ)
          (VBD-AOR εἰσῆλθεν)
          (NP-SBJ (NPR Ἰησοῦς))
          (PP (P εἰς)
              (NP (DA τὸ) (ADJA ἱερόν)))
          (, ,))
  (ID GreekNT,Matthew))

( (IP-MAT (CODE {VS:21_12a})
          (CONJ0 καὶ)
          (NP-SBJ *con*)
          (VBD-AOR ἐξέβαλεν)
          (NP-OB1 (QA πάντας)
                  (DSA τοὺς)
                  (RRC (VPRA-IMPF (VPRA-IMPF πωλοῦντας)
                                  (CONJ καὶ)
                                  (VPRA-IMPF ἀγοράζοντας))
                       (PP (P ἐν)
                           (NP (DD τῷ) (ADJD ἱερῷ))))))
  (ID GreekNT,Matthew))                    
                

Conjunction of Embedded IPs

The absence of a VP level in our annotation system forces us to treat both VP and IP conjunction in the same way—namely, as IP conjunction. An empty subject (NP-SBJ *con*) is included in any non-first clausal conjunct when its subject is elided.

(CP-ADV (C ὅτι)
        (IP-SUB (IP-SUB (NP-SBJ *pro*)
                        (VBD-AOR ἔκρυψας)
                        (NP-OB1 (DSA ταῦτα))
                        (PP (P ἀπὸ)
                            (NP (ADJ$ σοφῶν) (CONJ καὶ) (ADJ$ συνετῶν))))
                (, ,)
                (CONJP (CONJ καὶ)
                       (IP-SUB (NP-SBJ *con*)
                               (VBD-AOR ἀπεκάλυψας)
                               (NP-OB1 (PROA αὐτὰ))
                               (NP-OB2 (ADJD νηπίοις))))))                    
                

(Matthew 11.25)

For cases where more than the subject is elided from the second conjunct, see the section called “Gapping”.

Right-Node Raising

Right-node raising is treated differently depending on whether the right-node-raised constituent is smaller than NP or not. In the first case, the right-node-raised constituent will be immediately dominated by NP; in the second case, by some type of IP. In the examples below, the right-node-raised constituents are italicized.

  • an enlarging and a restraining power ← right-node raised constituent = N

  • She asked for, but she did not get, a raise. ← right-node raised constituent = NP

The noun phrase-internal type is treated like ordinary conjunction, except that the head of the first conjunct is not expressed.

(NP (NP (D an) (VAG enlarging))
    (CONJP (CONJ and)
           (NP (D a) (VAG restraining) (N power))))                    
                

In the clausal type, the incomplete IP is marked as a parenthetical (IP-MAT-PRN, IP-SUB-PRN, etc.). In addition, the incomplete clausal material immediately preceding the right-node raised constituent is coindexed (with a index preceded by an equal sign) to the clause from which its missing parts can be reconstructed. Within such incomplete clauses, no reconstruction at all is done. That is, no empty categories are included, and no information beyond part-of-speech and minimal phrasal labelling (including the annotation of grammatical functions) is provided. PPs in such clauses need not include a prepositional object.

( (IP-MAT-1 (NP-SBJ (PRO She))                  ← complete clause with right-node raised constituent
            (VBD asked)
            (PP (P for)
                (IP-MAT-PRN=1 (CONJ but)        ← index with = on incomplete clause
                              (NP-SBJ (PRO she))
                              (DOD did)
                              (NEG not)
                              (VB get))
                (NP (D a) (N raise)))))                    
                

Gapping

The treatment of gapping resembles that of right-node raising. The gapped clause is coindexed with the complete clause from which it can be reconstructed. If the gapped clause is sentence-final, it does not receive a -PRN dash tag.

( (IP-MAT (CODE {VS:2_3})
          (IP-MAT-1 (IP-PPL (VPR-AOR Ἀκούσας))
                    (CLPRT δὲ)
                    (NP-SBJ (D ὁ)
                            (N βασιλεὺς)
                            (NP-PRN (NPR Ἡρῴδης)))
                    (VBDP-AOR-PASS ἐταράχθη))
          (CONJP (CONJ καὶ)
                 (IP-MAT=1 (NP-SBJ (Q πᾶσα) (NPR Ἰεροσόλυμα))
                           (PP (P μετ')
                               (NP (PRO$ αὐτοῦ)))))
          (, ,))
  (ID GreekNT,Matthew))                    
                

In cases that are ambiguous between gapping and ordinary conjunction, the default is usually ordinary conjunction. A common case is the not (only) X but Y construction. In general, gapped clauses should dominate more than one constituent (not counting any conjunctions). However, in cases in which an extraposed CONJP is construed with the subject, the default is gapping. This is because it is possible for the extraposed CONJP to be construed with a silent subject or a subject yielding a mismatch in subject-verb agreement, in which cases an ordinary conjunction analysis is difficult or impossible to implement.

(CP-ADV (C ὅτε)                                             ← like this
        (IP-SUB (IP-SUB-2 (NP-SBJ *pro*)
                          (VBD-AOR ἐπείνασεν))
                (CONJP (CONJ καὶ)
                       (IP-SUB=2 (NP-SBJ (DS οἱ)
                                         (PP (P μετ')
                                             (NP (PRO$ αὐτοῦ))))))))

(CP-ADV (C ὅτε)                                             ← not like this
        (IP-SUB (NP-SBJ (PRO *pro*)
                        (CONJP *ICH*-1))
                (VBD-AOR ἐπείνασεν)
                (CONJP-1 (CONJ καὶ)
                         (NP (DS οἱ)
                             (PP (P μετ')
                                 (NP (PRO$ αὐτοῦ)))))))                    
                

(Matthew 12.3)

If the gapped clause is clause-internal, it is labeled as a parenthetical, with the exact label (IP-MAT-PRN, IP- SUB-PRN, etc.) depending on how its dominating clause is labeled, just as in the case of right-node raising. Gapping of this type is often used even without strict equivalence between the complete and the gapped clause. If the relationship is extremely loose, however, the second clause may be dominated by FRAG.

( (IP-MAT-1 (CONJ0 And)
            (NP-SBJ (NS men))
            (VBP comen)
            (PP (P fro)
                (ADJP (ADJ fer)))
            (PP (P by)
                (NP (N watre)))
            (PP (P in)
                (NP (NS schippes)))
            (IP-MAT-PRN=1 (CONJ and)
                          (PP (P be)
                              (NP (N londe)))
                          (PP (P with)
                              (NP (NS cartes))))
            (IP-INF-PRP (FOR for) (TO to)
                        (VB fetten)
                        (PP (P of)
                            (NP (D +tat) (N grauell)))))
  (ID CMMANDEV,19.464))                    
                

Backwards Gapping

In backwards gapping, the constituent that counts as gapped is the first conjunct (rather than the second conjunct, as with ordinary gapping). The conjunction is included under CONJP in either type. Backwards gapping is used to annotate ‘whether or not’ clauses as well as the conjunction of unlike categories that result in morphological mismatches.

(CP-THT (C ὅτι)
        (IP-SUB (IP-SUB=1 (NP-SBJ (PRO ἐγὼ))
                          (PP (P ἐν)
                              (NP (DD τῷ) (ND πατρὶ))))
                (CONJP (CONJ καὶ)
                       (IP-SUB-1 (NP-SBJ (D ὁ) (N πατὴρ))
                                 (PP (P ἐν)
                                     (NP (PROD ἐμοί)))
                                 (BEP-IMPF ἐστιν)))))                    
                

(John 14.10)

Conjoined Direct Speech

After a verb of saying in a matrix clause, all conjoined clauses containing direct speech after the first one are treated as separate tokens.

( (IP-MAT-SPE (CODE {VS:5_26})
              (INTJ ἀμὴν)
              (NP-SBJ *pro*)
              (VBP-IMPF λέγω)
              (NP-OB2 (CLPROD σοι))
              (, ,)
              (IP-MAT-SPE (NEG οὐ)
                          (NEG μὴ)
                          (NP-SBJ *pro*)
                          (VBS-AOR ἐξέλθῃς)
                          (ADVP-DIR (ADV ἐκεῖθεν))
                          (CP-ADV (C ἕως)
                                  (IP-SUB (PRTG ἂν)
                                          (NP-SBJ *pro*)
                                          (VBS-AOR ἀποδῷς)
                                          (NP-OB1 (DA τὸν) (ADJA ἔσχατον) (NA κοδράντην)))))
              (. .))
  (ID GreekNT,Matthew))

( (IP-MAT-SPE (CODE {VS:5_27})
              (NP-SBJ *pro*)
              (VBD-AOR Ἠκούσατε)
              (CP-THT (C ὅτι)
                      (IP-SUB (NP-SBJ *pro*)
                              (VBDP-AOR-PASS ἐρρέθη)
                              (IP-MAT-SPE (NEG Οὐ)
                                          (NP-SBJ *pro*)
                                          (VBP-FUT μοιχεύσεις))))
              (. .))
  (ID GreekNT,Matthew))                    
                

On the other hand, in embedded clauses, including parentheticals, all conjoined direct speech remains part of the same token.

( (IP-MAT-SPE (CODE {VS:5_33})
              (ADVP (ADV Πάλιν))
              (NP-SBJ *pro*)
              (VBD-AOR ἠκούσατε)
              (CP-THT (C ὅτι)
                      (IP-SUB (NP-SBJ *pro*)
                              (VBDP-AOR-PASS ἐρρέθη)
                              (NP-OB2 (DSD τοῖς) (ADJD ἀρχαίοις))
                              (IP-MAT-SPE (IP-MAT-SPE (NEG Οὐκ)
                                                      (NP-SBJ *pro*)
                                                      (VBP-FUT ἐπιορκήσεις))
                                          (, ,)
                                          (CONJP (IP-MAT-SPE (NP-SBJ *con*)
                                                             (VBP-FUT ἀποδώσεις)
                                                             (CLPRT δὲ)
                                                             (NP-OB2 (DD τῷ) (ND κυρίῳ))
                                                             (NP-OB1 (DSA τοὺς)
                                                                     (NSA ὅρκους)
                                                                     (NP-ATR (CLPRO$ σου))))))))
              (. .))
  (ID GreekNT,Matthew))                    
                

However, because of the prevalence of introducing direct speech with the participle λέγων in the GNT, conjoined direct speech embedded within participial or infinitival phrases in the PPCNTG is treated the same as conjoined direct speech introduced by a verb of saying in a matrix clause: all conjoined clauses containing direct speech after the first one are treated as separate tokens.

( (IP-MAT (CODE {VS:13_3})
          (CONJ0 καὶ)
          (NP-SBJ *pro*)
          (VBD-AOR ἐλάλησεν)
          (NP-OB2 (PROD αὐτοῖς))
          (NP-OB1 (QA πολλὰ))
          (PP (P ἐν)
              (NP (NSD παραβολαῖς)))
          (IP-PPL (VPR-IMPF λέγων)
                  (IP-MAT-SPE (INTJ Ἰδοὺ)
                              (VBD-AOR ἐξῆλθεν)
                              (NP-SBJ (D ὁ) (VPR-IMPF σπείρων))
                              (CP-PRP (C τοῦ)
                                      (IP-INF (VBN-IMPF σπείρειν)))))
          (. .))
  (ID GreekNT,Matthew))

( (IP-MAT-SPE (CODE {VS:13_4})
              (CONJ0 καὶ)
              (PP (P ἐν)
                  (NP (DD τῷ)
                      (IP-INF (VBN-IMPF σπείρειν)
                              (NP-SBJ (PROA αὐτὸν))
                              (NP-OB1 (CP-FRL (WNP-1 (WPROA ἃ)) 
                                              (CLPRT μὲν)
                                              (C 0)
                                              (IP-SUB (NP-SBJ *T*-1)
                                                      (VBD-AOR ἔπεσεν)
                                                      (PP (P παρὰ)
                                                          (NP (DA τὴν) (NA ὁδόν)))))))))
              (, ,)
              (CONJ καὶ)
              (IP-PPL (VPR-AOR ἐλθόντα))
              (NP-SBJ (DS τὰ) (ADJ πετεινὰ))
              (VBD-AOR κατέφαγεν)
              (NP-OB1 (PROA αὐτα))
              (. .))
  (ID GreekNT,Matthew))                    
                

Conjunction of CPs

Conjunction of embedded clauses is treated as CP conjunction (rather than as IP conjunction) only if the second conjunct has overt CP material, such as a wh- phrase or complementizer.

( (IP-IMP-SPE (CODE {VS:6_19})
              (NEG Μὴ)
              (VBI-IMPF θησαυρίζετε)
              (NP-OB2 (PROD ὑμῖν))
              (NP-OB1 (NSA θησαυροὺς))
              (PP (P ἐπὶ)
                  (NP (D$ τῆς)
                      (N$ γῆς)
                      (, ,)
                      (CP-REL (CP-REL (WADVP-1 (WADV ὅπου))
                                      (C 0)
                                      (IP-SUB (ADVP-LOC *T*-1)
                                              (NP-SBJ (N σὴς) (CONJ καὶ) (N βρῶσις))
                                              (VBP-IMPF ἀφανίζει)))
                              (, ,)
                              (CONJP (CONJ καὶ)
                                     (CP-REL (WADVP-2 (WADV ὅπου))
                                             (C 0)
                                             (IP-SUB (ADVP-LOC *T*-2)
                                                     (NP-SBJ (NS κλέπται))
                                                     (VBP-IMPF (VBP-IMPF διορύσσουσιν)
                                                               (CONJ καὶ)
                                                               (VBP-IMPF κλέπτουσιν))))))))
              (. .))
  (ID GreekNT,Matthew))                    
                

Clausal Conjunction and Shared Modifiers

Because of the way that we split matrix clauses, shared elements interpreted as belonging to both clauses must be annotated as belonging to one clause or the other, or one of the clauses must be treated as parenthetical. For adjuncts, we take the former approach (see examples below); for complements, the latter (see the section called “Conjunction of Embedded IPs”). As a result of our convention, pre-sentential adjuncts (most commonly, clause-initial WHEN and IF clauses) appear as constituents of the first conjunct only, while post-sentential adjuncts appear as constituents of the second conjunct only.

( (IP-IMP-SPE (CODE {VS:6_6})
              (CP-ADV (NP-LFD (PRO σὺ))                     ← potential shared adjunct
                      (CLPRT δὲ)
                      (C ὅταν)
                      (IP-SUB (NP-SBJ-RSP *pro*)
                              (VBSP-IMPF προσεύχῃ)))
              (, ,)
              (VBI-AOR εἴσελθε)
              (PP (P εἰς)
                  (NP (DA τὸ)
                      (NA ταμεῖόν)
                      (NP-ATR (CLPRO$ σου)))))
  (ID GreekNT,Matthew))

( (IP-IMP-SPE (CODE {VS:6_6a})
              (CONJ0 καὶ)
              (IP-PPL (VPR-AOR κλείσας)
                      (NP-OB1 (DA τὴν)
                              (NA θύραν)
                              (NP-ATR (CLPRO$ σου))))
              (VBIP-AOR πρόσευξαι)
              (NP-OB2 (DD τῷ)
                      (ND πατρί)
                      (NP-ATR (CLPRO$ σου))
                      (NP-PRN (DD τῷ)
                              (PP (P ἐν)
                                  (NP (DD τῷ) (ADJD κρυπτῷ)))))
              (. .))
  (ID GreekNT,Matthew))                    
                

( (IP-IMP-SPE (CODE {VS:6_17})
              (NP-SBJ (PRO σὺ))
              (CLPRT δὲ)
              (IP-PPL (VPR-IMPF νηστεύων))
              (VBIP-AOR ἄλειψαί)
              (NP-OB1 (NP-ATR (CLPRO$ σου))
                      (DA τὴν)
                      (NA κεφαλὴν)))
  (ID GreekNT,Matthew))

( (IP-IMP-SPE (CODE {VS:6_17a})
              (CONJ0 καὶ)
              (NP-OB1 (DA τὸ)
                      (NA πρόσωπόν)
                      (NP-ATR (CLPRO$ σου)))
              (VBIP-AOR νίψαι)
              (, ,)
              (CODE {VS:6_18})
              (CP-ADV (C ὅπως)                              ← potential shared adjunct
                      (IP-SUB (NEG μὴ)
                              (NP-SBJ *pro*)
                              (VBSP-AOR φανῇς)
                              (NP-OB2 (NP (DSD τοῖς) (NSD ἀνθρώποις))
                                      (CONJP *ICH*-1))
                              (IP-PPL-COM (VPR-IMPF νηστεύων))
                              (CONJP-1 (CONJ ἀλλὰ)
                                       (NP (DD τῷ)
                                           (ND πατρί)
                                           (NP-ATR (CLPRO$ σου))
                                           (NP-PRN (DD τῷ)
                                                   (PP (P ἐν)
                                                       (NP (DD τῷ) (ADJD κρυφαίῳ))))))))
              (. .))
  (ID GreekNT,Matthew))                    
                

Shared adjuncts in embedded conjoined clauses are treated analogously.

(IP-SUB (IP-SUB-5 (NP-SBJ *T*-4)
                  (CP-ADV (C ὅτε)                           ← potential shared adjunct
                          (IP-SUB (NP-SBJ *pro*)
                                  (BED-IMPF ἦν)
                                  (PP (P ἐν)
                                      (NP (DD τῇ) (NPRD Γαλιλαίᾳ)))))
                  (VBD-IMPF ἠκολούθουν)
                  (NP-OBQ (PROD αὐτῷ)))
        (CONJP (CONJ καὶ)
               (IP-SUB=5 (NP-SBJ *T*-4)
                         (VBD-IMPF διηκόνουν)
                         (NP-OBQ (PROD αὐτῷ)))))                    
                

(Mark 15.41)

Chapter 8. Empty Categories

In general, empty categories are treated the same way as overt ones except that they do not contain lexical material.

Empty Subjects

There are three types of empty subjects:

Due to the complete lack of overt expletive subjects, we assume that Greek does not contain empty expletive subjects either.

[Important] Important

All empty subjects should be placed immediately preceding the tensed verb in the clause.

If the clause does not contain a tensed verb (e.g., because of gapping), then the empty subject should come before all overt elements in the clause but after traces of wh- movement.

Subjects Elided under Conjunction

Subjects elided under conjunction contain *con*. Subjects are obligatory in most types of clauses in our system, and so elided subjects are indicated both in main clause tokens and in conjoined subordinate clauses. The empty subject of a matrix clause must be co-referential with the subject of the previous token in order to be treated as *con* (the clause must contain an overt CONJ and not merely a conjunctive sentential particle such as δέ); otherwise, it is treated as *pro*. In principle, elided objects could be treated analogously, but they aren’t, since too many cases are too difficult to decide.

( (IP-MAT-SPE (CODE {VS:2_2a})
              (NP-SBJ *pro*)                                ← first token with pro subject
              (VBD-AOR εἴδομεν)
              (CLPRT γὰρ)
              (NP-OB1 (NP-ATR (PRO$ αὐτοῦ))
                      (DA τὸν)
                      (NA ἀστέρα))
              (PP (P ἐν)
                  (NP (DD τῇ) (ND ἀνατολῇ))))
  (ID GreekNT,Matthew))

( (IP-MAT-SPE (CODE {VS:2_2b})
              (CONJ0 καὶ)
              (NP-SBJ *con*)                                ← following token with subject elided
              (VBD-AOR ἤλθομεν)
              (IP-INF-PRP (VBN-AOR προσκυνῆσαι)
                          (NP-OBP (PROD αὐτῷ)))
              (. .))
  (ID GreekNT,Matthew))                    
                

(CP-THT (C ὅτι)
        (IP-SUB-SPE (IP-SUB (NP-SBJ (Q πολλοὶ))
                            (PP (P ἀπὸ)
                                (NP (NS$ ἀνατολῶν) (CONJ καὶ) (NS$ δυσμῶν)))
                            (VBP-FUT ἥξουσιν))
                    (CONJP (CONJ καὶ)
                           (IP-SUB (NP-SBJ *con*)           ← subject elided in conjoined clause
                                   (VBPP-FUT ἀνακλιθήσονται)
                                   (PP (P μετὰ)
                                       (NP (NPR$ Ἀβραὰμ) (CONJ καὶ) 
                                           (NPR$ Ἰσαὰκ) (CONJ καὶ) 
                                           (NPR$ Ἰακὼβ)))
                                   (PP (P ἐν)
                                       (NP (DD τῇ)
                                           (ND βασιλείᾳ)
                                           (NP-ATR (DS$ τῶν) (NS$ οὐρανῶν))))))))                    
                

(Matthew 8.11)

Arbitrary Subjects in ECM Infinitives

The arbitrary subject of an infinitive governed by an ECM verb is labeled *arb*.

( (IP-MAT (CONJ0 And)
          (ADVP (ADV soo))
          (PP (P by)
              (NP (D the) (N counceil)
                  (PP (P of)
                      (NP (NPR Merlyn)))))
          (NP-SBJ (D the) (N kyng))
          (VBD lete)
          (IP-INF (NP-SBJ *arb*)
                  (VB calle)
                  (NP-OB1 (PRO$ his) (NS barons))
                  (PP (P to)
                      (NP (N counceil))))
          (. ,))
          (ID CMMALORY,14.419))                    
                

pro-Drop Subjects

An empty subject that is not co-referential with the subject of the previous clause (unless the clauses are connected by a conjunctive sentential particle and not a true conjunction) contains *pro*. This represents pro-drop, which is (unsurprisingly, given that Greek is a language with very rich verbal morphology) common in Greek.

( (IP-MAT-SPE (CODE {VS:2_5})
              (ADVP (ADV οὕτως))
              (CLPRT γὰρ)
              (NP-SBJ *pro*)
              (VBPP-PRF-PASS γέγραπται)
              (PP (P διὰ)
                  (NP (D$ τοῦ) (N$ προφήτου))))
  (ID GreekNT,Matthew))                    
                

Traces

Several kinds of traces are indicated in our system, including traces of A-movement, of wh-movement, and of other A’-movement. All traces bear an index matching the index on the trace’s antecedent. Not all instances of movement are explicitly indicated. In particular, we do not distinguish raising from subject control, and infinitive clauses ordinarily do not contain subject traces. See the section called “Infinitival Clauses” for details. In general, only phrases undergo movement. The three exceptions to this rule are discussed elsewhere:

Traces of A-Movement

Traces of A-movement contain * (which is also used as a generic empty category, but in this case never indexed). Not all instances of A-movement are explicitly indicated. In particular, our annotation scheme does not ordinarily express the distinction between raising and control. A-movement is indicated only when the moved element is extracted from within another phrase (usually PP) or clause. Thus, a passivized NP that is extracted from a PP or lower clause (small clause or infinitive) is coindexed with a trace, but an ordinary passive subject is not.

(CP-ADV (C μή)
        (IP-SUB (CLPRT ποτε)
                (IP-SMC (NP-SBJ *-1)
                        (ADJP-PRD (ADV καὶ)
                                  (ADJ θεομάχοι)))
                (NP-SBJ-1 *pro*)
                (VBSP-AOR-PASS εὑρεθῆτε)))                    
                

(Acts 5.39)

Traces of Wh-Movement

Wh- traces contain *T*; the POS label on the trace indicates the category of the gap. See Chapter 10, Subordinate Clauses for clause types in which wh- movement is assumed.

(CP-QUE-SPE (WADVP-1 (WADV Ποῦ))
            (C 0)
            (IP-SUB-SPE (ADVP-LOC *T*-1)
                        (BEP-IMPF ἐστὶν)
                        (NP-SBJ (D ὁ)
                                (RRC (VPRP-AOR-PASS τεχθεὶς)
                                     (NP-SPR (N βασιλεὺς)
                                             (NP-COM (DS$ τῶν) (ADJ$ Ἰουδαίων)))))))                    
                

(Matthew 2.2)

In pied-piping contexts, embedded wh- constituents recursively project a wh- feature all the way up to the phrase in SpecCP. In exceptional cases, this results in the categories WIP (and even WCP). Where a wh- phrase is not associated with a trace (as is possible in pied-piping contexts), a dash tag is exceptionally added to the wh- phrase’s label.

(CP-REL (WPP-1 (P ἐξ)
               (WNP (WPRO$ ἧς)))
        (C 0)
        (IP-SUB (PP *T*-1)
                (VBDP-AOR-PASS ἐγεννήθη)
                (NP-SBJ (NPR Ἰησοῦς)
                        (NP-PRN (D ὁ)
                                (RRC (VPRP-IMPF-PASS λεγόμενος)
                                     (IP-SMC (NP-SBJ *)
                                             (NP-PRD (NPR Χριστός))))))))                    
                

(Matthew 1.16)

Pied-piping is assumed only if no other analysis is possible. In particular, if a token can be treated as involving simple extraction from an adjunct, that analysis is the default. This is similar in spirit to our treatment of tokens that are ambiguous between an analysis with extraction and one with a parenthetical clause.

(NODE (CP-REL (WNP-1 (WPRO wich))                      ← WNP, not WIP-PPL-ABS
              (C 0)
              (IP-SUB (IP-PPL-ABS (NP-SBJ *T*-1)       ← extraction out of adjunct
                                  (BAG being)
                                  (VAN graunted))
                      (, ,)
                      (ADVP (ADV so))
                      (NP-SBJ (PRO she))
                      (VBD came)
                      (ADJP-SPR (ADJ like)
                                (NP (D a) (N frend)))))
      (ID EDWARD,290.167))

(NODE (CP-REL (WIP-PPL-ABS-1 (WNP-SBJ (WPRO wich))     ← movement of adjunct (WRONG)
                             (BAG being)
                             (VAN graunted))
              (, ,)
              (C 0)
              (IP-SUB (IP-PPL-ABS *T*-1)
                      (ADVP (ADV so))
                      (NP-SBJ (PRO she))
                      (VBD came)
                      (ADJP-SPR (ADJ like)
                                (NP (D a) (N frend))))))
      (ID EDWARD,290.167))                    
                

Traces of Other A'-Movement

Traces of A’-movement other than wh- movement, both leftward and rightward, contain *ICH* (= interpret constituent here). *ICH* traces are only indicated when a constituent has been extracted from within another constituent, making the extraction site clear. The grammatical function of the moved phrase is indicated on its trace in the clause of origin, but the moved phrase itself has no dash tag since it has no function in the clause into which it has moved.

( (IP-INF-THT (CODE {VS:4_3})
              (NP-SBJ (NPRSA Ἕλληνας-Ἕλλην))
              (CLPRT δὲ-δέ)
              (PP (ADJ$-1 Λακεδαιμονίης-Λακεδαιμόνιος)
                  (P εἵνεκεν-ἕνεκα)
                  (NP (ADJ$ *ICH*-1)
                      (N$ γυναικὸς-γυνή)))
              (NP-OB1 (NA στόλον-στόλος) (ADJA μέγαν-μέγας))
              (VBN-AOR συναγεῖραι-συναγείρω))
  (ID Herodotus,Histories))          
                

( (IP-MAT (CODE {VS:3_4})
          (NY-1 (D ἡ)
                (CLPRT δὲ)
                (N τροφὴ))
          (BED-IMPF ἦν)
          (NP-SBJ (NY *ICH*-1)
                  (NP-ATR (PRO$ αὐτοῦ)))
          (NP-PRD (NP (NS ἀκρίδες))
                  (CONJP (CONJ καὶ)
                         (NP (N μέλι) (ADJ ἄγριον))))
          (. .))
  (ID GreekNT,Matthew))                    
                

Other Empty Categories

Empty Objects

Empty objects (apart from traces) are not indicated since they are too difficult to distinguish from intransitive uses of transitive verbs. This is despite some examples in Greek that seem almost certain to contain pro-dropped objects:

(CP-ADV (C+AN ἐπὰν)
        (IP-SUB (CLPRT δὲ)
                (NP-SBJ *pro*)
                (VBS-AOR εὕρητε)))
(VBI-AOR ἀπαγγείλατέ)
(NP-OB2 (CLPROD μοι))                    
                

(Matthew 2.8)

Empty Wh- Phrases and Complementizers

Empty wh- phrases and complementizers contain 0 (zero).

  • empty wh- phrase:

    (CP-ADV (C ἵνα)
            (IP-SUB (NP-SBJ *pro*)
                    (VBS-AOR εὕρωσιν)
                    (NP-OB1 (CP-FRL (WNP-1 0)
                                    (C 0)
                                    (IP-INF (NP-OB1 *T*-1)
                                            (VBN-IMPF κατηγορεῖν)
                                            (NP-OB2 (PRO$ αὐτοῦ)))))))                                
                                

    (Luke 6.7)

  • empty complementizer:

    (NP-SBJ (D ὁ)
            (N ἀστὴρ)
            (CP-REL (WNP-1 (WPROA ὃν))
                    (C 0)
                    (IP-SUB (NP-OB1 *T*-1)
                            (NP-SBJ *pro*)
                            (VBD-AOR εἶδον)
                            (PP (P ἐν)
                                (NP (DD τῇ) (ND ἀνατολῇ))))))                                
                                

    (Matthew 2.9)

A Generic Empty Category

In addition to serving as a trace of (some instances of) A-movement, (X *) is a generic empty category, used in at least the following cases.

  • verbs elided when gapping is not possible:

    ( (IP-MAT (CODE {VS:2_37a}
              (NP-SBJ *pro*)
              (VBD-AOR εἶπάν)
              (CLTE τε)
              (PP (P πρὸς)
                  (NP (NP (DA τὸν) (NPRA Πέτρον))
                      (CONJP (CONJ καὶ)
                             (NP (DSA τοὺς) (ADJA λοιποὺς) (NSA ἀποστόλους)))))
              (CP-QUE-SPE (WNP-1 (WPROA Τί))
                          (C 0)
                          (IP-SUB (NP-OB1 *T*-1)
                                  (NP-SBJ *pro*)
                                  (VBS-AOR ποιήσωμεν))
                          (, ,)
                          (CODE {VS:2_38})
                          (NP-VOC (NS ἄνδρες)
                                  (NP-PRN (NS ἀδελφοί))))
              (. ;))
      (ID GreekNT,Acts))
    
    ( (IP-MAT (CODE {VS:2_38a})
              (NP-SBJ (NPR Πέτρος))
              (CLPRT δὲ)
              (VBD-AOR *)                                       ← εἶπεν elided
              (PP (P πρὸς)
                  (NP (PROA αὐτούς)))
              (IP-IMP-SPE (VBI-AOR Μετανοήσατε))
              (, ,))
      (ID GreekNT,Acts))                                
                                

  • null copula:

    ( (IP-MAT-SPE (CODE {VS:5_9})
                  (INTJ ἰδοὺ)
                  (NP-SBJ (DS οἱ)
                          (NS πόδες)
                          (NP-ATR (DS$ τῶν)
                                  (RRC (VPR$-AOR θαψάντων)
                                       (NP-OB1 (DA τὸν)
                                               (NA ἄνδρα)
                                               (NP-ATR (CLPRO$ σου))))))
                  (BEP-IMPF *)
                  (PP (P ἐπὶ)
                      (NP (DD τῇ) (ND θύρᾳ))))
      (ID GreekNT,Acts))                                
                                

  • empty subjects of small clauses within a reduced relative clause (see also small clauses):

    (PP (P ἀπὸ)
        (NP (N$ ὄρους)
            (NP-PRN (D$ τοῦ)
                    (RRC (VPRP$-IMPF-PASS καλουμένου)
                         (IP-SMC (NP-SBJ *)
                                 (NP-PRD (N$ Ἐλαιῶνος)))))                                
                                

    (Acts 1.12)

The Position of Empty Categories

Since empty categories are not visible, their position in the string is often not determinable. We therefore follow the following rules concerning their position.

The Position of the Null Copula

The null copula should appear between the subject and the predicate. If there is only a subject, then the null copula should appear before the subject.

  • subject and predicate:

    ( (IP-MAT-SPE (CODE {VS:5_9})
                  (INTJ ἰδοὺ)
                  (NP-SBJ (DS οἱ)
                          (NS πόδες)
                          (NP-ATR (DS$ τῶν)
                                  (RRC (VPR$-AOR θαψάντων)
                                       (NP-OB1 (DA τὸν)
                                               (NA ἄνδρα)
                                               (NP-ATR (CLPRO$ σου))))))
                  (BEP-IMPF *)
                  (PP (P ἐπὶ)
                      (NP (DD τῇ) (ND θύρᾳ))))
      (ID GreekNT,Acts))                                
                                

  • subject only:

    ( (IP-MAT (CODE {VS:4_15})
              (ADVP-LOC (CP-FRL (WADVP-1 (WADV οὗ))
                                (CLPRT δὲ)
                                (C 0)
                                (IP-SUB (ADVP-LOC *T*-1)
                                        (NEG οὐκ)
                                        (BEP-IMPF ἔστιν)
                                        (NP-SBJ (N νόμος)))))
              (, ,)
              (NEG+CONJ οὐδὲ)
              (BEP-IMPF *)
              (NP-SBJ (N παράβασις))
              (. .))
      (ID GreekNT,Romans))                                
                                

The Position of Empty Subjects

In general, empty subjects are put in the position immediately preceding the tensed verb in the clause. If the clause does not have a tensed verb (because of gapping, ellipsis, etc.), then the empty subject comes first in the clause after conjunctions, traces, and left-dislocated elements. (No good example of the latter yet, but it does seem to be a logical possibility.)

( (IP-MAT-SPE (CODE {VS:1_21})
              (NP-SBJ *pro*)
              (VBPP-FUT τέξεται)
              (CLPRT δὲ)
              (NP-OB1 (NA υἱὸν)))
  (ID GreekNT,Matthew))                    
                

The Position of Traces

Almost all traces (both wh and non-wh) appear in constituent-initial position:

  • traces in IP:

    (PP (P ἕως)
        (NP (CP-FRL (WNP-1 (WPRO$ οὗ))
                    (C 0)
                    (IP-SUB (NP-TMP *T*-1)
                            (NP-SBJ *pro*)
                            (VBD-AOR ἔτεκεν)
                            (NP-OB1 (NA υἱόν))))))                                
                                

    (Matthew 1.25)

  • traces in ADJP:

    (CP-ADV (C εἰ)
            (CLPRT δὲ)
            (IP-SUB (NP-SBJ *pro*)
                    (BEP-IMPF *CL*-3)
                    (PP (P χωρίς)
                        (BEP-IMPF-CL-3 ἐστε)
                        (NP (N$ παιδείας)
                              (CP-REL (WNP-2 (WPRO$ ἧς))
                                      (C 0)
                                      (IP-SUB (ADJP-PRD (NP-COM *T*-2)
                                                        (ADJ μέτοχοι))
                                              (VBP-PRF γεγόνασι)
                                              (NP-SBJ (Q πάντες))))))))                                
                                

    (Hebrews 12.8)

  • traces of modifiers:

    (CP-QUE-SPE (WQP-1 (WADJA Πόσους))
                (C 0)
                (IP-SUB (NP-SBJ *pro*)
                        (VBP-IMPF ἔχετε)
                        (NP-OB1 (QP *T*-1)
                                (NSA ἄρτους))))                                
                                

    (Mark 8.5)

Resumptive Pronouns

When a wh- phrase is associated with a resumptive pronoun, the pronoun receives an index, just like a trace. In addition, its syntactic tag receives the extended tag -RSP. See also the section called “Left Dislocation”.

(CP-REL (WNP-1 (WPRO$ οὗ))
        (C 0)
        (IP-SUB (NEG οὐκ)
                (BEP-IMPF εἰμὶ)
                (NP-SBJ (PRO ἐγὼ))
                (ADJP-PRD (ADJ ἄξιος)
                          (CP-COM (C ἵνα)
                                  (IP-SUB (NP-SBJ *pro*)
                                          (VBS-AOR λύσω)
                                          (NP-OB1 (NP-ATR-RSP-1 (PRO$ αὐτοῦ))
                                                  (DA τὸν)
                                                  (NA ἱμάντα)
                                                  (NP-ATR (NP-ATR-RSP *ICH*-1)
                                                          (D$ τοῦ)
                                                          (N$ ὑποδήματος))))))))                            
                        

(John 1.27)

(CP-REL (WNP-1 (WPRO ὃν))
        (C 0)
        (IP-SUB (NP-SBJ *pro*)
                (VBD-AOR ἀνέπεμψά)
                (NP-OB2 (CLPROD σοι))
                (NP-OB1-RSP-1 (PROA αὐτόν))))    

(Philemon 1.12)

Chapter 9. Clause-Level Constituents

Adjective Phrase

Predicate Adjectives

Bare ADJPs at the sentence level are predicates of the copular verbs listed below or of small clauses. The Greek semantic equivalents of the following English verbs count as copular verbs:

APPEAR, BE, BECOME, CHANCE, CONTINUE, FALL, FEEL, GET, GROW, KEEP, LOOK, PLEAD (in legal sense), PROVE, REMAIN, SAVOR, SEEM, SHOW, SMELL, SOUND, STAY, TASTE, TURN, WAX

Predicate adjectives carry the dash tag -PRD.

  • ADJP-PRD at the sentence level:

    (NP-SBJ (D ὁ)
            (CLPRT δὲ)
            (RRC (PP (P ὀπίσω)
                     (NP (CLPRO$ μου)))
                 (VPRP-IMPF ἐρχόμενος)))
    (ADJP-PRD (ADJR ἰσχυρότερός)
              (NP-CMP (CLPRO$ μου)))
    (BEP-IMPF ἐστίν)                                
                                

    (Matthew 3.11)

  • ADJP-PRD in a small clause:

    (IP-PPL (ADJA-1 ἴσον)
            (NP-2 (PROA+SLF ἑαυτὸν))
            (VPR-IMPF ποιῶν)
            (IP-SMC (ADJP-PRD (ADJA *ICH*-1)
                              (NP-COM (DD τῷ) (ND θεῷ)))
                    (NP-SBJ *ICH*-2)))
                                

    (John 5.18)

Other Clause-Level ADJPs (ADJP-SPR)

The dash tag -SPR is mnemonic for ‘secondary predicate’; see the section called “Secondary Predicate NP” for a list of verbs that take secondary predicates. This dash tag is also appended to all clause-level ADJPs that are not the predicate of a copular verb or small clause. In addition to secondary predicates, ADJP-SPRs thus also include adjectival adjuncts (parallel to adjunct participial clauses).

(ADJP-PRD (ADJ καλόν))
(NP-OB2 (PROD σοί))
(BEP-IMPF ἐστιν)
(IP-INF-SBJ (VBN-AOR εἰσελθεῖν)
            (PP (P εἰς)
                (NP (DA τὴν) (NA ζωὴν)))
            (ADJP-SPR (ADJA κυλλὸν) (CONJ ἢ) (ADJA χωλόν)))                    
                

(Matthew 18.8)

( (IP-MAT-SPE (CODE {VS:4_28})
              (IP-MAT-1 (ADJP-SPR (ADJ αὐτομάτη))
                        (NP-SBJ (D ἡ) (N γῆ))
                        (VBP-IMPF καρποφορεῖ)
                        (, ,)
                        (ADVP-TMP (ADV πρῶτον))
                        (NP-OB1 (NA χόρτον)))
              (, ,)
              (CONJP (IP-MAT=1 (ADVP-TMP (ADV εἶτεν))
                               (NP-OB1 (NA στάχυν))))
              (, ,)
              (CONJP (IP-MAT=1 (ADVP-TMP (ADV εἶτεν))
                               (NP-OB1 (ADJA πλήρη)
                                       (NA σῖτον)
                                       (PP (P ἐν)
                                           (NP (DD τῷ) (ND στάχυϊ))))))
              (. .))
  (ID GreekNT,Mark))                    
                

When a secondary predicate modifies a noun that is not the subject of the sentence, coindexation is employed to indicate which noun phrase the secondary predicate modifies.

(CP-ADV-SPE (WADVP-1 (WADV-AN ὅταν-ὅταν))
            (C 0)
            (IP-SUB-SPE (ADVP-TMP *T*-1)
                        (NP-OB1-2 (PROA αὐτὸ-αὐτός))
                        (NP-SBJ *pro*)
                        (VBS-IMPF πίνω-πίνω)
                        (ADJP-SPR-2 (ADJA καινὸν-καινός))
                        (PP (P ἐν-ἐν)
                            (NP (DD τῇ-ὁ)
                                (ND βασιλείᾳ-βασιλεία)
                                (NP-ATR (D$ τοῦ-ὁ) (N$ θεοῦ-θεός))))))                    
                

(Mark 14.25)

Movement out of ADJP

Our annotation system does not fix the order of syntactic dependents with respect to the head in ADJPs (which contrast in this respect with other phrasal categories), and so complements of adjectives that immediately precede the head are not treated as instances of movement. For examples of movement out of ADJP, see traces in ADJP.

Adverb Phrase

ADVPs must be headed by adverbs. Noun phrases that are used adverbially are labeled NP-ADV.

  • Directional adverbs (ADVP-DIR):

    (IP-MAT-SPE (NEG οὐ)
                (NEG μὴ)
                (NP-SBJ *pro*)
                (VBS-AOR ἐξέλθῃς)
                (ADVP-DIR (ADV ἐκεῖθεν))
                (CP-ADV (C ἕως)
                        (IP-SUB (AN ἂν)
                                (NP-SBJ *pro*)
                                (VBS-AOR ἀποδῷς)
                                (NP-OB1 (DA τὸν) (ADJS ἔσχατον) (NA κοδράντην)))))                            
                            

    (Matthew 5.26)

  • Locative adverbs (ADVP-LOC):

    ( (IP-MAT-SPE (CODE {VS:8_12})
                  (IP-MAT-1 (ADVP-LOC (ADV ἐκεῖ))
                            (BEP-FUT ἔσται)
                            (NP-SBJ (D ὁ) (N κλαυθμὸς)))
                  (CONJP (CONJ καὶ)
                         (IP-MAT=1 (NP-SBJ (D ὁ)
                                           (N βρυγμὸς)
                                           (NP-COM (DS$ τῶν) (NS$ ὀδόντων)))))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

  • Temporal adverbs (ADVP-TMP):

    ( (IP-MAT (CODE {VS:3_15})
              (ADVP-TMP (ADV τότε))
              (NP-SBJ *pro*)
              (VBP-IMPF ἀφίησιν)
              (NP-OB1 (PROA αὐτόν))
              (. .))
      (ID GreekNT,Matthew))                            
                            

  • Other adverbs (ADVP):

    ( (IP-MAT-SPE (CODE {VS:13_15})
                  (CONJ καὶ)
                  (NP-INS (DSD τοῖς) (NSD ὠσὶν))
                  (ADVP (ADV βαρέως))
                  (NP-SBJ *pro*)
                  (VBD-AOR ἤκουσαν)
                  (, ,))
      (ID GreekNT,Matthew))                            
                            

    ( (IP-MAT (CODE {VS:26_43})
              (CONJ καὶ)
              (IP-PPL (VPR-AOR ἐλθὼν)
                      (ADVP (ADV πάλιν)))
              (NP-SBJ *pro*)
              (VBD-AOR εὗρεν)
              (IP-SMC (NP-SBJ (PROA αὐτοὺς))
                      (VPRA-IMPF καθεύδοντας))
              (, ,))
      (ID GreekNT,Matthew))                            
                            

Noun Phrase

Subject

All clauses contain a marked subject except imperatives (where subjects are indicated only when overt) and infinitives with PRO subjects. Apart from these two cases, when a clause has no overt subject, an empty subject is added. The infinitival complements of ECM verbs have subjects; see the section called “IP-INF-COM” for a discussion of ECM versus object control.

  • overt subject in a declarative:

    ( (IP-MAT (CODE {VS:1_2})
              (NP-SBJ (NPR Ἀβραὰμ))
              (VBD-AOR ἐγέννησεν)
              (NP-OB1 (DA τὸν) (NPRA Ἰσαάκ))
              (, ,))
      (ID GreekNT,Matthew))                                
                                

  • null subject in a declarative:

    ( (IP-MAT-SPE (CODE {VS:1_21})
                  (NP-SBJ *pro*)
                  (VBPP-FUT τέξεται)
                  (CLPRT δὲ)
                  (NP-OB1 (NA υἱὸν)))
      (ID GreekNT,Matthew))                                
                                

  • imperative without a subject:

    (IP-IMP-SPE (VBI-IMPF φεῦγε)
                (PP (P εἰς)
                    (NP (NPRA Αἴγυπτον))))                                
                                

    (Matthew 2.13)

  • imperative with an overt subject (especially common in Greek with 3rd-person imperatives, as here):

    (ADVP (ADV οὕτως))
    (VBI-AOR λαμψάτω)
    (NP-SBJ (D τὸ)
            (N φῶς)
            (NP-ATR (PRO$ ὑμῶν)))
    (PP (P ἔμπροσθεν)
        (NP (DS$ τῶν) (NS$ ἀνθρώπων)))                                
                                

    (Matthew 5.16)

  • infinitive with an (unmarked) PRO subject:

    (ADJP-PRD (ADJ καλόν))
    (NP-OB2 (PROD σοί))
    (BEP-IMPF ἐστιν)
    (IP-INF-SBJ (VBN-AOR εἰσελθεῖν)
                (PP (P εἰς)
                    (NP (DA τὴν) (NA ζωὴν)))
                (ADJP-SPR (ADJA κυλλὸν) (CONJ ἢ) (ADJA χωλόν)))                                
                                

    (Matthew 18.8)

  • ECM infinitive with a (marked) subject:

    (CONJ καὶ)
    (IP-PPL (VPR-IMPF ἐπιτιμῶν))
    (NEG οὐκ)
    (NP-SBJ *pro*)
    (VBD-IMPF εἴα)
    (IP-INF-COM (NP-SBJ (PROA αὐτὰ))
                (VBN-IMPF λαλεῖν))                                
                                

    (Luke 4.41)

Object

NP-OB1

Because of the difficulties inherent in trying to label objects consistently by case, we label complement NPs in four different ways: as (1) first object (-OB1) for accusative direct objects, as (2) second object (-OB2) for accusative or dative indirect objects, as (3) prepositional object (-OBP) for objects that derive their case from a prepositional prefix on the verb, and as (4) quirky object (-OBQ) for direct objects in a quirky case (i.e., genitive or dative). The rules for deciding which object to label as first, second, prepositional, or quirky are intended to be unambiguous and to work well for automatic retrieval, but they do not always yield the linguistically correct analysis. Given the variability and changes in case marking in the course of the history of Greek, the approach taken seems the only practical one.

NP-OB1 with ordinary monotransitive verbs.

The single object of a monotransitive verb is generally labeled NP-OB1.

( (IP-MAT-SPE (CODE {VS:1_21})
              (NP-SBJ *pro*)
              (VBPP-FUT τέξεται)
              (CLPRT δὲ)
              (NP-OB1 (NA υἱὸν)))
  (ID GreekNT,Matthew))                                    
                                

NP-OB1 with ditransitive verbs.

The direct object of a ditransitive verb is labeled NP-OB1.

( (IP-MAT (CODE {VS:4_8})
          (CONJ καὶ)
          (NP-SBJ *con*)
          (VBP-IMPF δείκνυσιν)
          (NP-OB2 (PROD αὐτῷ))
          (NP-OB1 (NP (QA πάσας)
                      (DSA τὰς)
                      (NSA βασιλείας)
                      (NP-ATR (D$ τοῦ) (N$ κόσμου)))
                  (CONJP (CONJ καὶ)
                         (NP (DA τὴν)
                             (NA δόξαν)
                             (NP-ATR (PRO$ αὐτῶν)))))
          (, ,))
  (ID GreekNT,Matthew))                                    
                                

NP-OB1 + clause with monotransitives.

With monotransitives (i.e., verbs that do not occur with two NP objects), the NP object in cases with an NP and a clause is labeled NP-OB1. See the section called “IP-INF-COM” for further relevant discussion.

( (IP-MAT-SPE (CODE {VS:5_42})
              (CONJ ἀλλὰ)
              (NP-SBJ *pro*)
              (VBP-PRF ἔγνωκα)
              (NP-OB1 (PROA ὑμᾶς))
              (CP-THT (C ὅτι)
                      (IP-SUB (NP-OB1 (DA τὴν)
                                      (NA ἀγάπην)
                                      (NP-ATR (D$ τοῦ) (N$ θεοῦ)))
                              (NEG οὐκ)
                              (NP-SBJ *pro*)
                              (VBP-IMPF ἔχετε)
                              (PP (P ἐν)
                                  (NP (PROD+SLF ἑαυτοῖς)))))
              (. .))
  (ID GreekNT,John))                                    
                                

NP-OB1 + clause with complex predicates.

This may not translate to Greek! Complex predicates include all cases of HAVE + NP with a verbal equivalent (HAVE FEAR = FEAR, HAVE KNOWLEDGE = KNOW, HAVE MARVEL = MARVEL, etc.). Some other common complex predicates are:

  • bear witness

  • bow (one’s) heart

  • do (no) force/pain

  • give example/(one’s) heart/heed/(one’s) intent/leave/notice/occasion/order

  • have care/(great) comfort/head/(no) joy/leave/mind/need/occasion/stomach/(good) will/s.o. to witness

  • kindle (one’s) heart

  • lay (one’s) force

  • make countenance/effort/fiaunce/(no) mention/note/(no) semblant/shift

  • send word

  • send word

  • stir (one’s) heart

  • take care/counsel/(one’s) heart/heed/keep/note/notice/occasion/order/pains

  • turn (one’s) heart

The noun member of a complex predicate is labeled NP-OB1. The accompanying clause is treated as a clause-level constituent, rather than as a complement of the noun.

( (IP-MAT (CONJ and)
          (NP-SBJ *con*)
          (HVD had)
          (NP-OB1 (N mervayle))
          (CP-QUE (WADVP-1 (WADV whye))
                  (C 0)
                  (IP-SUB (ADVP *T*-1)
                          (NP-SBJ (PRO he))
                          (BED was)
                          (ADVP (ADV so))
                          (VAN abaysshed)))
          (. .))
  (ID CMMALORY,655.4467))                        
                    

NP-OB2

NP-OB2 with ditransitive verbs.

The indirect object of a ditransitive verb is labeled NP-OB2.

( (IP-MAT (CODE {VS:4_8})
          (CONJ καὶ)
          (NP-SBJ *con*)
          (VBP-IMPF δείκνυσιν)
          (NP-OB2 (PROD αὐτῷ))
          (NP-OB1 (NP (QA πάσας)
                      (DSA τὰς)
                      (NSA βασιλείας)
                      (NP-ATR (D$ τοῦ) (N$ κόσμου)))
                  (CONJP (CONJ καὶ)
                         (NP (DA τὴν)
                             (NA δόξαν)
                             (NP-ATR (PRO$ αὐτῶν)))))
          (, ,))
  (ID GreekNT,Matthew))                            
                        

NP-OB2 in copular constructions.

When the second NP in a copular construction is neither co-referential with nor predicated of the subject, it is labeled NP-OB2.

( (IP-MAT-SPE (CODE {VS:16_22})
              (NEG οὐ)
              (NEG μὴ)
              (BEP-FUT ἔσται)
              (NP-OB2 (CLPROD σοι))
              (NP-SBJ (D τοῦτο))
              (. .))
  (ID GreekNT,Matthew))                            
                        

We treat verbs like delight, help, pain, pray, and thank, etc. as taking an abstract direct object that incorporates into the verb. The superficial object is therefore annotated as NP-OB2.

( (IP-IMP-SPE (CODE {VS:6_6})
              (CONJ καὶ)
              (IP-PPL (VPR-AOR κλείσας)
                      (NP-OB1 (DA τὴν)
                              (NA θύραν)
                              (NP-ATR (CLPRO$ σου))))
              (VBIP-AOR πρόσευξαι)
              (NP-OB2 (DD τῷ)
                      (ND πατρί)
                      (NP-ATR (CLPRO$ σου))
                      (NP-PRN (DD τῷ)
                              (PP (P ἐν)
                                  (NP (DD τῷ) (ADJD κρυπτῷ)))))
              (. .))
  (ID GreekNT,Matthew))                            
                        

NP-OB2 + clause with ditransitives.

When a ditransitive verb takes an animate NP and a clause, rather than two NPs, the label of the animate NP remains the same as in the case with two NPs, preserving the parallelism between the two cases. In Greek, this only seems to occur with an NP-OB2 + infinitival clause.

(IP-SUB (NP-SBJ *T*-1)
        (VBD-AOR ὑπέδειξεν)
        (NP-OB2 (PROD ὑμῖν))
        (IP-INF-COM (VBN-AOR φυγεῖν)
                    (PP (P ἀπὸ)
                        (NP (D$ τῆς)
                            (RRC (VPR$-IMPF μελλούσης))
                            (N$ ὀργῆς)))))                            
                        

(Matthew 3.7)

NP-OB2 as the label for a genitive of price

For lack of a better label, the genitive of price with verbs that select for such an argument is labeled NP-OB2. A comment indicating that the genitive is a genitive of price should also be included.

( (IP-MAT-SPE (CODE {VS:26_9})
              (VBDP-IMPF ἐδύνατο-δύναμαι)
              (CLPRT γὰρ-γάρ)
              (IP-INF-SBJ (IP-INF-1 (NP-SBJ (D τοῦτο-οὗτος))
                                    (VBNP-AOR-PASS πραθῆναι-πιπράσκω)
                                    (NP-OB2 (CODE {COM:genitive_of_price})
                                            (Q$ πολλοῦ-πολύς)))
                          (CONJP (CONJ καὶ-καί)
                                 (IP-INF=1 (VBNP-AOR-PASS δοθῆναι-δίδωμι)
                                           (NP-OB2 (ADJD πτωχοῖς-πτωχός)))))
              (. .))
  (ID GreekNT,Matthew))                            
                        

NP-OBP

If the object of a verb gets its case from a prepositional prefix on the verb (check in the LSJ to make sure the case usage for that preposition is correct), the object is labeled NP-OBP. In the case below αὐτῷ is in the dative because of the prefix προσ- on the verb. According to the LSJ, πρός + dative is normal following verbs of motion.

( (IP-MAT (CODE {VS:5_1})
          (CONJ καὶ)
          (IP-ABS (VPR$-AOR καθίσαντος)
                  (NP-SBJ (PRO$ αὐτοῦ)))
          (VBD-AOR προσῆλθαν)
          (NP-OBP (PROD αὐτῷ))
          (NP-SBJ (DS οἱ)
                  (NS μαθηταὶ)
                  (NP-ATR (PRO$ αὐτοῦ)))
          (. .))
  (ID GreekNT,Matthew))                        
                    

NP-OBQ

If no other label is appropriate for an NP argument of a verb and the argument is a direct object that appears in the genitive or dative case, the NP is labeled NP-OBQ. One should check in the LSJ or a specifically NT Greek lexicon to be sure that the verb in question regularly takes an NP argument. If not, then possibly the NP should be labeled NP-ADV or NP-INS if it has a means or instrumental sense.

( (IP-MAT-SPE (CODE {VS:14_64})
              (NP-SBJ *pro*)
              (VBD-AOR ἠκούσατε)
              (NP-OBQ (D$ τῆς) (N$ βλασφημίας))
              (. ;))
  (ID GreekNT,Mark))                        
                    

(NP-SBJ *pro*)
(VBS-IMPF μιμνήσκῃ)
(NP-OBQ (PRO$ αὐτοῦ))                        
                    

(Hebrews 2.6)

( (IP-MAT (CODE {VS:9_9})
          (CONJ καὶ)
          (IP-PPL (VPR-AOR ἀναστὰς))
          (NP-SBJ *pro*)
          (VBD-AOR ἠκολούθησεν)
          (NP-OBQ (PROD αὐτῷ))
          (. .))
  (ID GreekNT,Matthew))                        
                    

( (IP-MAT-SPE (CODE {VS:21_32})
              (CONJ καὶ)
              (NEG οὐκ)
              (NP-SBJ *pro*)
              (VBD-AOR ἐπιστεύσατε)
              (NP-OBQ (PROD αὐτῷ))
              (. .))
  (ID GreekNT,Matthew))                        
                    

When a verb that normally takes an object with quirky case—e.g., ἀκούω, which normally takes the genitive—takes an accusative object instead, this object should be labeled NP-OB1 as is usual for accusative direct objects.

( (IP-MAT-SPE (CODE {VS:26_65})
              (INTJ ἴδε)
              (ADVP-TMP (ADV νῦν))
              (NP-SBJ *pro*)
              (VBD-AOR ἠκούσατε)
              (NP-OB1 (DA τὴν) (NA βλασφημίαν))
              (. .))
  (ID GreekNT,Matthew))                        
                    

Objects in Passive

Whichever object is promoted to subject in the passive of ditransitive verbs, the remaining object is treated in a way that maximizes the parallelism between active and passive.

  • direct object promoted to subject:

    ( (IP-MAT (NP-SBJ (D the) (N letter)))
              (BED was)
              (VAN given)
              (NP-OB2 (PRO him))
      (. .))                                    
                                    

  • indirect object promoted to subject:

    ( (IP-MAT (NP-SBJ (PRO he))
              (BED was)
              (VAN given)
              (NP-OB1 (D the) (N letter)))
      (. .))                                    
                                    

Non-Argument (Adjunct) NP

This is something of a catch-all category, which includes the following types, all or none of which may be evidenced in Greek:

  • fronted NPs not associated with a resumptive (-RSP) element in the body of the clause:

    ( (IP-IMP-SPE (CODE {VS:10_14})
                  (CONJ καὶ)
                  (NP-ADT (CP-FRL (WNP-1 (WPRO ὃς))
                                  (C 0)
                                  (IP-SUB (IP-SUB (NP-SBJ *T*-1)
                                                  (PRTG ἂν)
                                                  (NEG μὴ)
                                                  (VBSP-AOR δέξηται)
                                                  (NP-OB1 (PROA ὑμᾶς)))
                                          (CONJP (NEG+CONJ μηδὲ)
                                                 (IP-SUB (NP-SBJ *con*)
                                                         (VBS-AOR ἀκούσῃ)
                                                         (NP-OB1 (DSA τοὺς)
                                                                 (NSA λόγους)
                                                                 (NP-ATR (PRO$ ὑμῶν))))))))
                  (, ,)
                  (IP-PPL (VPRP-IMPF ἐξερχόμενοι)
                          (PP (P ἔξω)
                              (NP (NP (D$ τῆς) (N$ οἰκίας))
                                  (CONJP (CONJ ἢ)
                                         (NP (D$ τῆς) (N$ πόλεως) (D$ ἐκείνης))))))
                  (VBI-AOR ἐκτινάξατε)
                  (NP-OB1 (DA τὸν)
                          (NA κονιορτὸν)
                          (NP-COM (DS$ τῶν)
                                  (NS$ ποδῶν)
                                  (NP-ATR (PRO$ ὑμῶν))))
                  (. .))
      (ID GreekNT,Matthew))                                
                                

  • fronted NPs that mean more or less AS FOR NP (although these actually receive NP-ADV in the Gortyn Law Code):

    ( (IP-INF-IMP (NP-ADT (D$ το) (CLPRT δὲ) (N$ κρόνο))
                  (NP-SBJ (DA τὸν) (NA δι[κ]αστὰν))
                  (IP-PPL (VPRA-IMPF ὀμνύντα))
                  (VBN-IMPF κρίνεν)
                  (. .))
      (ID Gortyn,LawCode.6))                                
                                

  • traces of wh- phrases without an evident semantic relation to the body of the clause:

    (NODE (CP-CAR (WNP-1 (D the) (WPRO which))
                  (C 0)
                  (IP-SUB (NP-ADT *T*-1)
                          (NP-SBJ (PRO I))
                          (VBP marvell)
                          (ADVP (ADV greatly))
                          (CP-THT (C that)
                                  (IP-SUB (NP-SBJ (PRO I))
                                          (HVP have)
                                          (NP-OB1 (Q no) (N word))
                                          (PP (P from)
                                              (NP (PRO you)))))))
          (ID APLUMPT,186.92))                                
                                

Adverbial NP

NPs acting adverbially are labeled NP-ADV. The use of NP-ADV is essentially a catch-all that includes all adverbial uses (e.g., adverbial accusative) of nouns that do not fall into any other category such as NP-TMP, NP-LOC, etc. In most cases if trying to decide between NP-ADT and NP-ADV, the choice will be NP-ADV. In the traditional terms of Greek grammar[3], some of the uses which NP-ADV covers (but this list is probably not exhaustive) are:

  • dative of respect:

    (NP-SBJ (DS οἱ)
            (ADJP (ADJ πτωχοὶ)
                  (NP-ADV (DD τῷ) (ND πνεύματι))))                                
                                

    (Matthew 5.3)

  • adverbial accusative:

    (ADVP-TMP (ADV ποσάκις))
    (NP-SBJ *pro*)
    (VBD-AOR ἠθέλησα)
    (IP-INF-COM (VBN-AOR ἐπισυναγαγεῖν)
                (NP-OB1 (DSA τὰ)
                        (NSA τέκνα)
                        (NP-ATR (CLPRO$ σου)))
                (, ,)
                (NP-ADV (CP-FRL (WNP-1 (WDA ὃν) (NA τρόπον))
                                (C 0)
                                (IP-SUB (NP-ADV *T*-1)
                                        (NP-SBJ (N ὄρνις))
                                        (VBP-IMPF ἐπισυνάγει)
                                        (NP-OB1 (DSA τὰ)
                                                (NSA νοσσία)
                                                (NP-ATR (PRO$ αὐτῆς)))
                                        (PP (P ὑπὸ)
                                            (NP (DSA τὰς) (NSA πτέρυγας)))))))                                
                                

    (Matthew 23.37)

  • idiomatic usage:

    ( (IP-MAT (CODE {VS:6_40})
              (CONJ καὶ)
              (NP-SBJ *pro*)
              (VBD-AOR ἀνέπεσαν)
              (NP-ADV (NS πρασιαὶ) (NS πρασιαὶ))
              (PP (PP (P κατὰ)
                      (NP (NUM ἑκατὸν)))
                  (CONJP (CONJ καὶ)
                         (PP (P κατὰ)
                             (NP (NUM πεντήκοντα)))))
              (. .))
      (ID GreekNT,Mark))                                
                                

  • dative of manner:

    (CP-ADV (C+PRTG ἐὰν)
            (IP-SUB (NEG μὴ)
                    (NP-ADV (ND πυγμῇ))
                    (NP-SBJ *pro*)
                    (VBSP-AOR νίψωνται)
                    (NP-OB1 (DSA τὰς) (NSA χεῖρας))))                                
                                

    (Mark 7.3)

  • dative of interest (dative of advantage/disadvantage)

  • dative of reference

  • ethical dative

  • dative of cause

  • dative of (military) accompaniment

  • accusative of respect

Attributive (and/or Possessive) NP

A bare NP, usually in the genitive, that modifies another noun is often an attributive NP, labeled NP-ATR. Possessive genitives are included in this category. Subjective genitives are also included in this category, since they can be very difficult to distinguish from possessive genitives.

  • non-possessive NP-ATR:

    (IP-PPL (VPA-AOR ἄρξαντες)
            (CLPRT μὲν)
            (PP (P ἐπὶ)
                (NP (NP (NUMP (NUM δύο)
                              (CLTE τε)
                              (CONJ καὶ)
                              (NUM εἴκοσι))
                        (NSA γενεὰς)
                        (NP-ATR (NS$ ἀνδρῶν)))
                    (CONJP (NP (NSA ἔτεα)
                               (NUMP (NUM πέντε)
                                     (CLTE τε)
                                     (CONJ καὶ)
                                     (ADJA πεντακόσια)))))))                                
                                

    (Hdt. 1.7.4)

  • possessive NP-ATR:

    ( (IP-MAT (CODE {VS:1_8_1})
              (NP-SBJ (D οὗτος) 
                      (CLPRT δὴ)
                      (CLPRT ὦν)
                      (D ὁ)
                      (NPR Κανδαύλης))
              (VBDP-AOR ἠράσθη)
              (NP-OBQ (D$ τῆς)
                      (NP-ATR (PRO$+SLF ἑωυτοῦ))
                      (N$ γυναικός))
              (, ,))
      (ID Herodotus,Histories))                                
                                

  • dative of possession:

    (NP-SBJ (DA τὸ)
            (CLPRT δέ)
            (NP-ATR (CLPROD οἱ))
            (NA οὔνομα))                               
                                

    (Hdt. 1.1.3)

  • (potentially) subjective genitive:

    (PP (P πρὸ)
        (CLPRT δὲ)
        (NP (D$ τῆς)
            (NP-ATR (NPR$ Κροίσου))
            (N$ ἀρχῆς)))
                                

    (Hdt. 1.6.3)

Complement NP

A bare NP in the complement to a noun or adjective is labeled NP-COM. In the case of a nominal complement of a noun, the complement noun is most often an objective genitive, in the traditional terms of Greek grammar.

  • NP complement of a noun (objective genitive):

    (NP-OB1 (NP-COM (NPR$ Μηδείης))
            (DA τὴν)
            (NA ἁρπαγήν))                               
                                

    (Hdt. 1.3.2)

  • NP complement of an adjective:

    ( (IP-INF-THT (CODE {VS:1_2_1})
                  (PP (P μετὰ)
                      (CLPRT δὲ)
                      (NP (DSA ταῦτα)))
                  (NP-SBJ (NPRSA Ἕλληνας))
                  (ADJP-PRD (ADJA αἰτίους)
                            (NP-COM (D$ τῆς)
                                    (ADJ$ δευτέρης)
                                    (N$ ἀδικίης)))
                  (VBNP-AOR γενέσθαι)
    	      (. .))
      (ID Herodotus,Histories))                                
                                

Directional NP

A bare NP with a verb of motion indicating a direction (to or from) is labeled NP-DIR. This usage is found mainly in Homer.

Instrumental NP

A bare NP indicating the means or instrument with which something was accomplished is labeled NP-INS. This corresponds with the dative of means/instrumental dative in the terms of traditional Greek grammar[4].

(NP-SBJ (NPR Ἰωάνης))
(CLPRT μὲν)
(VBD-AOR ἐβάπτισεν)
(NP-INS (ND ὕδατι))                    
                

(Acts 1.5)

Locative NP

A bare NP indicating location is labeled NP-LOC. This corresponds with the dative of place where in the terms of traditional Greek grammar[5].

(NP-SBJ (NEG+Q Οὐδεὶς))
(CLPRT δὲ)
(IP-PPL (NP-OB1 (NA λύχνον))
        (VPR-AOR ἅψας))
(VBP-IMPF καλύπτει)
(NP-OB1 (PROA αὐτὸν))
(NP-LOC (ND σκεύει)))                    
                

(Luke 8.16)

Measure Phrase

[Important] Important

In cases that are ambiguous between complements (say, NP-OB1) and adjuncts (say, NP-MSR), the default is complement. A simple test for doubtful cases is to passivize the sentence; if the relevant noun phrase can be the subject of a passive, it should be annotated as NP-OB1.

( (IP-MAT (NP-SBJ (PRO He))
          (VBD ate)
          (NP-ACC (ADVR too) (Q much))
          (. .)))

( (IP-MAT (NP-SBJ (PRO He))
          (VBD ate)
          (NP-ACC (QP (ADVR too) (Q much))
                  (N spinach))
          (. .)))

( (IP-MAT (NP-SBJ (PRO He))
          (VBD ate)
          (NP-MSR (ADJP (ADVR too) (ADJ long)))
          (. .)))                        
                    

Measure phrases occur as sentential constituents. We also explicitly label measure phrase modifiers of other constituents (see, for instance, measure phrase modifiers of N). Only NP measure phrases are indicated as such. PP measure phrases (FOR THREE DAYS, etc.) are not indicated, following our policy of not specifying the function of PPs.

[Important] Important

The distinction between NP-MSR and QP, described immediately below, is not straightforward, and it is likely that some instances of one category are mistagged as the other. It is therefore wise to search for measure phrases in terms of the disjunction NP-MSR|QP.

At the sentence level, noun phrases expressing measure are labeled NP-MSR. As is generally the case with noun phrases, categories other than nouns may act as the head of NP-MSR. In the traditional terms of Greek grammar, NP measure phrases usually correspond to one of the following:

  • accusative of extent of space:

    (NP-SBJ (D Τὸ) (CLPRT δὲ) (N πλοῖον))
    (ADVP-TMP (ADV ἤδη))
    (NP-MSR (NSA σταδίους) (QA πολλοὺς))
    (PP (P ἀπὸ)
        (NP (D$ τῆς) (N$ γῆς)))
    (VBD-IMPF ἀπεῖχεν)                                
                                

    (Matthew 14.24)

  • accusative of extent of time:

    (CONJ καὶ)
    (NP-SBJ *con*)
    (VBD-IMPF περιέκρυβεν)
    (NP-OB1 (PROA+SLF ἑαυτὴν))
    (NP-MSR (NSA μῆνας) (NUM πέντε))                                
                                

    (Luke 1.25)

  • dative of time at which (?[6]):

    ( (IP-MAT (CODE {VS:8_27})
              (CONJ καὶ)
              (NP-MSR (ND χρόνῳ) (ADJR ἱκανῷ))
              (NEG οὐκ)
              (NP-SBJ *pro*)
              (VBDP-AOR ἐνεδύσατο)
              (NP-OB1 (NA ἱμάτιον))
              (, ,))
      (ID GreekNT,Luke))                                
                                

Measure phrases at levels lower than the sentence (notably, within ADJP and ADVP) are treated slightly differently. Here, measure expressions headed by Q are not surrounded by NP-MSR. But other measure phrases continue to be surrounded by NP-MSR brackets, even when they consist of a single word.

(ADJP (Q much) (ADJR higher))

(ADJP (QP (ADV very) (Q much))
      (ADJR higher))

(ADJP (QP (Q much) (QR more))
      (VAG imposing))

(ADJP (QP (QP (ADVR so) (Q much))
          (QR more))
      (VAG imposing))

(ADJP (NP-MSR (ADJ far))
      (ADJR higher))

(ADJP (NP-MSR (ADJP (ADV quite) (ADJ far))
      (ADJ distant))

(ADJP (NP-MSR (Q+WPRO somewhat))
      (ADJR higher))

(ADJP (NP-MSR (Q some) (WPRO what))
      (ADJR higher))

(ADJP (NP-MSR (Q many) (NS times))
      (ADJR higher))

(ADJP (QP (NP-MSR (Q many) (NS times))
          (QR more))
      (ADJR higher))

(ADVP (NP-MSR (NUM half))
      (ADVR so) (ADV sore))                    
                

Partitive NP

Bare NPs in the genitive that express partitivity—i.e., partitive genitives—are labeled NP-PAR.

( (IP-MAT (CODE {VS:1_1_1})
          (NP-SBJ (NP-PAR (NPRS$ Περσέων))
                  (CLPRT μέν)
                  (CLPRT νυν)
                  (DS οἱ)
                  (ADJ λόγιοι))
          (NP-1 (NPRSA Φοίνικας))
          (ADJA-2 αἰτίους)
          (VBP-IMPF φασὶ)
          (IP-INF-THT (NP-SBJ *ICH*-1)
                      (VBNP-AOR γενέσθαι)
                      (ADJP-PRD (ADJA *ICH*-2)
                                (NP-COM (D$ τῆς)
                                        (N$ διαφορῆς))))
          (. .))
  (ID Herodotus,Histories))                 
                

Predicate NP

NP predicates of copular verbs and NP predicates in small clauses are labeled NP-PRD.

  • with a copular verb:

    (CP-ADV (C ἵνα)
            (IP-SUB (NP-SBJ (DS οἱ) (NS λίθοι) (DS οὗτοι))
                    (NP-PRD (NS ἄρτοι))
                    (VBSP-AOR γένωνται)))                                
                                

    (Matthew 4.3)

  • in a small clause:

    (CP-REL-3 (WNP-4 (WPRO ὃς))
              (C 0)
              (IP-SUB (NP-SBJ *T*-4)
                      (NEG οὐ)
                      (VBP-IMPF γινώσκει)
                      (IP-PPL-THT (NP-SBJ (DA τὴν)
                                          (NP-ATR (ADJ$ Ἐφεσίων))
                                          (NA πόλιν))
                                  (NA-5 νεωκόρον)
                                  (BPR-IMPF οὖσαν)
                                  (NP-PRD (NA *ICH*-5)
                                          (NP-COM (NP (D$ τῆς) (ADJ$ μεγάλης) (NPR$ Ἀρτέμιδος))
                                                  (CONJP (CONJ καὶ)
                                                         (NP (D$ τοῦ) (ADJ$ διοπετοῦς)))))))                                
                                

    (Acts 19.35)

Note that a predicate NP in a copular construction may not always be in the nominative; genitive predicates (usually indicating possession, although not concretely in the example below) are also possible:

(IP-MAT-SPE (NEG Οὐχ)
            (NP-PRD (PRO$ ὑμῶν))
            (BEP-IMPF ἐστὶν)
            (IP-INF-SBJ (VBN-AOR γνῶναι)
                        (NP-OB1 (NSA χρόνους) (CONJ ἢ) (NSA καιροὺς)
                                (CP-REL (WNP-1 (WPROA οὓς))
                                        (C 0)
                                        (IP-SUB (NP-OB1 *T*-1)
                                                (NP-SBJ (D ὁ) (N πατὴρ))
                                                (VBDP-AOR ἔθετο)
                                                (PP (P ἐν)
                                                    (NP (DD τῇ) (ADJD ἰδίᾳ) (ND ἐξουσίᾳ))))))))                    
                

(Acts 1.7)

Secondary Predicate NP

See also small clauses.

The dash tag -SPR is appended to all NPs that act as secondary predicates.

(IP-INF (TO to)
        (VB stand/mount)
        (NP-SPR (N guard)))

(IP-INF (TO to)
        (VB elect)
        (NP-ACC (D a) (NPR Hapsburg))
        (NP-SPR (N emperor)))                    
                

The Greek semantic equivalents of the following verbs are treated as taking secondary predicates; the list is intended to be illustrative rather than exhaustive and should be compared with the list of verbs that we treat as taking small clause complements.

ADMIT, AGREE, ANOINT, APPOINT, BIND, BRING, CHOOSE, CHRISTEN, CLAIM, CONFIRM, CONSECRATE, CROWN, DEFEND, DEFINE, DEVOTE, ELECT, INTERPRET, KEEP, LEAVE, LOVE, NAME, NOMINATE, OFFER, ORDAIN, PRESERVE, SET, SUBSCRIBE, SUFFER, SURNAME, TAKE, TESTIFY, TRANSLATE, TRY, VOW, WITNESS

The general rule of thumb is that if the meaning of the verb does not change depending on whether a predicate-like item is in the sentence, the verb takes a secondary predicate. For example, the following do not differ fundamentally in their meaning (the second just contains more information):

  1. They named him.

  2. They named him John.

In contrast, verbs that we analyze as taking small clauses may have very different meanings (although not always) depending on whether a predicate-like item is present:

  1. We called him.

  2. We called him a loser.

In addition, the following Greek verbs should be considered to take secondary predicates.

  • τίκτω (more often passive—e.g., τεχθείς in Matthew 2:2) ‘give birth to’

Some examples from the PPCNTG:

  • ( (IP-MAT (CODE {VS:24_27})
              (IP-ABS (NP-SBJ (N$ Διετίας))
                      (CLPRT δὲ)
                      (VPRP$-AOR-PASS πληρωθείσης))
              (VBD-AOR ἔλαβεν)
              (NP-SPR (NA διάδοχον))
              (NP-SBJ (D ὁ) (NPR Φῆλιξ))
              (NP-OB1 (NPRA Πόρκιον)
                      (NP-PRN (NPRA Φῆστον)))
              (. .))
      (ID GreekNT,Acts))                            
                            
  • ( (IP-IMP (CODE {VS:3_15})
              (NP-OB1 (NA κύριον))
              (CLPRT δὲ)
              (NP-SPR (DA τὸν) (NPRA Χριστὸν))
              (VBI-AOR ἁγιάσατε)
              (PP (P ἐν)
                  (NP (DSD ταῖς)
                      (NSD καρδίαις)
                      (NP-ATR (PRO$ ὑμῶν))))
              (, ,))
      (ID GreekNT,1Peter))                            
                            

Temporal NP

NPs denoting points in time (THAT DAY, TUESDAY, YESTERDAY, etc.) are labeled NP- TMP. Dates are annotated with flat structure.

( (IP-MAT (CODE {VS:14_25})
          (NP-TMP (ADJD Τετάρτῃ)
                  (CLPRT δὲ)
                  (ND φυλακῇ)
                  (NP-PAR (D$ τῆς) (N$ νυκτὸς)))
          (NP-SBJ *pro*)
          (VBD-AOR ἦλθεν)
          (PP (P πρὸς)
              (NP (PROA αὐτοὺς)))
          (IP-PPL (VPR-IMPF περιπατῶν)
                  (PP (P ἐπὶ)
                      (NP (DA τὴν) (NA θάλασσαν))))
          (. .))
  (ID GreekNT,Matthew))                    
                

NPs denoting frequency (EVERY DAY, TUESDAYS, etc.) are labeled NP-TMP.

( (CP-QUE (CODE {VS:15_30})
          (WADVP-1 (WADV τί))
          (C 0)
          (IP-SUB (ADVP *T*-1)
                  (NP-SBJ (ADV καὶ) (PRO ἡμεῖς))
                  (VBP-IMPF κινδυνεύομεν)
                  (NP-TMP (QA πᾶσαν) (NA ὥραν)))
          (. ;))
  (ID GreekNT,1Corinthians))                    
                

In the terms of traditional Greek grammar, the NPs receiving the label NP-TMP correspond to genitive of time within which, the dative of time at which, and the accusative of extent of time[7].

Vocative

Vocatives are labeled NP-VOC. Accompanying interjections—i.e., ὦ—are treated as clause-level constituents rather than as constituents of the vocative NP.

( (IP-IMP (CODE {VS:6_11})
          (NP-SBJ (PRO Σὺ))
          (CLPRT δέ)
          (, ,)
          (INTJ ὦ)
          (NP-VOC (N ἄνθρωπε)
                  (NP-ATR (N$ θεοῦ)))
          (, ,)
          (NP-OB1 (DSA ταῦτα))
          (VBI-IMPF φεῦγε)
          (. .))
  (ID GreekNT,1Timothy))                    
                

Bare NP

The bare label NP is used for NPs without any of the above functions. These include:

  • NP complements of P

  • conjoined NPs

  • NPs extracted from another constituent

The topmost NP of a conjoined NP is labeled by function if necessary. The NPs immediately dominating the individual conjuncts are never marked for function.

( (IP-MAT (CODE {VS:1_3})
          (NP-SBJ (NPR Ἰούδας))
          (CLPRT δὲ)
          (VBD-AOR ἐγέννησεν)
          (NP-OB1 (NP (DA τὸν) (NPRA Φαρὲς))
                  (CONJP (CONJ καὶ)
                         (NP (DA τὸν) (NPRA Ζαρὰ))))
          (PP (P ἐκ)
              (NP (D$ τῆς) (NPR$ Θάμαρ)))
          (, ,))
  (ID GreekNT,Matthew))                    
                

NPs extracted from another constituent (phrase or clause) are labeled NP and coindexed with a trace within the constituent from which they are extracted. Examples include topicalization with preposition stranding and long-distance topicalization. If necessary, the function of the extracted NP is indicated on the trace.

(IP-PPL (NP-2 (N$ σαββάτου))
        (VPR-IMPF ἔχον)
        (NP-OB1 (NP-MSR *ICH*-2)
                (NA ὁδόν)))                    
                

(Acts 1.12)

Prepositional Phrase

For want of reliable criteria, our annotation system makes no distinction between complement and adjunct PPs.

(INTJ Ἄφες)
(NP-SBJ *pro*)
(VBS-AOR ἐκβάλω)
(NP-OB1 (DA τὸ) (NA κάρφος))
(PP (P ἐκ)
    (NP (D$ τοῦ)
        (N$ ὀφθαλμοῦ)
        (NP-ATR (CLPRO$ σου))))                
            

(Matthew 7.4)

(NP-SBJ (N ἄγγελος)
        (NP-ATR (N$ Κυρίου)))
(PP (P κατ')
    (NP (NA ὄναρ)))
(VBDP-AOR ἐφάνη)                
            

(Matthew 1.20)

By default, complements of prepositions that immediately precede the head are treated as constituents of the PP, and movement out of PP is indicated only if the word order leaves no alternative.

  • complements that precede the head:

    (IP-PPL (VPRP-AOR-PASS ἀποσχισθέντες)
            (NP-OBP (D$ τῆς)
                    (ADJ$ ἄλλης)
                    (N$ στρατιῆς))
            (PP (NP (DS$ τῶνδε))
                (P εἵνεκα)))                            
                            

    (Hdt. 8.35.2)

  • movement out of PP:

Although the trace of passivized constituents is not normally indicated, it is in the case of passivization from PP. See Traces of A-Movement for examples. It may be that passivization from PP does not occur in Greek.

Quantifier Phrase

In general, QP does not occur at the clausal level. There are two exceptions.

  • When a quantifier is part of a phrase that is comparable in meaning to a floated quantifier.

    ( (IP-MAT (CODE {VS:1_14_1})
              (NP-SBJ (NPR Γύγης))
              (CLPRT δὲ)
              (IP-PPL (VPR-AOR τυραννεύσας))
              (VBD-AOR ἀπέπεμψε)
              (NP-OB1 (NSA-1 ἀναθήματα))
              (PP (P ἐς)
                  (NP (NPRSA Δελφοὺς)))
              (QP=1 (NEG οὐκ) (QA ὀλίγα))
              (, ,))
      (ID Herodotus,Histories))                          
                            

  • When a QP is extracted out of a phrase.

    ( (CP-QUE (WNP-1 (WPRO what)
                     (QP *ICH*-2))
              (C 0)
              (IP-SUB (NP-OB1 *T*-1)
                      (MD would)
                      (NP-SBJ (PRO you))
                      (HV have)
                      (QP-2 (QR more)))
              (. ,))
      (ID PENNY,156.366))                        
                        

Words as Clausal Constituents

Adverbs

It is apparently possible for adverbs to scramble out of other phrases and become daughters of IP. Some instances look like permutations of the adverb and a preposition in the original, but we have not attempted to emend the text.

(IP-MAT-SPE (NEG Οὐ)
            (ADV-1 μακρὰν)
            (NP-SBJ *pro*)
            (BEP-IMPF εἶ)
            (PP (ADV *ICH*-1)
                (P ἀπὸ)
                (NP (D$ τῆς)
                    (N$ βασιλείας)
                    (NP-ATR (D$ τοῦ) (N$ θεοῦ)))))                    
                

(Mark 12.34)

Floated Quantifiers

See also clause-level QP. Floated quantifiers are labeled Q* without a phrase label, unless a phrase label is necessary because the quantifier itself has a modifier. The same notation as used for gapping is used to indicate the coindexation between a floated quantifier and the NP it modifies.

  • floated quantifier:

    (IP-MAT-SPE (NEG+QA=1 Οὐδὲν)
                (NP-SBJ *pro*)
                (VBP-IMPF εὑρίσκω)
                (NP-OB1 (ADJA-1 αἴτιον))
                (PP (P ἐν)
                    (NP (DD τῷ) (ND ἀνθρώπῳ) (DD τούτῳ))))                                
                                

    (Luke 23.4)

  • floated quantifier with a QP layer:

    ( (IP-MAT (CODE {VS:1_14_1})
              (NP-SBJ (NPR Γύγης))
              (CLPRT δὲ)
              (IP-PPL (VPR-AOR τυραννεύσας))
              (VBD-AOR ἀπέπεμψε)
              (NP-OB1 (NSA-1 ἀναθήματα))
              (PP (P ἐς)
                  (NP (NSA Δελφοὺς)))
              (QP=1 (NEG οὐκ) (QA ὀλίγα))
              (, ,))
      (ID Herodotus,Histories))                                
                                

If possible, the coindexation between a floated quantifier and the nominal that it modifies should be done at the word-level. When this is not possible because the modified nominal or the quantifier phrase contains more than one word, then the coindexation is done at the phrase level where necessary.

  • coindexation of a quantifier with a noun phrase:

    (IP-SUB-SPE (NP-SBJ (D ἡ-ὁ)
                        (N χήρα-χήρα)
                        (D αὕτη-οὗτος)
                        (NP-PRN (D ἡ-ὁ) (ADJ πτωχὴ-πτωχός)))
                        (ADJR-1 πλεῖον-πολύς)
                        (Q$=2 πάντων-πᾶς)
                        (VBP-PRF ἔβαλεν-βάλλω)
                        (NP-OB1 (ADJP (ADJR *ICH*-1)
                                      (NP-CMP-2 (DS$ τῶν-ὁ)
                                                (RRC (VPR$-IMPF βαλλόντων-βάλλω)
                                                     (PP (P εἰς-εἰς)
                                                     (NP (DA τὸ-ὁ) (NA γαζοφυλάκιον-γαζοφυλάκιον))))))))                               
                                

    (Mark 12.43)

[Important] Important

Where a phrase is ambiguous between an ordinary subject and a floated quantifier, it is treated as an ordinary subject.

( (IP-MAT-SPE (CODE {VS:11_27})
              (NP-SBJ (Q Πάντα))
              (NP-OB2 (CLPROD μοι))
              (VBDP-AOR-PASS παρεδόθη)
              (PP (P ὑπὸ)
                  (NP (D$ τοῦ)
                      (N$ πατρός)
                      (NP-ATR (CLPRO$ μου))))
	      (, ,))
  (ID GreekNT,Matthew))                        
                    

Negation

Negation never projects a phrase.

( (IP-MAT (CODE {VS:1_25})
          (CONJ καὶ)
          (NEG οὐκ)
          (NP-SBJ *con*)
          (VBD-IMPF ἐγίνωσκεν)
          (NP-OB1 (PROA αὐτὴν))
          (PP (P ἕως)
              (NP (CP-FRL (WNP-1 (WPRO$ οὗ))
                          (C 0)
                          (IP-SUB (NP-TMP *T*-1)
                                  (NP-SBJ *pro*)
                                  (VBD-AOR ἔτεκεν)
                                  (NP-OB1 (NA υἱόν))))))
          (. .))
  (ID GreekNT,Matthew))                    
                

Participles

At the clausal level, participles occur as complements of copular verbs, in reduced relative clauses, in small clauses, and in absolute clauses.

  • complement of copular verb:

    ( (IP-MAT (CODE {VS:1_9})
              (BED-IMPF Ἦν)
              (NP-SBJ (D τὸ)
                      (N φῶς)
                      (NP-PRN (D τὸ) (ADJ ἀληθινὸν))
                      (CP-REL (WNP-1 (WPRO ὃ))
                              (C 0)
                              (IP-SUB (NP-SBJ *T*-1)
                                      (VBP-IMPF φωτίζει)
                                      (NP-OB1 (QA πάντα) (NA ἄνθρωπον)))))
              (VPRP-IMPF ἐρχόμενον)
              (PP (P εἰς)
                  (NP (DA τὸν) (NA κόσμον)))
              (. .))
      (ID GreekNT,John))                                
                                

  • reduced relative clause:

    ( (IP-MAT (CODE {VS:2_9})
              (NP-SBJ (DS οἱ)
                      (CLPRT δὲ)
                      (RRC (VPR-AOR ἀκούσαντες)
                           (NP-OBQ (D$ τοῦ) (N$ βασιλέως))))
              (VBDP-AOR ἐπορεύθησαν)
              (, ,))
      (ID GreekNT,Matthew))                                
                                

  • small clause:

    ( (IP-MAT (CODE {VS:26_40})
              (CONJ καὶ)
              (NP-SBJ *con*)
              (VBP-IMPF εὑρίσκει)
              (IP-SMC (NP-SBJ (PROA αὐτοὺς))
                      (VPRA-IMPF καθεύδοντας))
              (, ,))
      (ID GreekNT,Matthew))                                
                                

  • absolute clause:

    ( (IP-MAT (CODE {VS:5_1})
              (CONJ καὶ)
              (IP-ABS (VPR$-AOR καθίσαντος)
                      (NP-SBJ (PRO$ αὐτοῦ)))
              (VBD-AOR προσῆλθαν)
              (NP-OBP (PROD αὐτῷ))
              (NP-SBJ (DS οἱ)
                      (NS μαθηταὶ)
                      (NP-ATR (PRO$ αὐτοῦ)))
              (. .))
      (ID GreekNT,Matthew))                                    
                                    

Extended Dash Tags

Left Dislocation

Left-dislocated constituents are associated with a resumptive element in the body of the clause, which is marked by the dash tag -RSP. The left-dislocated and the resumptive constituents are not coindexed since with very rare exceptions, there is only one -LFD/-RSP pair per clause. Function tags, if any, appear only on the resumptive element. In the absence of a resumptive associate, fronted constituents are labeled NP-ADT. In general, only phrases, and not words, are marked as resumptive, and so the -RSP label may be attached to the phrase that contains the resumptive element, rather than to the resumptive element itself. See also the section called “Prolepsis”.

  • left-dislocated NP:

    ( (IP-MAT-SPE (CODE {VS:7_40})
                  (NP-LFD (D ὁ)
                          (CLPRT γὰρ)
                          (NPR Μωυσῆς)
                          (D οὗτος)
                          (, ,)
                          (CP-REL (WNP-1 (WPRO ὃς))
                                  (C 0)
                                  (IP-SUB (NP-SBJ *T*-1)
                                          (VBD-AOR ἐξήγαγεν)
                                          (NP-OB1 (PROA ἡμᾶς))
                                          (PP (P ἐκ)
                                              (NP (N$ γῆς)
                                                  (NP-PRN (NPR$ Αἰγύπτου)))))))
                  (, ,)
                  (NEG οὐκ)
                  (NP-SBJ *pro*)
                  (VBP-PRF οἴδαμεν)
                  (CP-QUE (WNP-2 (WPRO τί))
                          (C 0)
                          (IP-SUB (NP-SBJ *T*-2)
                                  (VBDP-AOR ἐγένετο)
                                  (NP-OB2-RSP (PROD αὐτῷ))))
                  (. .))
      (ID GreekNT,Acts))                                
                                

  • non-NP left dislocation:

    ( (IP-MAT-SPE (CODE {VS:6_21})
                  (ADVP-LFD (CP-FRL (WADVP-1 (WADV ὅπου))
                                    (CLPRT γάρ)
                                    (C 0)
                                    (IP-SUB (ADVP-LOC *T*-1)
                                            (BEP-IMPF ἐστιν)
                                            (NP-SBJ (D ὁ)
                                                    (N θησαυρός)
                                                    (NP-ATR (CLPRO$ σου))))))
                  (, ,)
                  (ADVP-LOC-RSP (ADV ἐκεῖ))
                  (BEP-FUT ἔσται)
                  (NP-SBJ (ADV καὶ)
                          (D ἡ)
                          (N καρδία)
                          (NP-ATR (CLPRO$ σου)))
                  (. .))
      (ID GreekNT,Matthew))                                
                                

  • resumptive phrase:

    ( (IP-IMP-SPE (CODE {VS:21_23})
                  (NP-LFD (DA τοῦτο))
                  (CLPRT οὖν)
                  (VBI-AOR ποίησον)
                  (NP-OB1-RSP (CP-FRL (WNP-1 (WPROA ὅ))
                                      (C 0)
                                      (IP-SUB (NP-OB1 *T*-1)
                                              (NP-OB2 (CLPROD σοι))
                                              (NP-SBJ *pro*)
                                              (VBP-IMPF λέγομεν))))
    	      (. .))
      (ID GreekNT,Acts))                                
                                

Finally, there are cases where there is a category mismatch between the two elements that make up a correlative; these cases are analyzed as containing left dislocation and resumption in order to handle the category mismatch.

( (IP-IMP-SPE (CODE {VS:7_12})
              (NP-OB1-LFD (QA Πάντα)
                          (CLPRT οὖν)
                          (CP-REL (WNP-1 (WADJA ὅσα))
                                  (C+PRTG ἐὰν)
                                  (IP-SUB (NP-SBJ *pro*)
                                          (VBS-IMPF θέλητε)
                                          (CP-COM (C ἵνα)
                                                  (IP-SUB (NP-OB1 *T*-1)
                                                          (VBS-IMPF ποιῶσιν)
                                                          (NP-OB2 (PROD ὑμῖν))
                                                          (NP-SBJ (DS οἱ) (NS ἄνθρωποι)))))))
              (, ,)
              (ADVP-RSP (ADV οὕτως))
              (NP-SBJ (ADV καὶ) (PRO ὑμεῖς))
              (VBI-IMPF ποιεῖτε)
              (NP-OB2 (PROD αὐτοῖς))
              (. .))
  (ID GreekNT,Matthew))                    
                

Appositive or Parenthetical

See also ‘appositives and parentheticals’ in the section called “Modifiers of N”.

The extended label -PRN can be attached to any constituent label and indicates either an appositive or a parenthetical. If the appositive/parenthetical constituent immediately follows its co-constituent, it is included in the co-constituent’s brackets. If separated from its co-constituent, it is coindexed with a trace within the co-constituent. The two most common types are separated NP appositives (often probably right-dislocation structures) and clauses (usually that clauses) appositive to demonstratives (usually this or that). The latter is rare or does not occur in Greek.

  • included in the co-constituent’s brackets:

    (NP (NPRA Ἀνδρέαν)
        (NP-PRN (DA τὸν)
                (NA ἀδελφὸν)
                (NP-ATR (NPR$ Σίμωνος))))                                
                                

    (Mark 1.16)

  • co-indexed:

    (IP-MAT-SPE (NP-SBJ (N φωνὴ)
                        (NP-PRN *ICH*-1))
                (PP (P ἐν)
    	        (NP (NPRD Ῥαμὰ)))
                (VBDP-AOR-PASS ἠκούσθη)
                (, ,)
                (NP-1 (NP (N κλαυθμὸς))
                      (CONJP (CONJ καὶ)
                             (NP (N ὀδυρμὸς) (ADJ πολύς)))))                                
                                

    (Matthew 2.17)

  • Exception: no coindexation because speakers of imperatives are not normally indicated:

    (IP-IMP-SPE (NP-PRN (N φωνὴ)
                        (NP-ATR (RRC (VPR-IMPF βοῶντος)
                                     (PP (P ἐν)
                                         (NP (DD τῇ) (ADJD ἐρήμῳ))))))
                (VBI-AOR Ἑτοιμάσατε)
                (NP-OB1 (DA τὴν)
                        (NA ὁδὸν)
                        (NP-ATR (N$ Κυρίου))))                                
                                

    (Matthew 3.3)

Besides gapping and right-node raising, which are discussed separately, there are several types of parenthetical IPs:

  • Asides surrounded on both sides by parts of another clause are labeled IP-MAT-PRN.

    ( (IP-MAT-SPE (CODE {VS:10_42})
                  (CONJ καὶ)
                  (NP-SBJ (CP-FRL (WNP-1 (WPRO ὃς))
                                  (C 0)
                                  (IP-SUB (NP-SBJ *T*-1)
                                          (PRTG ἂν)
                                          (VBS-AOR ποτίσῃ)
                                          (NP-OB2 (NUM ἕνα)
                                                  (NP-PAR (DS$ τῶν) (ADJ$ μικρῶν) (DS$ τούτων)))
                                          (NP-OB1 (NA ποτήριον)
                                                  (NP-ATR (ADJ$ ψυχροῦ)))
                                          (ADVP (ADV μόνον))
                                          (PP (P εἰς)
                                              (NP (NA ὄνομα)
                                                  (NP-ATR (N$ μαθητοῦ)))))))
                  (, ,)
                  (IP-MAT-PRN (INTJ ἀμὴν)
                              (NP-SBJ *pro*)
                              (VBP-IMPF λέγω)
                              (NP-OB2 (PROD ὑμῖν)))
                  (, ,)
                  (NEG οὐ)
                  (NEG μὴ)
                  (VBS-AOR ἀπολέσῃ)
                  (NP-OB1 (DA τὸν)
                          (NA μισθὸν)
                          (NP-ATR (PRO$ αὐτοῦ)))
                  (. .))
      (ID GreekNT,Matthew))    
    

  • Bare reason adjuncts are labeled IP-MAT-PRN. These may not exist in Greek.

     ( (IP-MAT (NP-SBJ (D The) (N river))
               (VBD froze)
               (IP-MAT-PRN (NP-SBJ (PRO it))
                           (BED was)
                           (ADJP (ADVR so) (ADJ cold))))
               (. .))                               
                                

  • said X:

    ( (IP-MAT-SPE (CODE {VS:6_18})
                  (CONJ καὶ)
                  (NP-SBJ (PRO ὑμεῖς))
                  (VBPP-FUT ἔσεσθέ)
                  (NP-OB2 (CLPROD μοι))
                  (PP (P εἰς)
                      (NP (NSA υἱοὺς) (CONJ καὶ) (NSA θυγατέρας)))
                  (, ,)
                  (IP-MAT-PRN (VBP-IMPF λέγει)
                              (NP-SBJ (N Κύριος)
                                      (NP-PRN (NPR Παντοκράτωρ))))
                  (. .))
      (ID GreekNT,2Corinthians))                                
                                

  • Glosses or expansions on words or phrases introduced by that is, that is to say/understand, to wit, and the like are labeled as parenthetical clauses (IP-MAT-PRN or, in the case of to wit, IP-INF-PRN). The gloss itself is treated as a complement of the lowest verb. The category of the gloss is usually NP or PP, but may (at least in the English corpora) be VP (see “VP in Parentheticals” in the online annotation manual for the English corpora) or IP-MAT.

    (IP-SUB (PP (P διὰ)
                (NP (D$ τοῦ) (N$ θανάτου)))
            (NP-SBJ *pro*)
            (VBS-AOR καταργήσῃ)
            (NP-OB1 (DA τὸν)
                    (RRC (NX-1 (DA τὸ)
                               (NA κράτος))
                         (VPRA-IMPF ἔχοντα)
                         (NP-OB1 (NX *ICH*-1)
                                 (NP-ATR (D$ τοῦ) (N$ θανάτου)))
            (, ,)
            (IP-MAT-PRN (NP-SBJ (D τοῦτ'))
                        (BEP-IMPF ἔστι))
                        (NP-OB1 (DA τὸν) (ADJA διάβολον))))                                
                                

    (Hebrews 2.14)

Resumptive

Resumptive elements form a pair with left-dislocated constituents. See also the section called “Resumptive Pronouns”.

Direct Speech

Full sentences of direct speech (matrix clauses, imperatives and direct questions) are given the extended label -SPE. The first main clause following a verb of saying is included in its matrix clause. Any following main clauses are separated (and also labeled -SPE). See also the section called “Conjoined Direct Speech”, the section called “QTP”. The -SPE follows all other CP and finite IP tags—that is, it “trickles down”—except for grammatical function tags like -SBJ or -OB1, which it precedes. This is to facilitate searches for clausal subjects and objects along with ordinary subjects. Any sentence containing a parenthetical he said, etc. is labeled -SPE at the highest level.

  • -SPE “trickles down”:

    ( (IP-MAT (CODE {VS:8_25})
              (NP-SBJ *pro*)
              (VBD εἶπεν)
              (CLPRT δὲ)
              (NP-OB2 (PROD αὐτοῖς))
              (CP-QUE-SPE (WADVP-1 (WADV Ποῦ))
                          (C 0)
                          (IP-SUB-SPE (ADVP-LOC *T*-1)
                                      (BEP-IMPF *)
                                      (NP-SBJ (D ἡ)
                                              (N πίστις)
                                              (NP-ATR (PRO$ ὑμῶν)))))
              (. ;))
      (ID GreekNT,Luke))                                
                                

  • -SPE comes before function tags such as -SBJ:

    ( (IP-MAT (CODE {VS:8_20})
              (VBDP-AOR ἀπηγγέλη)
              (CLPRT δὲ)
              (NP-OB2 (PROD αὐτῷ))
              (IP-MAT-SPE-SBJ (NP-SBJ (NP (D Ἡ)
                                          (N μήτηρ)
                                          (NP-ATR (CLPRO$ σου)))
                                      (CONJP (CONJ καὶ)
                                             (NP (DS οἱ)
                                                 (NS ἀδελφοί)
                                                 (NP-ATR (CLPRO$ σου)))))
                              (VBP-PRF ἑστήκασιν)
                              (ADVP-LOC (ADV ἔξω))
                              (IP-PPL (VBN-AOR-1 ἰδεῖν)
                                      (VPR-IMPF θέλοντές)
                                      (IP-INF-COM (VBN-AOR *ICH*-1)
                                                  (NP-OB1 (CLPROA σε)))))
              (. .))
      (ID GreekNT,Luke))                                
                                

  • sentences containing parenthetical he said are labeled -SPE at the highest level:

    ( (IP-MAT-SPE (CODE {VS:6_18})
                  (CONJ καὶ)
                  (NP-SBJ (PRO ὑμεῖς))
                  (VBPP-FUT ἔσεσθέ)
                  (NP-OB2 (CLPROD μοι))
                  (PP (P εἰς)
                      (NP (NSA υἱοὺς) (CONJ καὶ) (NSA θυγατέρας)))
                  (, ,)
                  (IP-MAT-PRN (VBP-IMPF λέγει)
                              (NP-SBJ (N Κύριος)
                                      (NP-PRN (NPR Παντοκράτωρ))))
                  (. .))
      (ID GreekNT,2Corinthians))                                
                                

Unknown

The label X is used to mark parts of the sentence which are ungrammatical or of which the structure is unknown. Inside a constituent labeled X, words are usually labeled and partial structure may be given to the extent that it is clear. The parts, however, do not form a coherent whole. X is also used when the category of a constituent is unknown in general or is difficult to determine in a specific context.

  • unknown structure:

    ( (IP-MAT (CODE {VS:7_5})
              (IP-ABS (CONJ Καὶ)
                      (CLPRT γὰρ)
                      (VPR$-AOR ἐλθόντων)
                      (NP-SBJ (PRO$ ἡμῶν))
                      (PP (P εἰς)
                          (NP (NPRA Μακεδονίαν))))
              (QA=1 οὐδεμίαν)
              (VBD-PRF ἔσχηκεν)
              (NP-OB1 (NA-1 ἄνεσιν))
              (NP-SBJ (D ἡ)
                      (N σὰρξ)
                      (NP-ATR (PRO$ ἡμῶν)))
              (, ,)
              (CONJ ἀλλ')
              (IP-PPL (PP (P ἐν)
                          (NP (QD παντὶ)))
                      (VPRP-IMPF θλιβόμενοι))
              (X (X (ADVP-LOC (ADV ἔξωθεν))
                    (NP-SBJ (NS μάχαι))
                    (, ,)
                    (CONJP (X (ADVP-LOC (ADV ἔσωθεν))
                              (NP-SBJ (NS φόβοι)))))))
      (ID GreekNT,2Corinthians))                            
                            

  • unknown category of constituent:

    (NP-OB1 (DA τὸ)
            (X ἔντος)
            (NP-COM (NP (D$ τοῦ) (N$ ποτηρίου))
                    (CONJP (CONJ καὶ)
                           (NP (D$ τῆς) (N$ παροψίδος)))))                            
                            

    (Matthew 23.26)



[3] See the Appendix (pp. 699-707) in Hansen and Quinn’s Greek: An Intensive Course for descriptions of these terms.

[4] See Hansen and Quinn’s Greek: An Intensive Course pg. 701.

[5] See Hansen and Quinn’s Greek: An Intensive Course pg. 702.

[6] The usage here is quite odd, but no other type of dative in Hansen and Quinn’s Greek: An Intensive Course or Smyth’s Greek Grammar seems to fit. A Grammatical Analysis of the Greek New Testament by Max Zerwick, S.J. (translated, revised, and adapted by Mary Grosvenor in collaboration with the author, with the assistance of John Welch, S.J.) remarks that this is a case of the dative being used instead of the accusative of duration (p. 207).

[7] See pp. 697-706 in Hansen and Quinn’s Greek: An Intensive Course for explanations of these terms.

Chapter 10. Subordinate Clauses

Adverbial Clauses

Adverbial clauses are either introduced by a subordinating conjunction or not. In the most common cases, the subordinating conjunction may introduce different types of clauses, so deciding how the clause is functioning within a particular context and tagging it appropriately is the main purpose of the annotator’s task. CP-ADV is something of a catch all for all the clauses that do not fit into a more specific category (i.e., CP-THT or CP-PRP); it includes the antecedent (protasis) of conditionals, temporal clauses without a gap, and purpose clauses (excepting those following verbs of motion: see the section called “Purpose Clauses” below). In the case of an adverbial clause introduced by a subordinating conjunction, the subordinating conjunction is tagged as C and the remainder of the subordinate clause is treated as the IP-SUB complement of C. Adverbial ὡς clauses belong to this type and are discussed here in order to facilitate the comparison between these and other types of ὡς clauses. Bare adverbial clauses are very rare in Greek, but there is at least one example in the PPCNTG, which appears below.

  • adverbial clause introduced by ὅτι (which otherwise introduces CP-THT clauses):

    ( (IP-MAT-SPE (CODE {VS:5_5})
                  (ADJP-PRD (ADJ μακάριοι))
                  (BEP-IMPF *)
                  (NP-SBJ (DS οἱ) (ADJ πραεῖς))
                  (, ,)
                  (CP-ADV (C ὅτι)
                          (IP-SUB (NP-SBJ (PRO αὐτοὶ))
                                  (VBP-FUT κληρονομήσουσι)
                                  (NP-OB1 (DA τὴν) (NA γῆν))))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

  • an adverbial ὡς clause without a gap (can be paraphrased ‘when, while, since, as if’):

    (CP-ADV (C ὡς)
            (IP-SUB (PRTG ἂν)
                    (NP-SBJ *pro*)
                    (VBSP-IMPF πορεύωμαι)
                    (PP (P εἰς)
                        (NP (DA τὴν) (NPRA Σπανίαν)))))                            
                            

    (Romans 15.24)

  • a bare adverbial clause from the PPCNTG:

    ( (CP-QUE-SPE (CODE {VS:7_10})
                  (CONJ ἢ)
                  (C 0)
                  (IP-SUB-SPE (CP-ADV (C 0)
                                      (IP-SUB (NP-SBJ *pro*)
                                              (NP-OB1 (ADV καὶ) (NA ἰχθὺν))
                                              (VBP-FUT αἰτήσει)))
                              (NEG μὴ)
                              (NP-OB1 (NA ὄφιν))
                              (NP-SBJ *con*)
                              (VBP-FUT ἐπιδώσει)
                              (NP-OB2 (PROD αὐτῷ)))
                  (. ;))
      (ID GreekNT,Matthew))                            
                            

    Adverbial ὡς clauses with a gap fall into two types, depending on whether ὡς can be paraphrased (roughly) as in the same way that or as which. The first type, which often contains a specifier (just, like, right, so, etc.), is analyzed as a comparative clause in the PPCNTG, analogous to the “as...so...” correlative comparative (see the section called “Comparative Clauses” below, especially correlative comparatives). This represents a departure from the English corpora, in which only an overt correlative component triggers the CP-CMP analysis. In the second type of adverbial ὡς clause with a gap, the CP-ADV contains a nominal gap. This type includes ὡς continuative clauses, which might be treated as clause-adjoined relatives (CP-CAR), but are not so treated in our system. This type may or may not exist in Greek.

    ( (IP-MAT (CONJ and)
              (NP-SBJ (PRO it))
              (VBP serveth)
              (PP (P of)
                  (NP (Q many) (D a) (ADJ necessarie) (N conclusioun)
                      (PP (P in)
                          (NP (NS equacions)
                              (PP (P of)
                                  (NP (NS thinges)))))))
              (PP (P as)
                  (CP-ADV (WNP-1 0)
                          (C 0)
                          (IP-SUB (NP-SBJ *T*-1)
                                  (MD shal)
                                  (BE be)
                                  (VAN shewid))))
              (. .))
    (ID CMASTRO,669.C2.176))                            
                            

That-Clauses

The equivalent of that clauses in English are finite clauses containing the indicative or optative of secondary sequence and introduced by ὅτι or ὡς. Such clauses can appear as the complement to verbs, adjectives, and nouns.

  • complement of verb:

    (IP-PPL (VPR-AOR ἰδὼν)
            (CP-THT (C ὅτι)
                    (IP-SUB (NP-SBJ *pro*)
                            (VBDP-AOR-PASS ἐνεπαίχθη)
                            (PP (P ὑπὸ)
                                (NP-AGT (DS$ τῶν) (NS$ μάγων))))))                            
                            

    (Matthew 2.16)

  • complement of adjective: none in PPCNTG, but the common phrase δῆλον ὅτι ‘it is clear that...’ is a good example

  • complement of noun:

    ( (CP-QUE-SPE (CODE {VS:26_54})
                  (WADVP-1 (WADV πῶς))
                  (CLPRT οὖν)
                  (C 0)
                  (IP-SUB (ADVP *T*-1)
                          (VBSP-AOR-PASS πληρωθῶσιν)
                          (NP-SBJ (DS αἱ)
                                  (NS γραφαὶ)
                                  (CP-THT (C ὅτι)
                                          (IP-SUB (ADVP (ADV οὕτως))
                                                  (VBP-IMPF δεῖ)
                                                  (IP-INF-SBJ (VBNP-AOR γενέσθαι))))))
                 (. ;))
      (ID GreekNT,Matthew))                            
                            

    That clauses (and complement clauses more generally) that are extraposed from ADJP or NP are coindexed with an *ICH* trace.

That clauses (and complement clauses more generally) that are associated with an expletive object it are coindexed with the pronoun, as in the case of extraposed subject that clauses—this case may or may not exist in Greek. That clauses are often in apposition to a demonstrative or other NP. In this case, the NP is labeled as the argument of the verb and the that clause is labeled as an appositive/parenthetical. If the that clause is extraposed, as is often the case, a coindexed *ICH* trace indicates the relationship between the clause and the position where it is interpreted.

( (CP-QUE-SPE (CODE {VS:1_43})
              (CONJ καὶ)
              (WADVP-1 (WADV πόθεν))
              (C 0)
              (IP-SUB (ADVP *T*-1)
                      (NP-OB2 (CLPROD μοι))
                      (BEP-IMPF *)
                      (NP-SBJ (D τοῦτο)
                              (CP-THT-PRN (C ἵνα)
                                          (IP-SUB (VBS-AOR ἔλθῃ)
                                                  (NP-SBJ (D ἡ)
                                                          (N μήτηρ)
                                                          (NP-ATR (D$ τοῦ)
                                                                  (N$ κυρίου)
                                                                  (NP-ATR (CLPRO$ μου))))
                                                  (PP (P πρὸς)
                                                      (NP (PROA ἐμέ)))))))
              (. ;))
  (ID GreekNT,Luke))                
            

Complement Clauses

Complement clauses differ from that clauses in that they are non-veridical complements occurring following certain classes of verbs, mainly “verbs of the head” (i.e., verbs of thinking, speaking, etc.). They can be sub-classified into three types: fear clauses, object clauses of effort, and all other non-veridical complement clauses that sometimes occur—particularly in the New Testament—in place of control or ECM complements. Note that fear clauses as well as some of the “other”-category clauses have the normally POS-tagged with NEG particle μή as a complementizer tagged C.

  • fear clause:

    (IP-PPL (VPRP-IMPF φοβούμενοί)
            (CLTE τε)
            (CP-COM (C μὴ)
                    (IP-SUB (PP (P εἰς)
                                (NP (DA τὴν) (NPRA Σύρτιν)))
                            (NP-SBJ *pro*)
                            (VBS-AOR ἐκπέσωσιν))))                            
                            

    (Acts 27.17)

  • object clause of effort:

    ...μηχανᾶσθαι   [CP −COM    ὅκως   τὸ  σῶμα    τοῦ ἀδελφεοῦ    καταλύσας       κομιεῖ ]
    to-devise       [CP −COM    that   the body    the brother     cutting-down    bring-away ]
    “...to devise that having cut it down he might bring back his brother’s body.”
                            

    (Hdt. 2.121C.2)

  • other:

    ( (IP-MAT (CODE {VS:20_31})
              (NP-SBJ (D ὁ) (CLPRT δὲ) (N ὄχλος))
              (VBD-AOR ἐπετίμησεν)
              (NP-OB2 (PROD αὐτοῖς))
              (CP-COM (C ἵνα)
                      (IP-SUB (NP-SBJ *pro*)
                              (VBS-AOR σιωπήσωσιν)))
              (. .))
      (ID GreekNT,Matthew))                            
                            

  • other w/o complementizer:

    ( (IP-MAT (CODE {VS:20_32})
              (CONJ καὶ)
              (NP-SBJ *con*)
              (VBD-AOR εἶπεν)
              (CP-QUE-SPE (WNP-1 (WPRO Τί))
                          (C 0)
                          (IP-SUB (NP-SBJ *pro*)
                                  (VBP-IMPF θέλετε)
                                  (CP-COM (C 0)
                                          (IP-SUB (NP-OB1 *T*-1)
                                                  (NP-SBJ *pro*)
                                                  (VBS-AOR ποιήσω)
                                                  (NP-OB2 (PROD ὑμῖν))))))
              (. ;))
      (ID GreekNT,Matthew))                            
                            

Purpose Clauses

There are two classes of purpose clauses in Greek: purpose clauses introduced by ἵνα, ὡς, or ὅπως and containing the subjunctive (or optative in 2◦ sequence) serving as an ‘in order that’ adjunct to any type of matrix clause, and purpose infinitives, infinitives preceded by a τοῦ complementizer, and subjunctive clauses following verbs of motion. It is only the latter that take a -PRP tag in the PPCHiG. See infinitives of purpose for examples of the infinitival variety without a complementizer.

  • CP-PRP:

    ( (IP-MAT-SPE (CODE {VS:10_11})
                  (NP-SBJ (PRO ἐγὼ))
                  (VBD-AOR ἦλθον)
                  (CP-PRP (C ἵνα)
                          (IP-SUB (IP-SUB-1 (NP-OB1 (NA ζωὴν))
                                            (NP-SBJ *pro*)
                                            (VBS-IMPF ἔχωσιν))
                                  (CONJP (CONJ καὶ)
                                         (IP-SUB=1 (IP-SMC (ADJP-PRD (ADJA περισσὸν)))
                                                   (NP-SBJ *con*)
                                                   (VBS-IMPF ἔχωσιν)))))
                  (. .))
      (ID GreekNT,John))                            
                            

  • CP-PRP with τοῦ complementizer:

    (IP-MAT-SPE (VBD-AOR Ἐξῆλθεν)
                (NP-SBJ (D ὁ)
                        (RRC (VPR-IMPF σπείρων)))
                (CP-PRP (C τοῦ)
                        (IP-INF (VBN-AOR σπεῖραι)
                                (NP-OB1 (DA τὸν)
                                        (NA σπόρον)
                                        (NP-ATR (PRO$ αὐτοῦ))))))                            
                            

    (Luke 8.5)

  • no -PRP dash tag—CP-ADV instead—because not following verb of motion:

    ( (IP-MAT-SPE (CODE {VS:7_1})
                  (NEG Μὴ)
                  (VBI-IMPF κρίνετε)
                  (, ,)
                  (CP-ADV (C ἵνα)
                          (IP-SUB (NEG μὴ)
                                  (NP-SBJ *pro*)
                                  (VBSP-AOR-PASS κριθῆτε)))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

Degree Complements

Degree complement clauses differ from comparatives in not containing a gap. They are treated as sisters of the head that licenses them (οὕτω, ὡς, ὥστε, etc.). Separated so...that clauses are always treated as extraposed degree complements (rather than as adverbial result or purpose clauses). The traces of extraposed degree complements are treated in the same way as the degree complements themselves—namely, as sisters of the head they are associated with.

( (IP-MAT (CODE {VS:12_21})
          (CONJ0 καί)
          (, ,)
          (ADJP-PRD (ADV οὕτω) (ADJ φοβερὸν)
                    (CP-DEG *ICH*-1))
          (BED-IMPF ἦν)
          (NP-SBJ (D τὸ)
                  (RRC (VPRP-IMPF φανταζόμενον)))
          (, ,)
          (CP-DEG-1 (C 0)
                    (IP-SUB (NP-SBJ (NPR Μωυσῆς))
                            (VBD-AOR εἶπεν)
                            (IP-MAT-SPE (ADJP-PRD (ADJ Ἐκφοβός)
                                                  (CONJP *ICH*-1))
                                        (NP-SBJ *pro*)
                                        (BEP-IMPF εἰμι)
                                        (CONJP-1 (CONJ καὶ)
                                                 (ADJ ἔντρομος)))))
          (. .))
  (ID GreekNT,Hebrews))                
            

In Greek when the degree complement is introduced by ὥστε ‘so that’, there is often not a degree-related adverb or adjective (e.g., such) that can serve as a head taking the degree complement. In these cases the degree complement should be the sister of any adjective or adverb involved in the semantics of the degree relation:

( (IP-MAT (CODE {VS:16_26})
          (ADVP (ADV ἄφνω))
          (CLPRT δὲ)
          (N-1 σεισμὸς)
          (VBDP-AOR ἐγένετο)
          (NP-SBJ (N *ICH*-1)
                  (ADJ μέγας)
                  (CP-DEG (C ὥστε)
                          (IP-INF (VBNP-AOR-PASS σαλευθῆναι)
                                  (NP-SBJ (DSA τὰ)
                                          (NSA θεμέλια)
                                          (NP-ATR (D$ τοῦ) (N$ δεσμωτηρίου))))))
          (, ,))
  (ID GreekNT,Acts))                
            

Result Clauses

Clauses of natural and actual result are similar to degree clauses except that there is no adjectival or adverbial head that licenses them.

  • natural result (infinitival):

    ( (IP-MAT (CODE {VS:10_1})
              (CONJ Καὶ-καί)
              (IP-PPL (VPRP-AOR προσκαλεσάμενος-προσκαλέω)
                      (NP-OB1 (DSA τοὺς-ὁ)
                              (NUM δώδεκα-δώδεκα)
                              (NSA μαθητὰς-μαθητής)
                              (NP-ATR (PRO$ αὐτοῦ-αὐτός))))
              (NP-SBJ *pro*)
              (VBD-AOR ἔδωκεν-δίδωμι)
              (NP-OB2 (PROD αὐτοῖς-αὐτός))
              (NP-OB1 (NA ἐξουσίαν-ἐξουσία)
                      (NP-COM (NS$ πνευμάτων-πνεῦμα) (ADJ$ ἀκαθάρτων-ἀκάθαρτος)))
              (CP-RES (C-CLTE ὥστε-ὥστε)
                      (IP-INF (IP-INF (VBN-IMPF ἐκβάλλειν-ἐκβάλλω)
                                      (NP-OB1 (PROA αὐτὰ-αὐτός)))
                              (CONJP (CONJ καὶ-καί)
                                     (IP-INF (VBN-IMPF θεραπεύειν-θεραπεύω)
                                             (NP-OB1 (NP (QA πᾶσαν-πᾶς) (NA νόσον-νόσος))
                                                     (CONJP (CONJ καὶ-καί)
                                                            (NP (QA πᾶσαν-πᾶς) (NA μαλακίαν-μαλακία))))))))
              (. .))
      (ID GreekNT,Matthew:10_1.449))                            
                            

  • actual result (finite):

    ( (IP-MAT-SPE (CODE {VS:19_5b})
                  (CONJ καὶ-καί)
                  (BEPP-FUT ἔσονται-εἰμί)
                  (NP-SBJ (DS οἱ-ὁ) (NUM δύο-δύο))
                  (PP (P εἰς-εἰς)
                      (NP (NA σάρκα-σάρξ) (NUM μίαν-εἷς)))
                  (. ;)
                  (CODE {VS:19_6})
                  (CP-RES-SPE (C-CLTE ὥστε-ὥστε)
                              (IP-SUB-SPE (ADVP-TMP (NEG+ADV οὐκέτι-οὐκέτι))
                                          (NP-SBJ *pro*)
                                          (BEP-IMPF εἰσὶν-εἰμί)
                                          (NP-PRD (NP (NUM δύο-δύο))
                                                  (CONJP (CONJ ἀλλὰ-ἀλλά)
                                                         (NP (N σὰρξ-σάρξ) (NUM μία-εἷς))))))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

Comparative Clauses

The external structure of comparatives is as follows. The whole comparison is treated as single constituent, which is labeled according to the category of its head. For example, “as tall as...” would be labeled as an ADJP, “more cake than...” would be labeled as an NP, etc. The ὡς or ἤ that introduces the comparative clause is labeled C inside a CP-CMP complement. Internally, all instances of CP-CMP have an empty operator coindexed with a trace indicating the elided target of the comparison.

( (IP-MAT (CODE {VS:14_5})
          (ADJP-PRD (ADJR μείζων)
                    (CP-CMP *ICH*-1))
          (CLPRT δὲ)
          (BEP-IMPF *)
          (NP-SBJ (D ὁ)
                  (RRC (VPR-IMPF προφητεύων)))
          (CP-CMP-1 (WADJP-2 0)
                    (C ἢ)
                    (IP-SUB (ADJP-PRD *T*-2)
                            (BEP-IMPF *)
                            (NP-SBJ (D ὁ)
                                    (RRC (VPR-IMPF λαλῶν)
                                         (NP-INS (NSD γλώσσαις))))))))                
            

(1 Corinthians 14.5)

Often, parts of the comparative clause are elided. If only one element is missing, it is usually indicated by the generic empty category (X *). If more material is missing, gapping should not be employed.

( (IP-MAT (CODE {VS:7_9})
          (ADJP-PRD (ADJR κρεῖττον)
                    (CP-CMP *ICH*-1))
          (CLPRT γάρ)
          (BEP-IMPF ἐστιν)
          (IP-INF-SBJ (VBN-IMPF γαμεῖν))
          (CP-CMP-1 (WADJP-2 0)
                    (C ἢ)
                    (IP-SUB (ADJP-PRD *T*-2)
                            (BEP-IMPF *)
                            (IP-INF-SBJ (VBNP-IMPF πυροῦσθαι))))
          (. .))
  (ID GreekNT,1Corinthians))                
            

(IP-MAT-SPE (CP-ADV (C+PRTG Ἐὰν)
                    (IP-SUB (BES-IMPF ᾖ)
                            (NP-SBJ (D ὁ)
                                    (N ἀριθμὸς)
                                    (NP-ATR (DS$ τῶν)
                                            (NS$ υἱῶν)
                                            (NP-ATR (NPR$ Ἰσραὴλ))))
                            (ADVP (CP-CMP (WADVP-2 (WADV ὡς))
                                          (C 0)
                                          (IP-SUB (ADVP *T*-2)
                                                  (NP-SBJ (D ἡ)
                                                          (N ἄμμος)
                                                          (NP-ATR (D$ τῆς) (N$ θαλάσσης))))
                                  (, ,))))
            (NP-SBJ (D τὸ) (N κατάλειμμα))
            (VBPP-FUT-PASS σωθήσεται))                
            

(Romans 9.27)

Correlative Comparatives

The two comparative constituents in a correlative comparative construction are not treated as a pair of left-dislocated (-LFD) and resumptive (-RSP) elements. In a full correlative comparative (one in which both the demonstrative and relative elements are overt), the comparative clause headed by the relative is embedded within the phrase headed by the demonstrative, via an *ICH* trace if necessary. Note that the Greek corpora differ from the English corpora in this respect. (In the English corpora, the demonstrative and relative head separate phrases and the relative is usually tagged -SPR or -ADT depending on the category.)

  • ADVP correlative:

    ( (IP-MAT-SPE (CODE {VS:13_40})
                  (ADVP (CP-CMP (WADVP-1 (WADV ὥσπερ))
                                    (CLPRT οὖν)
                                    (C 0)
                                    (IP-SUB (IP-SUB-2 (ADVP *T*-1)
                                                      (VBPP-IMPF-PASS συλλέγεται)
                                                      (NP-SBJ (DS τὰ) (NS ζιζάνια)))
                                            (CONJP (CONJ καὶ)
                                                   (IP-SUB=2 (NP-INS (ND πυρὶ))
                                                             (VBPP-IMPF-PASS κατακαίεται)))))
                        (, ,)
                        (ADV οὕτως))
                  (BEP-FUT ἔσται)
                  (PP (P ἐν)
                      (NP (DD τῇ)
                          (ND συντελείᾳ)
                          (NP-ATR (D$ τοῦ) (N$ αἰῶνος))))
                  (. .))
      (ID GreekNT,Matthew))                                
                                

  • ADVP correlative, more suitable dash tag:

    ( (IP-MAT-SPE (CODE {VS:12_34})
                  (ADVP-LOC (CP-CMP (WADVP-1 (WADV ὅπου))
                                    (CLPRT γάρ)
                                    (C 0)
                                    (IP-SUB (ADVP-LOC *T*-1)
                                            (BEP-IMPF ἐστιν)
                                            (NP-SBJ (D ὁ)
                                                    (N θησαυρὸς)
                                                    (NP-ATR (PRO$ ὑμῶν)))))
                            (, ,)
                            (ADV ἐκεῖ))
                  (NP-SBJ (ADV καὶ)
                          (D ἡ)
                          (N καρδία)
                          (NP-ATR (PRO$ ὑμῶν)))
                  (BEP-FUT ἔσται)
                  (. .))
      (ID GreekNT,Luke))                                
                                

Correlative comparatives involving τοσοῦτος/ὅσος in Greek can be particularly tricky. When τοσοῦτος/ὅσος do not modify anything else they should be treated as substantives, with an -ADV dash tag if necessary. The POS tag of the relative of the pair (i.e., ὅσος) should be WADJ*. The POS tag of the resumptive/demonstrative of the pair should be Q*.

( (IP-IMP-SPE (CODE {VS:18_7})
              (NP-SPR (CP-CMP (WNP-1 (WADJA ὅσα)))
                              (C 0)
                              (IP-SUB (IP-SUB-2 (NP-ADV *T*-1)
                                                (NP-SBJ *pro*)
                                                (VBD-AOR ἐδόξασεν)
                                                (NP-OB1 (PROA αὑτὴν)))
                                      (CONJP (CONJ καὶ)
                                             (IP-SUB=2 (VBD ἐστρηνίασεν)))))
              (, ,)
              (QA=3 τοσοῦτον)
              (VBI-AOR δότε)
              (NP-OB2 (PROD αὐτῇ))
              (NP-OB1 (NA-3 βασανισμὸν) (CONJ καὶ) (NA-3 πένθος))
              (. .))
  (ID GreekNT,Revelation))                    
                

Questions

Direct and indirect questions are not differentiated by clause label. Questions contain a C position regardless of whether they exhibit subject-verb inversion. Direct questions in Greek generally do not exhibit inversion, nor do indirect questions. Indirect questions occur as complements.

  • ( (IP-MAT (CODE {VS:20_22})
              (IP-PPL (VPRP-AOR ἀποκριθεὶς))
              (CLPRT δὲ)
              (NP-SBJ (D ὁ) (NPR Ἰησοῦς))
              (VBD-AOR εἶπεν)
              (IP-MAT-SPE (NEG Οὐκ)
                          (NP-SBJ *pro*)
                          (VBP-PRF οἴδατε)
                          (CP-QUE (WNP-1 (WPROA τί))
                                  (C 0)
                                  (IP-SUB (NP-OB1 *T*-1)
                                          (NP-SBJ *pro*)
                                          (VBPP-IMPF αἰτεῖσθε))))
              (. .))
      (ID GreekNT,Matthew))                        
                        
  • ( (IP-MAT (CODE {VS:9_34})
              (PP (P πρὸς)
                  (NP (PROA ἀλλήλους)))
              (CLPRT γὰρ)
              (NP-SBJ *pro*)
              (VBDP-AOR διελέχθησαν)
              (PP (P ἐν)
                  (NP (DD τῇ) (ND ὁδῷ)))
              (CP-QUE (WNP-1 (WPRO τίς))
                      (IP-SUB (NP-SBJ *T*-1)
                              (BEP-IMPF *)
                              (ADJP-PRD (ADJR μείζων))))
              (. .))
      (ID GreekNT,Mark))                        
                        

The term “sluicing” refers to the elision of the entire IP-SUB part of an indirect question. Our annotation includes only the the wh- phrase and the CP-QUE immediately dominating it. The elided IP-SUB is completely omitted, as is the silent complementizer.

( (IP-MAT-SPE (CODE {VS:4_27})
              (CP-QUE (WADVP (WADV ὡς)))
              (NEG οὐκ)
              (VBP-PRF οἶδεν)
              (NP-SBJ (PRO αὐτός))
              (. .))
  (ID GreekNT,Mark))                
            

When indirect questions occur in apposition to a demonstrative or other NP, they carry the extended tag -PRN. If extraposed, they (like extraposed appositive/parenthetical that clauses) are coindexed with a trace within the NP in order to distinguish them from verbal complements.

( (IP-MAT (CODE {VS:3_13})
          (CONJ καὶ)
          (NP-LFD (NP-ATR (ADJ$ ἑκάστου))
                  (DA τὸ)
                  (NA ἔργον)
                  (CP-QUE-PRN (WADJP-2 (WADJ ὁποῖόν))
                              (C 0)
                              (IP-SUB (ADJP-PRD *T*-2)
                                      (NP-SBJ *pro*)
                                      (BEP-IMPF ἐστιν))))
          (NP-SBJ (D τὸ) (N πῦρ))
          (NP-OB1-RSP (PROA αὐτὸ))
          (VBP-FUT δοκιμάσει)
          (. .))
  (ID GreekNT,1Corinthians))                
            

When εἰ heads questions (including direct questions), it is tagged WQ. A complementizer position is always included, as usual.

  • indirect question:

    ( (IP-MAT (CODE {VS:15_44})
              (CONJ καὶ)
              (IP-PPL (VPRP-AOR προσκαλεσάμενος)
                      (NP-OB1 (DA τὸν) (NA κεντυρίωνα)))
              (NP-SBJ *pro*)
              (VBD-AOR ἐπηρώτησεν)
              (NP-OB1 (PROA αὐτὸν))
              (CP-QUE (WQ εἰ)
                      (C 0)
                      (IP-SUB (ADVP-TMP (ADV ἤδη))
                              (NP-SBJ *pro*)
                              (VBD-AOR ἀπέθανεν)))
              (. .))
      (ID GreekNT,Mark))                
                

  • direct question:

    ( (IP-MAT (CODE {VS:13_23})
              (VBD-AOR Εἶπεν)
              (CLPRT δέ)
              (NP-SBJ (PRO τις))
              (NP-OB2 (PROD αὐτῷ))
              (CP-QUE-SPE (NP-VOC (N Κύριε))
                          (, ,)
                          (WQ εἰ)
                          (C 0)
                          (IP-SUB (ADJP-PRD (Q ὀλίγοι))
                                  (BEP-IMPF *)
                                  (NP-SBJ (DS οἱ)
                                          (RRC (VPRP-IMPF-PASS σωζόμενοι)))))
              (. ;))
      (ID GreekNT,Luke))                            
                            

Whether questions that are not direct questions or the complement of a verb, but fill an adverbial function, are labeled CP-QUE-ADV.

( (IP-MAT-SPE (CONJ for)
              (NP-SBJ (PRO ye))
              (MD shal)
              (VB overcome)
              (NP-OB1 (PRO hem)
                      (QP (Q all)))
              (, ,)
              (CP-QUE-ADV (WQ whether)
                          (C 0)
                          (IP-SUB (IP-SUB-1 (NP-SBJ (PRO they))
                                            (VBP wille))
                                  (CONJP (CONJ or)
                                         (IP-SUB=1 (NEG+VBP nylle)))))
              (. .)
              (' '))
  (ID CMMALORY,13.372))                
            

Direct questions (both wh- and yes-no questions) are usually matrix clauses, but can appear in direct speech contexts as the complement of verbs, in which case the dash tag -SPE is added.

  • direct wh- question:

    ( (CP-QUE (CODE {VS:3_5})
              (CP-ADV (C εἰ)
                      (CLPRT δὲ)
                      (IP-SUB (NP-SBJ (D ἡ)
                                      (N ἀδικία)
                                      (NP-ATR (PRO$ ἡμῶν)))
                              (NP-OB1 (NP-ATR (N$ θεοῦ))
                                      (NA δικαιοσύνην))
                              (VBP-IMPF συνίστησιν)))
              (, ,)
              (WNP-1 (WPROA τί))
              (C 0)
              (IP-SUB (NP-OB1 *T*-1)
                      (NP-SBJ *pro*)
                      (VBP-FUT ἐροῦμεν))
              (. ;))
      (ID GreekNT,Romans))                            
                            

  • direct speech wh- question:

    ( (IP-MAT (CODE {VS:13_10})
              (CONJ Καὶ)
              (IP-PPL (VPR-AOR προσελθόντες))
              (NP-SBJ (DS οἱ) (NS μαθηταὶ))
              (VBD-AOR εἶπαν)
              (NP-OB2 (PROD αὐτῷ))
              (CP-QUE-SPE (WPP-1 (P Διὰ)
                                 (WNP (WPRO τί)))
                          (C 0)
                          (IP-SUB (PP *T*-1)
                                  (PP (P ἐν)
                                      (NP (NSD παραβολαῖς)))
                                  (NP-SBJ *pro*)
                                  (VBP-IMPF λαλεῖς)
                                  (NP-OB2 (PROD αὐτοῖς))))
              (. ;))
      (ID GreekNT,Matthew))                            
                            

  • direct yes-no question:

    ( (CP-QUE (CODE :VS:7_13})
              (C 0)
              (IP-SUB (NP-SBJ (D Τὸ) (CLPRT οὖν) (ADJ ἀγαθὸν))
                      (NP-OB2 (PROD ἐμοὶ))
                      (VBDP-AOR ἐγένετο)
                      (NP-PRD (N θάνατος)))
              (. ;))
      (ID GreekNT,Romans))                            
                            

In positively biased yes-no questions, the negative particle is tagged PRTQ and occurs in the C-domain of the question:

( (IP-MAT (CODE {VS:12_23})
          (CONJ καὶ)
          (NP-SBJ *con*)
          (VBD-IMPF ἔλεγον)
          (CP-QUE-SPE (PRTQ Μήτι)
                      (C 0)
                      (IP-SUB (NP-SBJ (D οὗτός))
                              (BEP-IMPF ἐστιν)
                              (NP-PRD (D ὁ)
                                      (N υἱὸς)
                                      (NP-ATR (NPR$ Δαυείδ)))))
          (. ;))
  (ID GreekNT,Matthew))                
            

Wh- in situ

Instances of questions or exclamatives that appear to lack wh- movement and have the wh- word in situ should be indicated by not including a CP-level but marking the wh- phrase with a prefixed W* as usual, although with no movement. Any dash tags should be indicated directly on the wh- phrase.

( (IP-EXL-SPE=1 (CODE {VS:6_23})
                (CP-ADV (C εἰ)
                        (CLPRT οὖν)
                        (IP-SUB-1 (NP-SBJ (D τὸ)
                                          (N φῶς)
                                          (NP-PRN (D τὸ)
                                                  (PP (P ἐν)
                                                      (NP (PROD σοὶ)))))
                                  (NP-PRD (N σκότος))
                                  (BEP-IMPF ἐστίν)))
                (, ,)
                (NP-SBJ (D τὸ) (N σκότος))
                (WADJP-PRD (WADJ πόσον)))
                (. .))
  (ID GreekNT,Matthew))                    
                

Exclamative Clauses

Exclamations are tagged CP-EXL. Their internal syntax is that of questions.

( (IP-MAT (CODE {VS:10_23})
          (CONJ Καὶ)
          (IP-PPL (VPRP-AOR περιβλεψάμενος))
          (NP-SBJ (D ὁ) (NPR Ἰησοῦς))
          (VBP-IMPF λέγει)
          (NP-OB2 (DSD τοῖς)
                  (NSD μαθηταῖς)
                  (NP-ATR (PRO$ αὐτοῦ)))
          (CP-EXL-SPE (WADVP-1 (WADV Πῶς) (ADV δυσκόλως))
                      (C 0)
                      (IP-SUB (ADVP *T*-1)
                              (NP-SBJ (DS οἱ)
                                      (RRC (NP-OB1 (DSA τὰ) (NSA χρήματα))
                                           (VPR-IMPF ἔχοντες)))
                              (PP (P εἰς)
                                  (NP (DA τὴν)
                                      (NA βασιλείαν)
                                      (NP-ATR (D$ τοῦ) (N$ θεοῦ))))
                              (VBPP-FUT εἰσελεύσονται)))
          (. .))
  (ID GreekNT,Mark))                
            

Ordinary Relative Clauses

This section focuses on finite relative clauses. Our annotation system does not distinguish restrictive and non-restrictive relative clauses. Both types are contained within the brackets of their antecedent. Relative clauses that are not contained in (or do not leave a trace of extraposition in) the brackets of their antecedents are given a separate label: CP-CAR (CAR = clause-adjoined relative). As usual, both the wh- constituent and the trace indicate the category of the gap, with additional functional information appearing on the trace only.

  • empty complementizer:

    ( (IP-MAT-SPE (CODE {VS:2_6})
                  (PP (P ἐκ)
                      (NP (PRO$ σοῦ)))
                  (CLPRT γὰρ)
                  (VBPP-FUT ἐξελεύσεται)
                  (NP-SBJ (N ἡγούμενος)
                          (, ,)
                          (CP-REL (WNP-1 (WPRO+CLQ ὅστις))
                                  (C 0)
                                  (IP-SUB (NP-SBJ *T*-1)
                                          (VBP-FUT ποιμανεῖ)
                                          (NP-OB1 (DA τὸν)
                                                  (NA λαόν)
                                                  (NP-ATR (CLPRO$ μου))
                                                  (NP-PRN (DA τὸν) (NPRA Ἰσραήλ)))
                                          (. .)))))
      (ID GreekNT,Matthew))                            
                            

  • rare null operator with ὅτι complementizer:

    (CP-QUE-SPE (WNP-1 (WPRO Τίς))
                (CLPRT ἄρα)
                (C 0)
                (IP-SUB (NP-PRD *T*-1)
                        (NP-SBJ (D οὗτός)
                                (CP-REL *ICH*-2))
                        (BEP-IMPF ἐστιν)
                        (CP-REL-2 (WNP-3 0)
                                  (C ὅτι)
                                  (IP-SUB (NP-SBJ *T*-3)
                                          (NP-OBP (ADV καὶ)
                                                  (NP (DSD τοῖς)
                                                      (NSD ἀνέμοις)
                                                      (CONJP *ICH*-4)))
                                          (VBP-IMPF ἐπιτάσσει)
                                          (CONJP-4 (CONJ καὶ)
                                                   (NP (DD τῷ) (ND ὕδατι)))))))                            
                            

    (Luke 8.25)

  • As usual, extraposed relative clauses are coindexed to an *ICH* trace within the brackets of their antecedent.

    ( (IP-MAT (CODE {VS:12_18})
              (CONJ Καὶ)
              (VBPP-IMPF ἔρχονται)
              (NP-SBJ (NPRS Σαδδουκαῖοι)
                      (CP-REL *ICH*-1))
              (PP (P πρὸς)
                  (NP (PROA αὐτόν)))
              (, ,)
              (CP-REL-1 (WNP-2 (WPRO+CLQ οἵτινες))
                        (C 0)
                        (IP-SUB (NP-SBJ *T*-2)
                                (VBP-IMPF λέγουσιν)
                                (IP-INF-THT (NP-SBJ (NA ἀνάστασιν))
                                            (NEG μὴ)
                                            (BEN-IMPF εἶναι))))
              (, ,))
      (ID GreekNT,Mark))                            
                            

  • As relatives have the same structure as other relative clauses. They always contain an empty operator rather than an overt wh- word, and the complementizer is ὡς.

    ( (IP-MAT (CODE {VS:22_61})
              (CONJ καὶ)
              (VBDP-AOR ὑπεμνήσθη)
              (NP-SBJ (D ὁ) (NPR Πέτρος))
              (NP-OBQ (D$ τοῦ)
                      (N$ ῥήματος)
                      (NP-ATR (D$ τοῦ) (N$ κυρίου))
                      (CP-REL (WNP-1 0)
                              (C ὡς)
                              (IP-SUB (NP-OB1 *T*-1)
                                      (NP-SBJ *pro*)
                                      (VBD-AOR εἶπεν)
                                      (NP-OB2 (PROD αὐτῷ))))
                      (CP-THT-PRN (C ὅτι)
                                  (IP-SUB-SPE (PP (P Πρὶν)
                                                  (IP-INF (NP-SBJ (NA ἀλέκτορα))
                                                          (VBN-AOR φωνῆσαι)
                                                          (ADVP-TMP (ADV σήμερον))))
                                              (NP-SBJ *pro*)
                                              (VBPP-FUT ἀπαρνήσῃ)
                                              (NP-OB1 (CLPROA με))
                                              (ADVP (NUM τρίς)))))
              (. .))
      (ID GreekNT,Luke))                            
                            

Reduced Relative Clauses

[Important] Important

Reduced relatives (RRC) headed by participles are not always easy to distinguish from participial clauses (IP-PPL) (see below for some discussion). In searches for one category, it is wise to include the other.

Reduced relatives (“wh+is” deletion) are labeled RRC (= reduced relative clause). They can be restrictive or non-restrictive, but almost always immediately follow their antecedent. RRCs contain no operator or gap of their own. They may, however, contain a small clause with a subject gap; in this case, a placeholder for the subject (which in a full relative clause would contain the trace of the wh- word) is added in order to make clear the structure. The predicate of a reduced relative is ordinarily a present or past participle, but nonverbal predicates are also possible.

  • typical case:

    ( (IP-MAT-SPE (CODE {VS:12_44})
                  (NP-SBJ (Q πάντες))
                  (CLPRT γὰρ)
                  (NP-OB1 (PP (P ἐκ)
                              (NP (D$ τοῦ)
                                  (RRC (VPR$-IMPF περισσεύοντος)
                                       (NP-OB2 (PROD αὐτοῖς))))))
                  (VBD-AOR ἔβαλον)
                  (, ,))
      (ID GreekNT,Mark))                            
                            

  • containing small clause with subject gap:

    ( (IP-MAT (CODE {VS:1_12})
              (ADVP-TMP (ADV Τότε))
              (NP-SBJ *pro*)
              (VBD-AOR ὑπέστρεψαν)
              (PP (P εἰς)
                  (NP (NA Ἰερουσαλὴμ)))
              (PP (P ἀπὸ)
                  (NP (N$ ὄρους)
                      (NP-PRN (D$ τοῦ)
                              (RRC (VPRP$-IMPF-PASS καλουμένου)
                                   (IP-SMC (NP-SBJ *)
                                           (NP-PRD (N$ Ἐλαιῶνος)))))
                      (, ,)
                      (CP-REL (WNP-1 (WPRO ὅ))
                              (C 0)
                              (IP-SUB (NP-SBJ *T*-1)
                                      (BEP-IMPF ἐστιν)
                                      (PP (P ἐγγὺς)
                                          (NP (NPRA Ἰερουσαλὴμ)))
                                      (IP-PPL (NP-2 (N$ σαββάτου))
                                              (VPR-IMPF ἔχον)
                                              (NP-OB1 (NP-MSR *ICH*-2)
                                                      (NA ὁδόν)))))))
              (. .))
      (ID GreekNT,Acts))                            
                            

Reduced Relative Clauses versus Participial Clauses

As mentioned earlier, reduced relatives are not always easy to distinguish from participial clauses (IP-PPL). We apply the following rules:

  • Certain participial clauses are clearly not reduced relative clauses. For instance, clauses headed by λέγων in the New Testament are almost always instances of IP-PPL.

  • Generally, a participial clause immediately following a subject is labeled RRC, except when it clearly describes an action that takes place simultaneously with that of the main verb.

    (NODE (IP-SUB (PP *T*-2)
                  (NP-SBJ (D an) (N angel))
                  (, ,)
                  (IP-PPL (VAG entryngge)	          ← IP-PPL, not RRC
                          (PP (P in-to)
                              (NP (N prysoun))))
                  (, ,)
                  (VBD turnede)
                  (NP-OB1 (N derknesse))
                  (PP (P in-to)
                      (NP (N li+gt))))
          (ID CMAELR3,28.74))                                
                                

    In Greek, a participle that describes an action taking place immediately before the action of the main verb is also likely to be an IP-PPL—this is the “Taking the money, run!” construction.

    ( (IP-MAT (CODE {VS:2_14})
              (CONJ καὶ)
              (IP-PPL (VPR-AOR ἀναστὰς))
              (NP-SBJ *pro*)
              (VBD-AOR ἠκολούθησεν)
              (NP-OBQ (PROD αὐτῷ))
              (. .))
      (ID GreekNT,Mark))                    
                    

  • Participials serving as the complement of a verb should be treated as small clauses (IP-SMC).

    ( (IP-MAT (CODE {VS:26_43})
              (CONJ καὶ)
              (IP-PPL (VPR-AOR ἐλθὼν)
                      (ADVP (ADV πάλιν)))
              (NP-SBJ *pro*)
              (VBD-AOR εὗρεν)
              (IP-SMC (NP-SBJ (PROA αὐτοὺς))
                      (VPRA-IMPF καθεύδοντας))
              (, ,))
      (ID GreekNT,Matthew))                                
                                

  • Participials in other positions are always labeled IP-PPL except when their interpretation is clearly analogous to that of a restrictive relative clause.

Clause-Adjoined Relative Clauses

The following types of relative clauses are treated as clause-adjoined. With the exception noted below, clause-adjoined relative clauses are daughters of IP.

  • Continuative relatives, where the antecedent of relative clause is the VP or the entire clause.

    ( (IP-MAT-SPE (CODE {VS:19_44})
                  (CONJ καὶ)
                  (NEG οὐκ)
                  (NP-SBJ *con*)
                  (VBP-FUT ἀφήσουσιν)
                  (IP-SMC (NP-SBJ (NA λίθον))
                          (PP (P ἐπὶ)
                              (NP (NA λίθον))))
                  (PP (P ἐν)
                      (NP (PROD σοί))) 
                  (, ,)
                  (CP-CAR (WPP-1 (P ἀνθ')
                                 (WNP (WPRO$ ὧν)))
                          (C 0)
                          (IP-SUB (PP *T*-1)
                                  (NEG οὐκ)
                                  (NP-SBJ *pro*)
                                  (VBD-AOR ἔγνως)
                                  (NP-OB1 (DA τὸν)
                                          (NA καιρὸν)
                                          (NP-ATR (D$ τῆς)
                                                  (N$ ἐπισκοπῆς)
                                                  (NP-ATR (CLPRO$ σου))))))
                  (. .))
      (ID GreekNT,Luke))                            
                            

  • All relative clauses introduced by a wh- modified nominal. Some instances of this type may in fact be matrix clauses with a preposed wh- phrase functioning as a topicalized constituent, but we do not attempt to distinguish the two cases.

    ( (IP-MAT (NP-SBJ (D +Tis) (N vertu)
                      (PP (P of)
                          (NP (N meknesse))))
              (BEP is)
              (NP-OB1 (D a) (ADJ sur) (N fundement)
                      (PP (P of)
                          (NP (Q alle) (N maner)
                              (NP-COM (NS vertues)))))
              (, ,)
              (CP-CAR (WPP-1 (P whit-oute)
                             (WNP (WD which) (N fundement)))
                      (, ,)
                      (C 0)
                      (IP-SUB (PP *T*-1)
                              (NP-LFD (CP-FRL (WNP-2 (WPRO what) (ADV so) (ADV euere))
                                              (C 0)
                                              (IP-SUB (NP-OB1 *T*-2)
                                                      (NP-SBJ (PRO +tu))
                                                      (MD wylt)
                                                      (ADVP (ADV gostly))
                                                      (VB reyse)
                                                      (RP vp))))
                              (, ,)
                              (NP-SBJ-RSP (PRO hit))
                              (VBP falle+t)
                              (RP doun)))
              (. .))
      (ID CMAELR3,32.181))                            
                            

  • Ordinary relative clauses that are separated from their antecedent by one or more clauses. Such relative clauses are treated as clause-adjoined relatives forming a token of their own.

    ( (IP-MAT-SPE (CODE {VS:6_4})
                  (CONJ καὶ)
                  (NP-SBJ *con*)
                  (VBD-AOR ἔδωκεν)
                  (NP-OB2 (DSD τοῖς)
                          (PP (P μετ')
                              (NP (PRO$ αὐτοῦ))))
                  (, ,)
                  (CP-CAR (WNP-1 (WPROA οὓς))
                          (C 0)
                          (IP-SUB (NEG οὐκ)
                                  (VBP-IMPF ἔξεστιν)
                                  (IP-INF-SBJ (NP-OB1 *T*-1)
                                              (VBN-AOR φαγεῖν))
                                  (PP (P εἰ)
                                      (NEG μὴ)
                                      (NP (QA μόνους) (DSA τοὺς) (NSA ἱερεῖς)))))
                  (. ;))
      (ID GreekNT,Luke))                            
                            

  • When the wh- phrase introducing the clause-adjoined relative has no evident relation to the body of the clause, the wh- trace is labeled NP-ADT:

    (NODE (CP-CAR (WNP-1 (D the) (WPRO which))
                  (C 0)
                  (IP-SUB (NP-ADT *T*-1)
                          (NP-SBJ (PRO I))
                          (VBP marvell)
                          (ADVP (ADV greatly))
                          (CP-THT (C that)
                                  (IP-SUB (NP-SBJ (PRO I))
                                          (HVP have)
                                          (NP-OB1 (Q no) (N word))
                                          (PP (P from)
                                              (NP (PRO you)))))))
          (ID APLUMPT,186.92))                            
                            

Free Relatives

Internally, free relative clauses are treated in the same way as indirect questions or ordinary relative clauses. A wh- position and a complementizer position are included in every case. Externally, free relatives are bracketed by function (as subjects, objects, complements of prepositions, adverb phrases, etc.). The basic category enclosing the free relative is identical with the gap in the free relative, but the dash tags on the two categories may differ. This is particularly common with free relative clauses containing WNP. Internally, the WNP is generally coindexed with an argument trace, but externally, the entire free relative often has a non-argument function (NP-ADT, NP-ADV, NP-LFD).

In certain cases, a free relative is preceded by a preposition that does double duty in the sense that it governs the wh- phrase within the free relative, but is also part of the matrix clause. Prepositions doing such double duty are bracketed within the free relative, as in the following examples. These cases should not be confused with the ordinary case of a free relative functioning as the complement of a preposition.

  • free relative:

    ( (IP-IMP-SPE (CODE {VS:10_27})
                  (NP-OB1 (CP-FRL (WNP-1 (WPROA ὃ))
                                  (C 0)
                                  (IP-SUB (NP-OB1 *T*-1)
                                          (NP-SBJ *pro*)
                                          (VBP-IMPF λέγω)
                                          (NP-OB2 (PROD ὑμῖν))
                                          (PP (P ἐν)
                                              (NP (DD τῇ) (ND σκοτίᾳ))))))
                  (, ,)
                  (VBI-AOR εἴπατε)
                  (PP (P ἐν)
                      (NP (DD τῷ) (ND φωτί)))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

  • left-dislocated free relative with resumptive:

    ( (IP-MAT-SPE (CODE {VS:10_33})
                  (NP-LFD (CP-FRL (WNP-1 (WPRO ὅστις)) 
                                  (CLPRT δὲ)
                                  (C 0)
                                  (IP-SUB (NP-SBJ *T*-1)
                                          (VBSP-AOR ἀρνήσηταί)
                                          (NP-OB1 (CLPROA με))
                                          (PP (P ἔμπροσθεν)
                                              (NP (DS$ τῶν) (NS$ ἀνθρώπων))))))
                  (, ,)
                  (VBPP-FUT ἀρνήσομαι)
                  (NP-SBJ (CONJ+PRO κἀγὼ))
                  (NP-OB1-RSP (PROA αὐτὸν))
                  (PP (P ἔμπροσθεν)
                      (NP (D$ τοῦ)
                          (N$ πατρός)
                          (NP-ATR (CLPRO$ μου))
                          (NP-PRN (D$ τοῦ)
                                  (PP (P ἐν)
                                      (NP (DSD τοῖς) (NSD οὐρανοῖς))))))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

  • free relative with preposition not doing “double duty”:

    ( (IP-IMP-SPE (CODE {VS:10_11})
                  (PP (P εἰς)
                      (NP (CP-FRL (WDA-1 ἣν)
                                  (CLPRT δ')
                                  (C 0)
                                  (IP-SUB (PRTG ἂν)
                                          (NP-OB1 (DA *T*-1)
                                                  (NA πόλιν)
                                                  (CONJ ἢ)
                                                  (NA κώμην))
                                          (NP-SBJ *pro*)
                                          (VBS-AOR εἰσέλθητε)))))
                  (, ,)
                  (VBI-AOR ἐξετάσατε)
                  (CP-QUE (WNP-2 (WPRO τίς))
                          (C 0)
                          (IP-SUB (NP-SBJ *T*-2)
                                  (PP (P ἐν)
                                      (NP (PROD αὐτῇ)))
                                  (ADJP-PRD (ADJ ἄξιός))
                                  (BEP-IMPF ἐστιν)))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

  • free relative with preposition doing “double duty”:

    ( (IP-MAT (CODE {VS:7_39})
              (NP-SBJ (N Γυνὴ))
              (VBPP-PRF δέδεται)
              (PP *T*-1)
              (PP (CP-FRL (WPP-1 (P ἐφ')
                                 (WNP (WADJA ὅσον) (NA χρόνον)))
                          (C 0)
                          (IP-SUB (PP *T*-1)
                                  (VBS-IMPF ζῇ)
                                  (NP-SBJ (D ὁ)
                                          (N ἀνὴρ)
                                          (NP-ATR (PRO$ αὐτῆς))))))
              (. .))
      (ID GreekNT,1Corinthians))                            
                            

ὡς Clauses with a Nominal Gap

ὡς clauses with a nominal gap that act as arguments of the matrix verb are treated as free relatives with a nominal empty operator and with ὡς filling the complementizer position.

( (IP-MAT (CODE {VS:26_19})
          (CONJ καὶ)
          (VBD-AOR ἐποίησαν)
          (NP-SBJ (DS οἱ) (NS μαθηταὶ))
          (NP-OB1 (CP-FRL (WNP-1 0)
                          (C ὡς)
                          (IP-SUB (NP-OB1 *T*-1)
                                  (VBD-AOR συνέταξεν)
                                  (NP-OB2 (PROD αὐτοῖς))
                                  (NP-SBJ (D ὁ) (NPR Ἰησοῦς)))))
          (, ,))
  (ID GreekNT,Matthew))                
            

Empty Operator Clauses

Infinitival relatives without introductory wh- phrases are often difficult or impossible to distinguish from purpose clauses, and we have not attempted to distinguish the two. By definition, these clauses contain a gap, but are not introduced by an overt wh- phrase. Purpose infinitives without a gap are discussed elsewhere.

Purpose/relative infinitive clauses are labeled CP-EOP (EOP = empty operator) and are contained within the brackets of their potential antecedent. Like other clauses with gaps, they contain a wh- operator (empty by definition in the case of CP-EOP) coindexed with a trace. Both subject and object gaps are indicated.

  • infinitival relative without an overt wh-phrase:

    ( (IP-MAT (CODE {VS:8_8})
              (IP-PPL (NP-OB1 (DA Ταῦτα))
                      (VPR-IMPF λέγων))
              (NP-SBJ *pro*)
              (VBD-IMPF ἐφώνει)
              (IP-IMP-SPE (NP-SBJ (D Ὁ)
                                  (RRC (VPR-IMPF ἔχων)
                                       (NP-OB1 (NSA ὦτα)
                                               (CP-EOP (WNP-1 0)
                                                       (C 0)
                                                       (IP-INF (NP-SBJ *T*-1)
                                                               (VBN-IMPF ἀκούειν))))))
                          (VBI-IMPF ἀκουέτω))
              (. .))
      (ID GreekNT,Luke))                            
                            

  • an especially tricky case:

    ( (IP-MAT (CODE {VS:7_40})
              (CONJ καὶ)
              (IP-PPL (VPRP-AOR ἀποκριθεὶς))
              (NP-SBJ (D ὁ) (NPR Ἰησοῦς))
              (VBD-AOR εἶπεν)
              (PP (P πρὸς)
                  (NP (PROA αὐτόν)))
              (IP-MAT-SPE (NP-VOC (NPR Σίμων))
                          (, ,)
                          (NP-SBJ *pro*)
                          (VBP-IMPF ἔχω)
                          (NP-CL-1 (CLPROD σοί))
                          (NP-OB1 (PROA τι)
                                  (CP-EOP (WNP-2 0)
                                          (C 0)
                                          (IP-INF (NP-OB2 *CL*-1)
                                                  (NP-OB1 *T*-2)
                                                  (VBN εἰπεῖν)))))
              (. .))
      (ID GreekNT,Luke))                            
                            

Infinitival Clauses

Infinitival syntax in Ancient Greek is very rich. Besides the specific types of infinitival clauses laid out in the following paragraphs, infinitival clauses (IP-INF) can occur in certain temporal clauses, clauses of natural result, articular infinitives, etc.

  • temporal clause:

    ( (IP-MAT (CODE {VS:1_18})
              (IP-ABS (VPRP-AOR-PASS Μνηστευθείσης)
                      (NP-SBJ (D$ τῆς)
                              (N$ μητρὸς)
                              (NP-ATR (PRO$ αὐτοῦ))
                              (NP-PRN (NPR$ Μαρίας)))
                      (NP-OB2 (DD τῷ) (NPRD Ἰωσήφ)))
              (, ,)
              (PP (P πρὶν)
                  (CP-ADV (C ἢ)
                          (IP-INF (VBN-AOR συνελθεῖν)
                                  (NP-SBJ (PROA αὐτοὺς)))))
              (NP-SBJ-1 *pro*)
              (VBDP-AOR-PASS εὑρέθη)
              (IP-SMC (NP-SBJ *-1)
                      (PP (P ἐν)
                          (NP (ND γαστρὶ)))
                      (VPR-IMPF ἔχουσα)
                      (PP (P ἐκ)
                          (NP (N$ πνεύματος) (ADJ$ ἁγίου))))
    	  (. .))
      (ID GreekNT,Matthew))                            
                            

  • clause of natural result:

    ( (IP-MAT (CODE {VS:8_24})
              (CONJ καὶ)
              (INTJ ἰδοὺ)
              (NP-SBJ (N σεισμὸς)
                      (ADJP (ADJ μέγας)
                            (CP-DEG *ICH*-1)))
              (VBDP-AOR ἐγένετο)
              (PP (P ἐν)
                  (NP (DD τῇ) (ND θαλάσσῃ)))
              (, ,)
              (CP-DEG-1 (C ὥστε)
                        (IP-INF (NP-SBJ (DA τὸ) (NA πλοῖον))
                                (VBNP-IMPF καλύπτεσθαι)
                                (PP (P ὑπὸ)
                                    (NP (DS$ τῶν) (NS$ κυμάτων)))))
              (. .))
      (ID GreekNT,Matthew))                            
                            

  • articular infinitive:

    ( (IP-MAT-SPE (CODE {VS:6_8})
                  (VBP-PRF οἶδεν)
                  (CLPRT γὰρ)
                  (NP-SBJ (D ὁ)
                          (N θεὸς)
                          (NP-PRN (D ὁ)
                                  (N πατὴρ)
                                  (NP-ATR (PRO$ ὑμῶν))))
                  (NP-OB1 (CP-FRL (WNP-1 (WPRO$ ὧν))
                                  (C 0)
                                  (IP-SUB (NP-OB1 (NP-COM *T*-1)
                                                  (NA χρείαν))
                                          (NP-SBJ *pro*)
                                          (VBP-IMPF ἔχετε))))
                  (PP (P πρὸ)
                      (NP (D$ τοῦ)
                          (IP-INF (NP-SBJ (PROA ὑμᾶς))
                                  (VBN-AOR αἰτῆσαι)
                                  (NP-OB1 (PROA αὐτόν)))))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

Extraction out of infinitives is treated in the usual way. A-movement (passivization) out of an infinitive is indicated by an (NP-SBJ *) trace coindexed with the moved NP.

( (IP-MAT (CODE {VS:11_5})
          (NP-SBJ-1 *pro*)
          (VBPP-IMPF λογίζομαι)
          (CLPRT γὰρ)
          (IP-INF-THT (NP-SBJ *-1)
                      (NP-ADV (QA μηδὲν))
                      (VBN-PRF ὑστερηκέναι)
                      (NP-OBQ (DS$ τῶν) (ADV ὑπερλίαν) (NS$ ἀποστόλων)))
          (. .))
  (ID GreekNT,2Corinthians))                
            

IP-INF-ADT

[Important] Important

Infinitival adjuncts are likely to lack the -ADT dash tag.

Infinitival adjuncts other than purpose clauses are labeled IP-INF-ADT. There are no examples of IP-INF-ADT in the PPCNTG, although this is likely a mistake.

IP-INF-COM

Infinitives serve as the complements of verbs, nouns, and adjectives. We do not distinguish between raising and control, nor between object control and ECM, and subjects of infinitives are ordinarily not indicated, with the following exceptions:

  • when the subject is extracted out of the clause (probably not consistent in the PPCNTG due to incorrect annotation)

  • in the complements of some object control verbs (namely, those that are ambiguous between object control and ECM) and ECM verbs—this is the always subject rule for accusatives that are co-referent with the subject of an infinitive in the complement of a verb

  • in empty-operator infinitival relative (purpose) clauses

In line with our general principles, when the infinitive clause is an argument of a verb, there is no bracketing to indicate its relation to the governing predicate. The same is true when the infinitive is a complement of a complex predicate. But infinitives that are complements of a noun or an adjective are bracketed together with the governing word.

  • infinitival complements of verbs:

    • complement μέλλω ‘be about to’ (subject control or potentially even raising):

      ( (IP-MAT (CODE {VS:2_13})
                (VBP-IMPF μέλλει)
                (CLPRT γὰρ)
                (NP-SBJ (NPR Ἡρῴδης))
                (IP-INF-COM (VBN-IMPF ζητεῖν)
                            (NP-OB1 (DA τὸ) (NA παιδίον)))
                (CP-PRP (C τοῦ)
                        (IP-INF (VBN-AOR ἀπολέσαι)
                                (NP-OB1 (PROA αὐτό))))
                (. .))
        (ID GreekNT,Matthew))                                            
                                              

    • complement to ἐθέλω ‘want’ (subject control):

      ( (IP-MAT (CODE {VS:2_18})
                (NP-LFD (NPR Ῥαχὴλ)
                        (RRC (VPR-IMPF κλαίουσα)
                             (NP-OB1 (DSA τὰ)
                                     (NSA τέκνα)
                                     (NP-ATR (PRO$ αὐτῆς))))
                (, ,)
                (CONJ καὶ)
                (NEG οὐκ)
                (NP-SBJ-RSP *pro*)
                (VBD-IMPF ἤθελεν)
                (IP-INF-COM (VBNP-AOR-PASS παρακληθῆναι))
                (CP-ADV (C ὅτι)
                        (IP-SUB (NEG οὐκ)
                                (NP-SBJ *pro*)
                                (BEP-IMPF εἰσίν)))
                (. .))
        (ID GreekNT,Matthew))                                            
                                              

    • complement to πέμπω ‘send’ (probably object control, but annotated like ECM because somewhat ambiguous):

      ( (IP-MAT-SPE (CODE {VS:22_16})
                    (NP-SBJ (PRO Ἐγὼ)
                            (NP-PRN (NPR Ἰησοῦς)))
                    (VBD-AOR ἔπεμψα)
                    (IP-INF-COM (NP-SBJ (DA τὸν)
                                        (NA ἄγγελόν)
                                        (NP-ATR (CLPRO$ μου)))
                                (VBN-AOR μαρτυρῆσαι)
                                (NP-OB2 (PROD ὑμῖν))
                                (NP-OB1 (DSA ταῦτα))
                                (PP (P ἐπὶ)
                                    (NP (DSD ταῖς) (NSD ἐκκλησίαις))))
                    (. .))
        (ID GreekNT,Revelation))                                            
                                              

    • complement to δίδωμι (unambiguous object control because dative object is co-referent with subject of infinitive):

      ( (IP-MAT-SPE (CODE {VS:5_26})
                    (ADVP (CP-CMP (WADVP-1 (WADV ὥσπερ))
                                  (CLPRT γὰρ)
                                  (C 0)
                                  (IP-SUB (ADVP *T*-1)
                                          (NP-SBJ (D ὁ) (N πατὴρ))
                                          (VBP-IMPF ἔχει)
                                          (NP-OB1 (NA ζωὴν))
                                          (PP (P ἐν)
                                              (NP (PROD+SLF ἑαυτῷ)))))
                          (, ,)
                          (ADV οὕτως))
                    (NP-OB2 (ADV καὶ) (DD τῷ) (ND υἱῷ))
                    (NP-SBJ *pro*)
                    (VBD-AOR ἔδωκεν)
                    (IP-INF-COM (NP-OB1 (NA ζωὴν))
                                (VBN-IMPF ἔχειν)
                                (PP (P ἐν)
                                    (NP (PROD+SLF ἑαυτῷ))))
                    (. .))
        (ID GreekNT,John))                                            
                                              

  • infinitival complements of nouns:

    • without movement away from the infinitive clause:

      ( (IP-IMP-SPE (CODE {VS:12_5})
                    (VBSP-AOR φοβήθητε)
                    (NP-OB1 (DA τὸν)
                            (RRC (PP (P μετὰ)
                                     (NP (DA τὸ)
                                         (IP-INF (VBN-AOR ἀποκτεῖναι))))
                                 (VPRA-IMPF ἔχοντα)
                                 (NP-OB1 (NA ἐξουσίαν)
                                         (IP-INF-COM (VBN-AOR ἐμβαλεῖν)
                                                     (PP (P εἰς)
                                                         (NP (DA τὴν) (NA γέενναν)))))))
                    (. .))
        (ID GreekNT,Luke))                                        
                                          

    • with movement away from the infinitive clause (very common):

      ( (IP-MAT-SPE (CODE {VS:14_18})
                    (CONJ καὶ)
                    (NP-SBJ *pro*)
                    (VBP-IMPF ἔχω)
                    (NA-1 ἀνάγκην)
                    (IP-PPL (VPR-AOR ἐξελθὼν))
                    (NP-OB1 (NA *ICH*-1)
                            (IP-INF-COM (VBN-AOR ἰδεῖν)
                                        (NP-OB1 (PROA αὐτόν))))
                    (. .))
        (ID GreekNT,Luke.1117))                                        
                                          

  • infinitival complements of adjectives:

    • without movement away from the infinitive clause:

      ( (IP-MAT-SPE (CODE {VS:15_18})
                    (ADVP-TMP (NEG+ADV οὐκέτι))
                    (NP-SBJ *pro*)
                    (BEP-IMPF εἰμὶ)
                    (ADJP-PRD (ADJ ἄξιος)
                              (IP-INF-COM (VBNP-AOR-PASS κληθῆναι)
                                          (IP-SMC (NP-SBJ *)
                                                  (NP-PRD (N υἱός)
                                                          (NP-ATR (CLPRO$ σου)))))
                    (. .))
        (ID GreekNT,Luke))                                            
                                              

    • with movement away from the infinitive clause (quite common):

      ( (IP-MAT-SPE (CODE {VS:3_15})
                    (ADJ-1 ὀξεῖς)
                    (BEP-IMPF *)
                    (NP-SBJ (DS οἱ)
                            (NS πόδες)
                            (NP-ATR (PRO$ αὐτῶν)))
                    (ADJP-PRD (ADJ *ICH*-1)
                              (IP-INF-COM (VBN-AOR ἐκχέαι)
                                          (NP-OB1 (NA αἷμα))))
                    (, ,))
        (ID GreekNT,Romans))                                            
                                              

IP-INF-DEG

All degree infinitives are treated as sisters of the degree word that licenses them; in English, these are enough, so, too.

  • Degree infinitives with a gap. (CP-DEG) Degree infinitives with a gap are labeled CP-DEG and take an IP-INF complement.

    ( (IP-MAT-SPE (CONJ and)
                  (NP-SBJ *con*)
                  (ADVP-TMP (ADV now))
                  (HVP hast)
                  (VBN seyd)
                  (NP-OB1 (ADJR j-now)
                          (CP-DEG (WNP-1 0)
                                  (IP-INF (TO to)
                                          (BE be)
                                          (VAN brent)
                                          (PP (P for)
                                              (NP *T*-1)))))
                  (. .)
                  (" "))
      (ID CMKEMPE,133.3127))                                
                                

  • Degree infinitives without a gap. (IP-INF-DEG) Degree infinitives without a gap are labeled IP-INF- DEG. Note that the infinitive is a sister to the phrasal projection of its licensing head, if there is a phrasal projection.

    (NODE (IP-SUB (NP-SBJ (NPR Crist))
                  (BEP is)
                  (ADJP (ADVR so) (ADJ merciable)
                        (IP-INF-DEG (TO to)
                                    (VB for+gyue)
                                    (NP-OB2 (D a) (ADJ synful) (N man))
                                    (NP-OB1 (PRO$ his) (NS synnes)))))
          (ID CMAELR3,43.513))                                
                                

IP-INF-PRP

Purpose infinitives without a gap and following verbs of motion (see the section called “Purpose Clauses” above) are labeled IP-INF-PRP.

( (IP-MAT-SPE (CODE {VS:2_2})
              (CONJ καὶ)
              (NP-SBJ *con*)
              (VBD-AOR ἤλθομεν)
              (IP-INF-PRP (VBN-AOR προσκυνῆσαι)
                          (NP-OBP (PROD αὐτῷ)))
              (. .))
  (ID GreekNT,Matthew))                    
                

IP-INF-SBJ

In addition to serving as complements, infinitives also act as subjects. Infinitival subjects in Ancient Greek are often articular infinitives. In contrast to the English corpora, extraposed infinitival subjects are not co- indexed with a null expletive since there is no evidence for overt expletives in Greek at all.

  • articular:

    ( (IP-MAT (CODE {VS:1_21})
              (NP-OB2 (PROD Ἐμοὶ))
              (CLPRT γὰρ)
              (NP-SBJ (D τὸ)
                      (IP-INF (VBN-IMPF ζῇν)))
              (BEP-IMPF *)
              (NP-PRD (NPR Χριστὸς)))
      (ID GreekNT,Philippians))                                
                                

  • non-articular:

    ( (IP-MAT (CODE {VS:27_6})
              (NP-SBJ (D Οἱ) (CLPRT δὲ) (NS ἀρχιερεῖς))
              (IP-PPL (VPR-AOR λαβόντες)
                      (NP-OB1 (DSA τὰ) (NSA ἀργύρια)))
              (VBD-AOR εἶπαν)
              (IP-MAT-SPE (NEG Οὐκ)
                          (VBP-IMPF ἔξεστιν)
                          (IP-INF-SBJ (VBN-AOR βαλεῖν)
                                      (NP-OB1 (PROA αὐτὰ))
                                      (PP (P εἰς)
                                          (NP (DA τὸν) (NA κορβανᾶν))))
                          (, ,)
                          (CP-ADV (C ἐπεὶ)
                                  (IP-SUB (NP-PRD (N τιμὴ)
                                                  (NP-ATR (N$ αἵματός)))
                                          (NP-SBJ *pro*)
                                          (BEP-IMPF ἐστιν))))
              (. .))
      (ID GreekNT,Matthew))                                
                                

IP-INF-THT

Infinitival clauses of indirect discourse in Greek are labeled IP-INF-THT. At least the following verbs may (but perhaps not always!) introduce infinitives of indirect discourse[8]:

  • φήμι ‘say, declare’

  • νομίζω ‘consider’

  • λέγω ‘say, speak’

( (IP-MAT (CODE {VS:22_23})
          (PP (P Ἐν)
              (NP (DD ἐκείνῃ) (DD τῇ) (ND ἡμέρᾳ)))
          (VBD-AOR προσῆλθον)
          (NP-OBP (PROD αὐτῷ))
          (NP-SBJ (NPRS Σαδδουκαῖοι))
          (, ,)
          (IP-PPL (VPR-IMPF λέγοντες)
                  (IP-INF-THT (NEG μὴ)
                              (BEN-IMPF εἶναι)
                              (NP-SBJ (NA ἀνάστασιν))))
          (, ,))
  (ID GreekNT,Matthew))                    
                

Small Clauses

[Important] Important

See also the sections on secondary predicates (ADJP-SPR, NP-SPR).

Small clauses are instances of predication where the predicate is a noun phrase, adjective phrase (possibly headed by a participle), participle, or other plausible category for predication (i.e., including predicates that consist in just a prepositional phrase). This marks a departure from the criteria used in the English corpora, where only noun phrases, adjective phrases, and passive participles are allowed as small clause predicates. Crucially, the predicate in a small clause should not be an infinitive.

  • noun phrase predicate:

    ( (IP-MAT (CODE {VS:6_20})
              (NP-SBJ (D ὁ) (CLPRT γὰρ) (NPR Ἡρῴδης))
              (VBDP-IMPF ἐφοβεῖτο)
              (NP-OB1 (DA τὸν) (NPRA Ἰωάνην))
              (, ,)
              (IP-PPL (VPR-PRF εἰδὼς)
                      (IP-SMC (NP-SBJ (PROA αὐτὸν))
                              (NP-PRD (NA ἄνδρα)
                                      (ADJP (ADJA δίκαιον) (CONJ καὶ) (ADJA ἅγιον)))))
              (, ,))
      (ID GreekNT,Mark))                            
                            

  • adjective phrase predicate:

    ( (IP-IMP-SPE (CODE {VS:3_8})
                  (VBI-AOR ποιήσατε)
                  (CLPRT οὖν)
                  (IP-SMC (NP-SBJ (NA καρπὸν))
                          (ADJP-PRD (ADJA ἄξιον)
                                    (NP-COM (D$ τῆς) (N$ μετανοίας))))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

  • prepositional phrase predicate:

    (CONJ καὶ)
    (NEG οὐκ)
    (NP-SBJ *con*)
    (VBP-FUT ἀφήσουσιν)
    (IP-SMC (NP-SBJ (NA λίθον))
            (PP (P ἐπὶ)
                (NP (NA λίθον))))
    (PP (P ἐν)
        (NP (PROD σοί)))                            
                            

    (Luke 19.44)

  • active participle predicate:

    ( (IP-MAT (CODE {VS:26_43})
              (CONJ καὶ)
              (IP-PPL (VPR-AOR ἐλθὼν)
                      (ADVP (ADV πάλιν)))
              (NP-SBJ *pro*)
              (VBD-AOR εὗρεν)
              (IP-SMC (NP-SBJ (PROA αὐτοὺς))
                      (VPRA-IMPF καθεύδοντας))
              (, ,))
      (ID GreekNT,Matthew))                            
                            

  • passive participle predicate:

    ( (IP-MAT (CODE {VS:7_30})
              (CONJ καὶ)
              (IP-PPL (VPR-AOR ἀπελθοῦσα)
                      (PP (P εἰς)
                          (NP (DA τὸν)
                              (NA οἶκον)
                              (NP-ATR (PRO$ αὐτῆς)))))
              (NP-SBJ *pro*)
              (VBD-AOR εὗρεν)
              (IP-SMC (IP-SMC (NP-SBJ (DA τὸ) (NA παιδίον))
                              (VPRPA-PRF βεβλημένον)
                              (PP (P ἐπὶ)
                                  (NP (DA τὴν) (NA κλίνην))))
                      (CONJP (CONJ καὶ)
                             (IP-SMC (NP-SBJ (DA τὸ) (NA δαιμόνιον))
                                     (VPRA-PRF ἐξεληλυθός))))
              (. .))
      (ID GreekNT,Mark))                            
                            

The Greek equivalents of the following English verbs should be considered to take small clauses. (Cf. the list of verbs that take secondary predicates in the section called “Secondary Predicate NP”.)

ACKNOWLEDGE, ACCOUNT, ADJUDGE, AFFIRM, ALLOW, APPROVE, ARGUE, BELIEVE, BETHINK, BID, CALL, CAUSE, CONCEIVE, CONCLUDE, CONFESS, CONFIRM, CONSIDER, CONSTITUTE, CONTINUE, COUNT, CREATE, DECLARE, DEEM, DENOMINATE, DENY, DESIGN, DETERMINE, DISCLOSE, DISCOVER, DO, DOUBT, DUB, ESPY, EFFORM, ESTEEM, EXPECT, EXPRESS, FANCY, FEEL, FEIGN, FIND, FORESEE, FORM, GET, GRANT, GUESS, HAVE, HEAR, HOLD, HOPE, IMAGINE, IMPLY, INTEND, JUDGE, KNOW, LEARN, LET, MAKE, MEAN, NOTE, NUMBER, OBSERVE, OWN, OVERSEE, PERCEIVE, PHRASE, PRESUME, PRETEND, PROCLAIM, PROFESS, PRONOUNCE, PROTEST, PROVE, RECKON, RELATE, REMEMBER, RENDER, REPORT, REPUTE, SAY, SEE, SHOW, SPEAK, SPY, STYLE, SUPPOSE, SUSPECT, TERM, THINK, TITLE, TRUST, TURN, UNDERSTAND, VALUE, WARRANT, WISH, WRITE, WYST, YIELD

In addition, the following Greek verbs should be considered to take small clauses.

Participial Clauses

[Important] Important

Participial clauses (IP-PPL) are not always easy to distinguish from reduced relatives (RRC) headed by participles (see above for some discussion). In searches for one category, it is wise to include the other.

Participial clauses without overt subjects are labeled IP-PPL. As usual in our system, the PRO subject, which is generally co-referential with the matrix subject, is not indicated. On rare occasions, an IP-PPL will have an overt subject, but for the most part participial clauses with overt subjects are genitive absolutes and labeled IP-ABS.

IP-PPLs in Greek are adjuncts (IP-PPL) or complements (covered below in the section called “Supplementary Participles” and the section called “Participles of Indirect Discourse”).

  • adjunct participial clause:

    ( (IP-MAT (CODE {VS:2_8})
              (CONJ καὶ)
              (IP-PPL (VPR-AOR πέμψας)
                      (NP-OB1 (PROA αὐτοὺς))
                      (PP (P εἰς)
                          (NP (NPRA Βηθλεὲμ))))
              (VBD-AOR εἶπεν)
              (IP-IMP-SPE (IP-PPL (VPR-AOR Πορευθέντες))
                          (VBI-AOR ἐξετάσατε)
                          (ADVP (ADV ἀκριβῶς))
                          (PP (P περὶ)
                              (NP (D$ τοῦ) (N$ παιδίου))))
              (. .))
      (ID GreekNT,Matthew))                            
                            

  • adjunct participial clause with a (rare) subject:

    ( (IP-MAT (CODE {VS:2_36})
              (CONJ Καὶ)
              (BED-IMPF ἦν)
              (NP-SBJ (NP-PRN (NPR Ἇννα))
                      (N προφῆτις)
                      (, ,)
                      (NP-PRN (N θυγάτηρ)
                              (NP-ATR (NPR$ Φανουήλ)))
                      (, ,)
                      (PP (P ἐκ)
                          (NP (N$ φυλῆς)
                              (NP-ATR (NPR$ Ἀσήρ)))))
              (, ,)
              (IP-PPL (IP-PPL (NP-SBJ (D αὕτη))
                              (VPR-PRF προβεβηκυῖα)
                              (PP (P ἐν)
                                  (NP (NSD ἡμέραις) (QD πολλαῖς))))
                      (, ,)
                      (CONJP (IP-PPL (VPR-AOR ζήσασα)
                                     (PP (P μετὰ)
                                         (NP (N$ ἀνδρὸς)))
                                     (NP-MSR (NSA ἔτη) (NUM ἑπτὰ))
                                     (PP (P ἀπὸ)
                                         (NP (D$ τῆς)
                                             (N$ παρθενίας)
                                             (NP-ATR (PRO$ αὐτῆς)))))))
              (, ,))
      (ID GreekNT,Luke))                            
                            

Absolute Clauses

IP-ABS is the default for any participial clause with a subject, unless the subject of the participle is the same as the subject of the whole clause (as in the example above, which is just labeled IP-PPL). In other words, the distinguishing feature of an absolute clause is that (1) its subject is marked and (2) its subject is distinct from the subject of the matrix clause. Most absolute participial clauses in Greek occur in the genitive—these are the genitive absolutes of the traditional grammatical terminology for Greek. Occasionally, accusative absolutes also occur (although none seem to occur in the PPCNTG), but these are not marked any differently since it is possible to find them in a search by specifying the case of the subject and/or participle.

( (IP-MAT (CODE {VS:4_35})
          (CONJ Καὶ)
          (NP-SBJ *con*)
          (VBP-IMPF λέγει)
          (NP-OB2 (PROD αὐτοῖς))
          (PP (P ἐν)
              (NP (DD ἐκείνῃ) (DD τῇ) (ND ἡμέρᾳ)))
          (IP-ABS (NP-SBJ (ADJ$ ὀψίας))
                  (VPRP$-AOR γενομένης))
          (IP-MAT-SPE (NP-SBJ *pro*)
                      (VBS-AOR Διέλθωμεν)
                      (PP (P εἰς)
                          (NP (DA τὸ) (ADV πέραν))))
          (. .))
  (ID GreekNT,Mark))                    
                

Supplementary Participles

“Supplementary participles,” in the traditional terminology for Greek grammar, are actually participial complements of a small number of verbs; they are labeled IP-PPL-COM.

( (IP-MAT (CODE {VS:5_4})
          (CP-ADV (C ὡς)
                  (CLPRT δὲ)
                  (IP-SUB (NP-SBJ *pro*)
                          (VBDP-AOR ἐπαύσατο)
                          (IP-PPL-COM (VPR-IMPF λαλῶν))))
          (, ,)
          (NP-SBJ *pro*)
          (VBD-AOR εἶπεν)
          (PP (P πρὸς)
              (NP (DA τὸν) (NPRA Σίμωνα)))
          (IP-IMP-SPE (VBI-AOR Ἐπανάγαγε)
                      (PP (P εἰς)
                          (NP (DA τὸ) (NA βάθος)))))
  (ID GreekNT,Luke))                    
                

There are three major categories of verbs that may select for supplementary participles:

Participles of Indirect Discourse

Participial clauses in Greek also occur as a third alternative to finite clauses and infinitives in indirect discourse; these are labeled IP-PPL-THT. Only certain verbs are able to take participles of indirect discourse; similar-looking cases with different verbs should be considered small clauses (IP-SMC).

( (IP-MAT (CODE {VS:17_6})
          (CONJ0 καὶ)
          (NP-SBJ *pro*)
          (VBD-AOR εἶδον)
          (IP-PPL-THT (NP-SBJ (DA τὴν) (NA γυναῖκα))
                      (VPRA-IMPF μεθύουσαν)
                      (PP (PP (P ἐκ)
                              (NP (D$ τοῦ)
                                  (N$ αἵματος)
                                  (NP-ATR (DS$ τῶν) (ADJ$ ἁγίων))))
                          (CONJP (CONJ καὶ)
                                 (PP (P ἐκ)
                                     (NP (D$ τοῦ)
                                         (N$ αἵματος)
                                         (NP-ATR (DS$ τῶν)
                                                 (NS$ μαρτύρων)
                                                 (NP-ATR (NPR$ Ἰησοῦ))))))))
          (. .))
  (ID GreekNT,Revelation))                    
                

The verbs that (optionally) select for participles of indirect discourse are the following[9]:



[8] From CUNY Summer Greek Institute Day 21, Unit 16 Handout by Hardy Hansen.

[9] From CUNY Summer Greek Institute Day 21, Unit 16 Handout by Hardy Hansen.

Chapter 11. Special Constructions

Prolepsis

Smyth defines prolepsis as the case when “The subject of the dependent clause is...anticipated and made the object of the verb of the principal clause.” This construction is especially common—again, according to Smyth—following verbs of saying, seeing, hearing, knowing, fearing, and effecting. In our annotation scheme, prolepsis is indicated by retaining the proleptic object within the outer scope of the dependent clause and giving it a -LFD tag, while the *pro* subject of the dependent clause receives a -RSP tag.

( (IP-IMP-SPE (CODE {VS:6_28})
              (VBI-AOR καταμάθετε)
              (CP-QUE (NP-OB1-LFD (DSA τὰ)
                                  (NSA κρίνα)
                                  (NP-COM (D$ τοῦ) (N$ ἀγροῦ)))
                      (WADVP-1 (WADV πῶς))
                      (C 0)
                      (IP-SUB (ADVP *T*-1)
                              (NP-SBJ-RSP *pro*)
                              (VBP-IMPF αὐξάνουσιν)))
              (. .))
  (ID GreekNT,Matthew))                
            

Clausal Subjects

Clausal subjects that contain a complementizer are labeled CP-SBJ.

( (IP-MAT-SPE (CODE {VS:5_29})
              (VBP-IMPF συμφέρει)
              (CLPRT γάρ)
              (NP-OBP (CLPROD σοι))
              (CP-SBJ (C ἵνα)
                      (IP-SUB (IP-SUB (VBSP-AOR ἀπόληται)
                                      (NP-SBJ (NUM ἓν)
                                              (NP-PAR (DS$ τῶν)
                                                      (NS$ μελῶν)
                                                      (NP-ATR (CLPRO$ σου)))))
                              (CONJP (CONJ καὶ)
                                     (IP-SUB (NEG μὴ)
                                             (NP-SBJ (ADJ ὅλον)
                                                     (D τὸ)
                                                     (N σῶμά)
                                                     (NP-ATR (CLPRO$ σου)))
                                             (VBSP-AOR-PASS βληθῇ)
                                             (PP (P εἰς)
                                                 (NP (NA γέενναν)))))))
              (. .))
  (ID GreekNT,Matthew))                
            

In the GNT, however, matrix clauses without a CP projection can also occur as the subject of a verb like ἐγένετο ‘(it) happened.’ In this case, the clausal subject is simply indicated with the tag IP-MAT-SBJ.

( (IP-MAT (CODE {VS:7_28})
          (CONJ Καὶ)
          (VBDP-AOR ἐγένετο)
          (CP-ADV (C ὅτε)
                  (IP-SUB (VBD-AOR ἐτέλεσεν)
                          (NP-SBJ (D ὁ) (NPR Ἰησοῦς))
                          (NP-OB1 (DSA τοὺς) (NSA λόγους) (DSA τούτους))))
          (, ,)
          (IP-MAT-SBJ (VBDP-IMPF-PASS ἐξεπλήσσοντο)
                      (NP-SBJ (DS οἱ) (NS ὄχλοι))
                      (PP (P ἐπὶ)
                          (NP (DD τῇ)
                              (ND διδαχῇ)
                              (NP-ATR (PRO$ αὐτοῦ)))))
          (. .))
  (ID GreekNT,Matthew))                
            

'If not' = 'except'

The combination of εἰ + μή when it means 'except' can be tricky to represent. There are two strategies we employ to make it easy to find instead of εἰ + μή = 'except.'

  • Instaces of εἰ in combination with μή to mean 'except' receive a special lemma: εἰ#1.

    The negator μή is always represented as the immediate sister of εἰ, no matter what the larger structure of the phrase is. The common types of εἰ + μή constructions are as follows:

    • εἰ + μή in a CP-ADV introducing a gapped IP-SUB

      ( (IP-MAT-SPE-1 (CODE {VS:11_27a})
                      (CONJ καὶ-καί)
                      (NP-SBJ (NEG+Q οὐδεὶς-οὐδείς))
                      (VBP-IMPF ἐπιγινώσκει-ἐπιγινώσκω)
                      (NP-OB1 (DA τὸν-ὁ) (NA υἱὸν-υἱός))
                      (CP-ADV (C εἰ-εἰ#1)
                              (NEG μὴ-μή)
                              (IP-SUB=1 (NP-SBJ (D ὁ-ὁ) (N πατήρ-πατήρ))))
                      (, ,))
        (ID GreekNT,Matthew))                           
                              

    • εἰ + μή in a PP serving as the object of the main verb with a null head noun:

      ( (IP-MAT (CODE {VS:14_17})
                (NP-SBJ (DS οἱ-ὁ))
                (CLPRT δὲ-δέ)
                (VBP-IMPF λέγουσιν-λέγω)
                (NP-OB2 (PROD αὐτῷ-αὐτός))
                (IP-MAT-SPE (NEG Οὐκ-οὐ)
                            (NP-SBJ *pro*)
                            (VBP-IMPF ἔχομεν-ἔχω)
                            (ADVP-LOC (ADV ὧδε-ὧδε))
                            (NP-OB1 (PP (P εἰ-εἰ#1)
                                        (NEG μὴ-μή)
                                        (NP (NP (NUM πέντε-πέντε) (NSA ἄρτους-ἄρτος))
                                            (CONJP (CONJ καὶ-καί)
                                                   (NP (NUM δύο-δύο) (NSA ἰχθύας-ἰχθύς)))))))
                (. .))
        (ID GreekNT,Matthew))                            
                              

    • εἰ + μή in a CP-ADV serving as a sort of associate to the object of the main verb:

      ( (IP-MAT-1 (CODE {VS:21_19a})
                  (CONJ καὶ-καί)
                  (NP-OB1 (NEG+QA οὐδὲν-οὐδείς))
                  (NP-SBJ *con*)
                  (VBD-AOR εὗρεν-εὑρίσκω)
                  (PP (P ἐν-ἐν)
                       (NP (PROD αὐτῇ-αὐτός)))
                  (CP-ADV (C εἰ-εἰ#1)
                          (NEG μὴ-μή)
                          (IP-SUB=1 (NP-OB1 (NSA φύλλα-φύλλον) (ADV μόνον-μόνος))))
                  (, ,))
        (ID GreekNT,Matthew))                                        
                                          

    • εἰ + μή in a recursive PP (CP-ADV and IP-SUB gapping doesn't seem appropriate):

      ( (IP-MAT (CODE {VS:15_24})
                (NP-SBJ (D ὁ-ὁ))
                (CLPRT δὲ-δέ)
                (IP-PPL (VPRP-AOR ἀποκριθεὶς-ἀποκρίνομαι))
                (VBD-AOR εἶπεν-λέγω)
                (IP-MAT-SPE (NEG Οὐκ-οὐ)
                            (NP-SBJ *pro*)
                            (VBDP-AOR ἀπεστάλην-ἀποστέλλω)
                            (PP (P εἰ-εἰ#1)
                                (NEG μὴ-μή)
                                (PP (P εἰς-εἰς)
                                    (NP (DSA τὰ-ὁ)
                                        (NSA πρόβατα-πρόβατον)
                                        (NP-PRN (DSA τὰ-ὁ)
                                                (RRC (VPRA-PRF ἀπολωλότα-ἀπόλλυμι)))
                                                (NP-ATR (N$ οἴκου-οἶκος)
                                                        (NP-ATR (NPR$ Ἰσραήλ-Ἰσραήλ)))))))
                (. .))
        (ID GreekNT,Matthew))                                        
                                          

Chapter 12. Discontinuous Phrases

This section gives guidelines for representing discontinuous phrases—most commonly, discontinuous noun phrases—in the PPCHiG. The basic guiding principle of representing discontinuous phrases in these parsed corpora is that all movement should be represented as leftward movement if possible. Certain defined exceptions to this generalization are represented as rightward movement or clitic displacement.

Rightward Extraposition

It appears that restricting rightward movement of phrases to the same categories in which rightward movement is annotated in the English corpora is sufficient for Ancient Greek. Rightward movement of phrases is thus allowed only with:

  • CONJP

    ( (IP-MAT (CODE {VS:14_2})
              (NP-SBJ (D ὁ)
                      (CLPRT γὰρ)
                      (RRC (VPR-IMPF λαλῶν)
                           (NP-INS (ND γλώσσῃ))))
              (NEG οὐκ)
              (NP-OB2 (NSD ἀνθρώποις)
                      (CONJP *ICH*-1))
              (VBP-IMPF λαλεῖ)
              (CONJP-1 (CONJ ἀλλὰ)
                       (ND θεῷ))
              (, ,))
      (ID GreekNT,1Corinthians))                            
                            

  • PP

    ( (IP-MAT (CODE {VS:20_17})
              (IP-PPL (VPR-IMPF Μέλλων)
                      (CLPRT δὲ)
                      (IP-INF-COM (VBN-IMPF ἀναβαίνειν)
                                  (PP *ICH*-1)))
              (NP-SBJ (NPR Ἰησοῦς))
              (PP-1 (P εἰς)
                    (NP (NPRA Ἰεροσόλυμα)))
              (VBD-AOR παρέλαβεν)
              (NP-OB1 (DSA τοὺς) (NUM δώδεκα) (NSA μαθητὰς))
              (PP (P κατ')
                  (NP (ADJA ἰδίαν)))
              (, ,))
      (ID GreekNT,Matthew))                         
                            

  • CP-REL

    ( (IP-MAT-SPE (CODE {VS:1_8_4})
                  (ADVP-TMP (ADV πάλαι))
                  (CLPRT δὲ)
                  (NP-SBJ (DS τὰ)
                          (ADJ καλὰ)
                          (CP-REL *ICH*-1))
                  (NP-AGT (NSD ἀνθρώποισι))
                  (VBPP-PRF-PASS ἐξεύρηται)
                  (, ,)
                  (CP-REL-1 (WPP-2 (P ἐκ)
                                   (WNP (WPRO$ τῶν)))
                            (IP-SUB (IP-INF-SBJ (PP *T*-2)
                                                (VBN-IMPF μανθάνειν))
                                    (VBP-IMPF δεῖ)))
                  (. ·))
      (ID Herodotus,Histories))                            
                            

  • RRC (reduced relative clauses)

    ( (IP-MAT-SPE (CODE {VS:5_14})
                  (NEG οὐ)
                  (VBPP-IMPF δύναται)
                  (NP-SBJ (N πόλις)
                          (RRC *ICH*-1))
                  (IP-INF-COM (VBNP-AOR-PASS κρυβῆναι))
                  (RRC-1 (PP (P ἐπάνω)
                             (NP (N$ ὄρους)))
                         (VPAP-IMPF κειμένη))
                  (. .))
      (ID GreekNT,Matthew))                            
                            

  • CP-CMP (comparative clauses)

    ( (IP-MAT (CODE {VS:14_5})
              (ADJP-PRD (ADJR μείζων)
                        (CP-CMP *ICH*-1))
              (CLPRT δὲ)
              (BEP-IMPF *)
              (NP-SBJ (D ὁ)
                      (RRC (VPR-IMPF προφητεύων)))
              (CP-CMP-1 (WADJP-2 0)
                        (C ἢ)
                        (IP-SUB (ADJP-PRD *T*-2)
                                (BEP-IMPF *)
                                (NP-SBJ (D ὁ)
                                        (RRC (VPR-IMPF λαλῶν)
                                             (NP-INS (NSD γλώσσαις))))))
              (, ,)
              (PP (P ἐκτὸς)
                  (CP-ADV (C εἰ)
                          (IP-SUB (NEG μὴ)
                                  (NP-SBJ *pro*)
                                  (VBS-IMPF διερμηνεύῃ)
                                  (, ,)
                                  (CP-ADV (C ἵνα)
                                          (IP-SUB (NP-SBJ (D ἡ) (N ἐκκλησία))
                                                  (NP-OB1 (NA οἰκοδομὴν))
                                                  (VBS-AOR λάβῃ))))))
    	  (. .))
      (ID GreekNT,1Corinthians))                            
                            

  • CP-DEG (degree complements)

    ( (CP-QUE-SPE (CODE {VS:1_32_1})
                  (" ")
                  (INTJ ὦ)
                  (NP-VOC (N ξεῖνε)
                          (ADJ Ἀθηναῖε))
                  (, ,)
                  (IP-SUB (NP-SBJ (D ἡ)
                                  (CLPRT δ’)
                                  (ADJ-POS ἡμετέρη)
                                  (N εὐδαιμονίη))
                          (ADVP (ADV οὕτω)
                                (CP-DEG *ICH*-1))
                          (NP-AGT (PROD τοι))
                          (VBPP-PRF-PASS ἀπέρριπται)
                          (PP (P ἐς)
                              (NP (DA τὸ)
                                  (NEG+Q μηδὲν)))
                          (CP-DEG-1 (C ὥστε)
                                    (IP-SUB (NEG+CONJ οὐδὲ)
                                            (IP-SMC (ADJP-PRD (NP-COM (NS$ ἰδιωτέων)
                                                                      (NS$ ἀνδρῶν))
                                                              (ADJA ἀξίους))
                                                    (NP-SBJ (PROA ἡμέας)))
                                            (NP-SBJ *pro*)
                                            (VBD-AOR ἐποίησας))))
                  (. ;)
                  (" "))
      (ID Herodotus,Histories))                            
                            

  • CP-QUE

    ( (IP-MAT (CODE {VS:26_22})
              (CONJ καὶ)
              (IP-PPL (VPRP-IMPF λυπούμενοι)
                      (ADVP (ADV σφόδρα)))
              (VBDP-AOR ἤρξαντο)
              (IP-INF-COM (VBN-IMPF λέγειν)
                          (NP-OB2 (PROD αὐτῷ))
                          (CP-QUE-SPE *ICH*-1))
              (NP-SBJ (NUM εἷς) (Q ἕκαστος))
              (CP-QUE-SPE-1 (PRTQ Μήτι)
                            (IP-SUB (NP-SBJ (PRO ἐγώ))
                                    (BEP-IMPF εἰμι)
                                    (, ,)
                                    (NP-VOC (N κύριε))))
              (. ;))
      (ID GreekNT,Matthew))                            
                            

  • *-PRN (appositives and parentheticals)

    (IP-MAT-SPE (NP-SBJ (N φωνὴ)
                        (NP-PRN *ICH*-1))
                (PP (P ἐν)
                    (NP (NPRD Ῥαμὰ)))
                (VBDP-AOR-PASS ἠκούσθη)
                (, ,)
                (NP-1 (NP (N κλαυθμὸς))
                      (CONJP (CONJ καὶ)
                             (NP (N ὀδυρμὸς) (ADJ πολύς)))))                            
                            

    (Matthew 2.18)

Care should be taken to represent rightward movement only as a last resort. For example, the following very common (in the Greek New Testament) construction of participle fronting should be represented as constituting leftward movement of the participle, not rightward movement of the PP.

  • like this:

    ( (IP-MAT (CODE {VS:8_14})
              (CONJ0 Καὶ)
              (VPR-AOR-1 ἐλθὼν)
              (NP-SBJ (D ὁ) (NPR Ἰησοῦς))
              (IP-PPL (VPR-AOR *ICH*-1)
                      (PP (P εἰς)
                          (NP (DA τὴν)
                              (NA οἰκίαν)
                              (NP-ATR (NPR$ Πέτρου)))))
              (VBD-AOR εἶδεν)
              (NP-OB1 (DA τὴν)
                      (NA πενθερὰν)
                      (NP-ATR (PRO$ αὐτοῦ))
                      (RRC (VPRPA-PRF βεβλημένην)
                           (CONJ καὶ)
                           (VPRA-IMPF πυρέσσουσαν)))
              (. .))
      (ID GreekNT,Matthew))                            
                            

  • not like this:

    ( (IP-MAT (CODE {VS:8_14})
              (CONJ0 Καὶ)
              (IP-PPL (VPR-AOR ἐλθὼν)
                      (PP *ICH*-1))
              (NP-SBJ (D ὁ) (NPR Ἰησοῦς))
              (PP-1 (P εἰς)
                    (NP (DA τὴν)
                        (NA οἰκίαν)
                        (NP-ATR (NPR$ Πέτρου))))
              (VBD-AOR εἶδεν)
              (NP-OB1 (DA τὴν)
                      (NA πενθερὰν)
                      (NP-ATR (PRO$ αὐτοῦ))
                      (RRC (VPRPA-PRF-PASS βεβλημένην)
                           (CONJ καὶ)
                           (VPRA-IMPF πυρέσσουσαν)))
              (. .))
      (ID GreekNT,Matthew))                            
                            

Default Leftward Movement

Word-Level Fronting

Both modifiers and heads of phrases can be moved leftward in Classical Greek (the latter case may be a case of remnant movement).

  • fronted modifier:

    (ADJD-1 δευτέρῃ)
    (CLPRT δὲ)
    (NP-SBJ *pro*)
    (VBP-IMPF λέγουσι)
    (NP-TMP (ADJD *ICH*-1)
            (ND γενεῇ)
            (PP (P μετὰ)
                (NP (DSA ταῦτα))))                                
                                

    (Hdt. 1.3.1)

  • fronted head:

    ( (IP-MAT (CODE {VS:1_6_2})
              (NP-SBJ (D οὗτος)
                      (D ὁ)
                      (NPR Κροῖσος))
                      (ADJP (NP-PAR (ADJ$ βαρβάρων)
                                    (CP-REL *ICH*-1))
                            (ADJ πρῶτος))
                            (CP-REL-1 (WNP-2 (WPRO$ τῶν))
                                      (C 0)
                                      (IP-SUB (NP-OB1 *T*-2)
                                              (NP-SBJ (PRO ἡμεῖς))
                                              (VBP-PRF ἴδμεν)))
              (DSA-3 τοὺς)
              (CLPRT μὲν)
              (VBDP-AOR κατεστρέψατο)
              (NP-OB1 (DSA *ICH*-3)
                      (NP-PAR (NPRS$ Ἑλλήνων)))
              (PP (P ἐς)
                   (NP (NP-COM (N$ φόρου))
                       (NA ἀπαγωγήν)))
    	  (, ,))
      (ID Herodotus,Histories))                                
                                

Word-level wh- movement is also possible.

(NP-LFD (CP-FRL (WADJ-1 ὅσος-ὅσος)
                (C 0)
                (IP-SUB (BED-IMPF ἦν-εἰμί)
                        (PP (P ἐν-ἐν)
                            (NP (DD τῷ-ὁ) (ND ἄστεϊ-ἄστυ)))
                        (NP-SBJ (ADJ *T*-1)
                                (N σῖτος-σῖτος)
                                (NP-ATR (CONJ καὶ-καί)
                                        (NP-ATR (PRO$+SLF ἑωυτοῦ-ἑαυτοῦ))
                                        (CONJP (CONJ καὶ-καί)
                                               (ADJP (ADJ ἰδιωτικός-ἰδιωτικός))))))))            
                

(Hdt. 1.21.2)

Fronting of Non-Constituents

On occasion, more than one word from a phrase—the words together not forming a constituent—will move to the left, stranding behind other element(s) of the phrase. This occurs most commonly with noun phrases, but it can occur with other types of phrases as well, including inflectional phrases (IPs). The non-constituent grouping is given a label suffixed with -Y: IY, NY, PY etc.

  • nominal non-constituent (NY):

    ( (IP-MAT (CODE {VS:1_5_4})
              (NP-LFD (CP-FRL (WNP-1 (WPRO τὰ))
                              (CLPRT γὰρ)
                              (IP-SUB (NP-SBJ *T*-1)
                                      (NP-TMP (DA τὸ)
                                              (ADV πάλαι))
                                      (ADJP-PRD (ADJ μεγάλα))
                                      (BED-IMPF ἦν))))
              (, ,)
              (NY-2 (DS τὰ)
                    (Q πολλὰ))
              (ADJP-PRD (ADJ σμικρὰ))
              (NP-SBJ (NY *ICH*-2)
                      (NP-PAR-RSP (PRO$ αὐτῶν)))
              (VBP-PRF γέγονε)
              (. ·))
      (ID Herodotus,Histories))                                
                                

  • inflectional non-constituent (IY):

    (IP-PPL (IY-2 (VBN-AOR ἐπικατασφάξαι)
                  (NP-OB1 (CLPROA μιν)))
            (VPR-IMPF κελεύων)
            (IP-INF-COM (IY *ICH*-2)
                        (NP-OBP (DD τῷ)
                                (ADJD νεκρῷ))))                            
                                

    (Hdt. 1.45.1)

Wh- movement of a non-constituent is also possible.

( (IP-IMP-SPE (CODE {VS:10_11})
              (NP-LFD (CP-FRL (WPP (WP-1 εἰς)
                                   (WNP (WDA-2 ἣν)))
                              (CLPRT δ')
                              (C 0)
                              (IP-SUB (PRTG ἂν)
                                      (PP (P *T*-1)
                                          (NP (DA *T*-2)
                                              (NA (NA πόλιν)
                                                  (CONJ ἢ)
                                                  (NA κώμην))))
                                      (NP-SBJ *pro*)
                                      (VBS-AOR εἰσέλθητε))))
              (, ,)
              (VBI-AOR ἐξετάσατε)
              (CP-QUE (WNP-3 (WPRO τίς))
                      (C 0)
                      (IP-SUB (NP-SBJ *T*-3)
                              (PP (P ἐν)
                                  (NP-RSP (PROD αὐτῇ)))
                              (ADJP-PRD (ADJ ἄξιός))
                              (BEP-IMPF ἐστιν)))
	      (. .))
  (ID GreekNT,Matthew))                    
                

Phrase-Level Fronting

A whole phrase can also be moved leftward. As with the examples above, the fronted phrase often ends up in pre-verbal position.

( (IP-INF-THT (CODE {VS:1_2_3})
              (NP-SBJ (DSA τοὺς))
              (CLPRT δὲ)
              (VBNP-AOR ὑποκρίνασθαι)
              (CP-THT (C ὡς)
                      (IP-SUB (NEG+CONJ οὐδὲ)
                              (NP-SBJ (DS ἐκεῖνοι))
                              (NP-1 (NP-PRN (NPR$ Ἰοῦς))
                                    (D$ τῆς)
                                    (ADJ$ Ἀργείης))
                              (VBD-AOR ἔδοσάν)
                              (NP-OB2 (CLPROD σφι))
                              (NP-OB1 (NSA δίκας))
                              (NP-OBQ (NP-COM *ICH*-1)
                                      (D$ τῆς)
                                      (N$ ἁρπαγῆς))))
              (. ·))
  (ID Herodotus,Histories))                    
                

Clitic Displacement

Discontinuities resulting from the intervention of prosodically weak clitic elements that break up otherwise continuous phrases are a different case from those considered thus far. The “movement” or, rather, displacement of clitics is a different process that has been argued to occur at a later stage in the derivation of a sentence than syntactic movement operations like those considered above—namely, during the process that takes the output of syntactic operations as its input, linearizes this input, and maps it onto a phonological form.

Sentence-Position Particles

The very common second-position particles are represented with the special POS tag CLPRT alone.

( (IP-MAT-SPE (CODE {VS:1_21})
              (NP-SBJ *pro*)
              (VBPP-FUT τέξεται)
              (CLPRT δὲ)
              (NP-OB1 (NA υἱὸν)))
  (ID GreekNT,Matthew))                    
                

When a second-position particle intervenes in a phrase, it is simply attached as high as possible.

( (IP-MAT (CODE {VS:1_8_1})
          (NP-SBJ (D οὗτος) 
                  (CLPRT δὴ)
                  (CLPRT ὦν)
                  (D ὁ)
                  (NPR Κανδαύλης))
          (VBDP-AOR ἠράσθη)
          (NP-OBQ (D$ τῆς)
                  (NP-ATR (PRO$+SLF ἑωυτοῦ))
                  (N$ γυναικός))
          (, ,))
  (ID Herodotus,Histories))                       
                

Clitic Pronouns and Verbs

Discontinuities also occur when the intervention of clitic pronouns and clitic forms of the verbs εἰμί ‘be’ and φημί ‘say’ results in a discontinuous phrase. The clitic pronouns and verbal forms have a more complex distribution, and so these are represented with a dash tag -CL and a special type of movement trace *CL* at the proper hierarchical level where the clitic element is interpreted. There are no restrictions on whether this trace occurs to the left or to the right of the displaced element, but an effort is made to put the trace in the likely position where the clitic originated in the syntax given that clitics often appear in second position, inverted in linear order with either the first (rightward movement) or the last element (leftward movement) of some phrase.

  • intervening clitic pronoun:

    ( (IP-MAT-SPE (CODE {VS:1_37_2})
                  (ADVP-TMP (ADV νῦν))
                  (CLPRT δὲ)
                  (NP-OB1 *CL*-1)
                  (NP-OBP (Q$ ἀμφοτέρων)
                          (NP-CL-1 (CLPROA με))
                          (DS$ τούτων))
                  (VPR-AOR ἀποκληίσας)
                  (NP-SBJ *pro*)
                  (VBP-IMPF ἔχεις)
                  (, ,)
                  (IP-PPL (NP-OB1 (NEG+CLTE οὔτε)
                                  (NP (CLQA τινὰ)
                                      (NA δειλίην))
                                  (CONJP *ICH*-2))
                          (NP-OBP (CLPROD μοι))
                          (VPR-AOR παριδὼν)
                          (CONJP-2 (NEG+CLTE οὔτε)
                                   (NP (NA ἀθυμίην)))))
      (ID Herodotus,Histories))                                
                                

  • intervening clitic verb:

    ( (IP-MAT-SPE (CODE {VS:12_8})
                  (BEP-IMPF *CL*-1)
                  (NP-PRD (N κύριος)
                          (CLPRT γάρ)
                          (BEP-IMPF-CL-1 ἐστιν)
                          (NP-ATR (D$ τοῦ) (N$ σαββάτου)))
                  (NP-SBJ (D ὁ)
                          (N υἱὸς)
                          (NP-ATR (D$ τοῦ) (N$ ἀνθρώπου)))
                  (. .))
      (ID GreekNT,Matthew.579))                                
                                

Chapter 13. Disfluencies

Table of Contents

Break
False Start
Repetition
Tag

The following labels and dash tags are used to annotate disfluencies.

Break

BREAK dominates 0 and indicates that a phrase or sentence breaks off or is otherwise left unfinished.

( (IP-MAT (NP-SBJ (PRO They))
          (MD $could)
          (NEG $n't)
          (VB believe)
          (CP-THAT (C 0)
                   (IP-SUB (NP-SBJ (PRO we))
                           (VBD saw)
                           (NP-OB1 (D a) (BREAK 0))))
           (. ...)))                
            

False Start

FS indicates a false start. Within a false start, internal structure is indicated as far as is feasible, but breaks are not specially indicated.

( (IP-MAT (FS (IP-MAT (NP-SBJ (PRO I))
                      (DOD $did)
                      (NEG $n't)
                      (VB know)))
          (, -)
          (FS (IP-MAT (NP-SBJ (CP-FRL (WNP-1 (WPRO what))
                                      (C 0)
                                      (IP-SUB (NP-SBJ (PRO I))
                                              (VBP mean)
                                              (IP-INF (NP-ACC *T*-1)
                                                      (TO to)
                                                      (VB say)))))
                      (BEP is)))
          (NP-SBJ (PRO I))
          (BEP am)
          (ADJP (ADV very) (ADJ sorry)
                (IP-INF (TO to)
                        (VB hear)
                        (PP (P of)
                            (NP (PRO$ your) (N loss)))))
          (. .)))                
            

Repetition

REP is a full-fledged tag indicating the exact repetition of a constituent for rhetorical or other reasons.

( (IP-MAT (NP-SBJ (PRO It))
          (BEP is)
          (NP-OB1 (D a) (N trifle)
                  (REP (NP (D a) (N trifle))))
          (. .)))                
            

Tag

The dash tag -TAG on a question indicates that it is a tag question.

( (IP-MAT (NP-SBJ-1 (EX There))
          (HVP have)
          (BEN been)
          (NP=1 (NS riots))
          (PP (P in)
              (NP (NPR Calabria))
          (CP-QUE-TAG (IP-SUB (HVP $have)
                              (NEG $n't)
                              (NP-SBJ (EX there))))
          (? .)))                
            

Part III. Planned Corpus Development and Known Issues

Chapter 14. Known Issues

State of the Corpora

As of November 1, 2013, the Penn Parsed Corpora of Historical Greek are no longer being developed. This chapter gives the details of the state of the corpus and advice on proceeding with the work to get the corpus into a state where it can be used—with confidence—for research.

State of Each Corpus

  • Herodotus, Histories: Books 1-3 parsed, some TODOs on individual sentences remaining (indicated with (CODE {TODO:[...desc. of TODO]}) tags), but otherwise corrected and sanity-checked.

    Filenames:

    • Book 1: Histories1.psd

    • Book 2: Histories2.psd

    • Book 3: Histories3.psd

  • Greek New Testament: all parsed, in various states of correction

    Filenames and progress details:

New POS Tags (and Parsing Guidelines)

Since the Penn Parsed Corpora of Historical Greek were designed following the model of the Penn Parsed Corpora of Historical English, at first the POS tags used mirrored the English tags quite closely and—given the vast differences between English and Greek—inappropriately. At a later stage, the corpora were switched to use an expanded POS tagging and parsing scheme (as described, for the most part, in this manual, although Matthew and Mark from the Greek New Testament and the first book of Herodotus' Histories should be taken as the gold standard), but this transition was not completed for the Gospels beyond Matthew and Mark, and from Acts to the remainder of the New Testament, the corpus remains in the original, English-style format.

To Be Revised

When revising the Greek New Testament corpus beyond Matthew and Mark, in general the following things all need to be checked:

  • add NP-ATR and NP-PAR where appropriate

  • add NP-OBP and NP-OBQ where appropriate

  • add -PASS where appropriate

  • fix CP-CMPs where needed

  • add -TAG dash tag to tag questions where appropriate

  • fix movement (see "How to Check Movement" in docs/ directory of GitHub repo)

  • syntactic and POS correction queries (see "GNT Correction Queries" in docs/ directory of GitHub repo)

  • check that words are split correctly (in particular, find all cases of crasis and fix any cases where breathing marks are not in the first part of the split word)

  • collect D lemmas from Matthew and Mark for D-to-DEM, etc. queries

  • write and run Matthew and Mark D-to-DEM, etc. queries

Revisions to Herodotus

Some things should also be checked in Herodotus' Histories.

  • make sure all books of HDT have D vs. DEM distinction

  • check that all clitic pronouns in HDT have CL* tags

  • fix TODOs in HDT

Appendix A. Appendix

Copular Verbs

APPEAR, BE, BECOME, CHANCE, CONTINUE, FALL, FEEL, GET, GROW, KEEP, LOOK, PLEAD (in legal sense), PROVE, REMAIN, SAVOR, SEEM, SHOW, SMELL, SOUND, STAY, TASTE, TURN, WAX

Verbs taking Small Clauses

ACKNOWLEDGE, ACCOUNT, ADJUDGE, AFFIRM, ALLOW, APPROVE, ARGUE, BELIEVE, BETHINK, BID, CALL, CAUSE, CONCEIVE, CONCLUDE, CONFESS, CONFIRM, CONSIDER, CONSTITUTE, CONTINUE, COUNT, CREATE, DECLARE, DEEM, DENOMINATE, DENY, DESIGN, DETERMINE, DISCLOSE, DISCOVER, DO, DOUBT, DUB, ESPY, EFFORM, ESTEEM, EXPECT, EXPRESS, FANCY, FEEL, FEIGN, FIND, FORESEE, FORM, GET, GRANT, GUESS, HAVE, HEAR, HOLD, HOPE, IMAGINE, IMPLY, INTEND, JUDGE, KNOW, LEARN, LET, MAKE, MEAN, NOTE, NUMBER, OBSERVE, OWN, OVERSEE, PERCEIVE, PHRASE, PRESUME, PRETEND, PROCLAIM, PROFESS, PRONOUNCE, PROTEST, PROVE, RECKON, RELATE, REMEMBER, RENDER, REPORT, REPUTE, SAY, SEE, SHOW, SPEAK, SPY, STYLE, SUPPOSE, SUSPECT, TERM, THINK, TITLE, TRUST, TURN, UNDERSTAND, VALUE, WARRANT, WISH, WRITE, WYST, YIELD

Verbs taking Secondary Predicates

ADMIT, AGREE, ANOINT, APPOINT, BIND, BRING, CHOOSE, CHRISTEN, CLAIM, CONFIRM, CONSECRATE, CROWN, DEFEND, DEFINE, DEVOTE, ELECT, INTERPRET, KEEP, LEAVE, LOVE, NAME, NOMINATE, OFFER, ORDAIN, PRESERVE, SET, SUBSCRIBE, SUFFER, SURNAME, TAKE, TESTIFY, TRANSLATE, TRY, VOW, WITNESS

Cognate Object Verbs

DELIGHT, HELP, PAIN, PRAY, THANK