Subject-Verb Agreement
A **finite verb** normally agrees with its subject in **person and number**, not gender or case. Its inflection identifies first, second, or third person and singular or plural in NT Greek. The inherited dual is not a productive inflectional category in the NT; reference to two does not require a dual form. Participles, unlike finite forms, have gender, number, and case. The following sections explain established qualifications to simple singular–singular and plural–plural agreement.
Greek permits an unexpressed subject. **Constructed paradigm forms:** **λέγω** ('I say'), **λέγομεν** ('we say'), **λέγει** ('he/she/it says'), **λέγουσι(ν)** ('they say'). The ending identifies person/number, but a third-person ending alone does not identify the referent or its gender. Overt nouns and pronouns have several discourse functions; their presence does not automatically prove emphasis. See Chapters 50–51.
When a coordinated subject receives resolved plural agreement, first person normally takes precedence over second/third, and second over third. **ἐγὼ καὶ ὁ πατὴρ ἕν ἐσμεν** (John 10:30) has first plural **ἐσμεν**. A verb can instead agree with one conjunct in appropriate constructions; the hierarchy is not an exceptionless prediction for every coordinated phrase. Identify the subject before using a clause as agreement evidence.
**HBS BYZ04:** **John 12:19:** **οἱ οὖν Φαρισαῖοι εἶπον πρὸς ἑαυτούς**, 'So the Pharisees said to one another': **εἶπον** is third plural aorist active indicative here, agreeing with the plural subject. Do not quote εἶπον but parse a different spelling εἶπαν. **Matthew 14:5:** **ἐφοβήθη τὸν ὄχλον**, 'he feared the crowd': **ἐφοβήθη** is third singular aorist passive-form indicative of φοβέομαι with the contextual sense 'feared'; Herod is the understood subject and the crowd the object.
Noun-Adjective Agreement
An adjective normally agrees with its controller in **gender, number, and case (GNC)**. The features must be distinguished from the shapes of endings: different declensions can express the same features. **Constructed examples:** **ὁ ἀγαθὸς λόγος** (masculine nominative singular), **τῇ καλῇ ἡμέρᾳ** (feminine dative singular), and **τὸ ἀγαθὸν ἔργον** (neuter nominative/accusative singular).
A clearer cross-declension example is **τὴν ἁγίαν πόλιν** (Matthew 4:5): the adjective's feminine accusative singular form belongs to its first-declension pattern, while **πόλιν** is a third-declension feminine accusative singular noun. **πᾶσα σάρξ** also illustrates differing endings, but **πᾶς, πᾶσα, πᾶν** is a mixed third/first-declension adjective, not an ordinary first/second-declension adjective.
Grammatical gender normally governs a directly modifying adjective. Common-gender nouns and agreement according to sense require their own analysis. Do not switch a modifier solely from an English description of the referent. Predicate adjectives normally agree with their controller, but **predicate nouns retain their lexical gender and can have different number**; they are not subject to the same matching rule. Substantival adjectives do not require a written noun. Apparent exceptions should prompt investigation of attachment and construction, not an absolute 'no exceptions' slogan.
**HBS BYZ04:** **Matthew 4:5:** **τὴν ἁγίαν πόλιν**, feminine accusative singular agreement across different declension patterns. **Matthew 5:9:** **μακάριοι οἱ εἰρηνοποιοί**, masculine nominative plural predicate adjective and substantival subject. **John 1:14:** **πλήρης χάριτος καὶ ἀληθείας** is commonly related to the masculine singular **ὁ Λόγος** across the intervening material. Because attachment and the Koine behavior of πλήρης require discussion, it should not be the sole unqualified elementary proof of attributive agreement.
Article Agreement
The article normally matches its nominal expression in **gender, number, and case**. A shared surface form may encode more than one feature combination: **τοῦ** is masculine or neuter genitive singular, and **τῶν** is genitive plural across genders. Identify the whole expression rather than assigning a unique gender from the article ending alone. **τὸ παιδίον** is grammatically neuter even when its referent is a boy or girl.
Article agreement is separate from **article distribution and reference**. Distinct coordinated referents do not invariably require repeated articles, and a single article before a coordination is not universally proof of one referent. The restricted **Granville Sharp** formulation concerns two singular, personal, non-proper substantives in the same case joined by καί, with an article before the first and none before the second. Plurals, proper names, and impersonal expressions fall outside that formulation; see Chapter 43.23.
Apposition does not require each noun to carry an article. Proper names, titles, and substantival adjectives/participles have their own distributions. An article with a substantival adjective or participle agrees with that expression, but article presence alone does not classify all participial syntax. In Matthew 28:19 the repeated genitive articles belong to their respective expressions; their case, gender, and number must be distinguished from theological interpretation.
**HBS BYZ04:** **John 1:1:** **ὁ Λόγος**, masculine nominative singular article and noun. **Matthew 28:19:** **τοῦ Πατρὸς καὶ τοῦ Υἱοῦ καὶ τοῦ Ἁγίου Πνεύματος**. Each expression is genitive singular; Πατρός and Υἱοῦ are masculine, Πνεύματος neuter. The spelling **τοῦ** is shared across masculine and neuter, not one article token governing all three nouns.
Pronoun-Antecedent Agreement
A gender-inflected **anaphoric pronoun** commonly takes gender and number from its antecedent or referent, while case follows its own syntactic construction. A relative pronoun normally has the case required inside its relative clause, though case attraction and agreement according to sense qualify the pattern. First- and second-person personal pronouns do not have masculine/feminine/neuter paradigms and need not have a textual antecedent. Thus 'every pronoun agrees in gender with an antecedent' is too broad.
**Constructed relative example:** **ὁ ἄνθρωπος ὃν εἶδον**, 'the man whom I saw': **ὃν** is masculine singular accusative as the object of **εἶδον**, although **ὁ ἄνθρωπος** is nominative. A separate constructed clause, **ὁ ἄνθρωπος βλέπει αὐτόν**, normally means 'the man sees him', referring to another person. For the ordinary same-subject reflexive sense 'the man sees himself', use **ἑαυτόν**; do not teach **αὐτόν** as the unrestricted reflexive form.
In John 14:6, **δι᾽ ἐμοῦ** has a first singular genitive pronoun governed by διά; it does not exemplify grammatical-gender matching. In John 6:9, **ὃς** is nominative as subject of **ἔχει** and masculine with reference to the boy, despite neuter **παιδάριον**. Case selection and reference-based gender are separate observations. See Chapters 45 and 61.
**HBS BYZ04:** **John 14:6:** **δι᾽ ἐμοῦ**, a first singular personal pronoun in the genitive after διά; no grammatical gender is expressed. **John 6:9:** **παιδάριον ἓν ὧδε, ὃς ἔχει**, neuter noun/numeral and masculine relative. **ὃς** is nominative because it is subject in its clause, not because every relative copies its antecedent's case.
Constructio ad Sensum
**Constructio ad sensum** means agreement according to a referent's notional properties rather than direct matching of a nominal's grammatical form. It can involve the gender/number of a pronoun or modifier, or the **number of a finite verb** with a collective subject. A finite verb has no gender category, so lexical personification is not finite grammatical-gender agreement. Not every apparent mismatch belongs to this category: first check the controller, attachment, and possible antecedents.
The Johannine Spirit passages should not be introduced as though **πνεῦμα must be the grammatical antecedent of every pronoun referring to the same referent**. In **John 14:26**, masculine **ὁ παράκλητος** is specified by neuter **τὸ Πνεῦμα τὸ Ἅγιον**; **ὃ** is neuter, while the resumed **ἐκεῖνος** is masculine. Ordinary reference back to **παράκλητος** explains the masculine without requiring an exceptional gender override. **John 15:26** similarly uses masculine **ὃν** with παράκλητος, neuter **ὃ** with πνεῦμα, and masculine **ἐκεῖνος**. Different noun descriptions of the same referent have different grammatical genders.
In **John 16:13**, **ἐκεῖνος, τὸ Πνεῦμα τῆς ἀληθείας** occurs after the Helper discourse, including **παράκλητος** in 16:7. Analyze the longer reference chain rather than impose a nearest-noun rule or present personification and local 'attraction' as the only choices. The pronoun's gender is not, by itself, proof or disproof of personhood.
By contrast, **παιδάριον … ὃς** (John 6:9) clearly contrasts neuter noun form and masculine personal reference, and Matthew 21:8 illustrates notional finite-verb number with a collective subject. **τὸ πνεῦμα ὅπου θέλει πνεῖ** (John 3:8) has ordinary third singular verbs; its imagery and lexical interpretation do not create a grammatical-gender mismatch.
**HBS BYZ04:** **John 14:26** has **ὁ δὲ παράκλητος, τὸ Πνεῦμα τὸ Ἅγιον**, followed by neuter **ὃ** and masculine **ἐκεῖνος**. **John 15:26** has **ὁ παράκλητος ὃν**, then **τὸ Πνεῦμα τῆς ἀληθείας ὃ**, and masculine **ἐκεῖνος**. Ordinary anaphora to masculine παράκλητος must be considered. **John 16:13:** **ἐκεῖνος, τὸ Πνεῦμα τῆς ἀληθείας**, within the continuing Helper discourse. These forms do not alone prove a theological conclusion about personhood.
Neuter Plural Subjects
A **neuter plural subject** can occur with a **singular finite verb** in Greek. Plural finite agreement also occurs in NT Greek. These are established constructional patterns, not mistakes to be corrected from English. Animacy, lexical expression, word order, style, and context can influence distribution; number alone does not prove that an author is emphasizing a unified mass or separate individuals. No new frequency ranking is assumed here.
**John 6:9**, **ταῦτα τί ἐστιν**, has neuter plural **ταῦτα** and third singular **ἐστιν**. Translate naturally 'what are these?' without converting the Greek pronoun to singular. In **Luke 4:41**, neuter plural **δαιμόνια** occurs with singular **ἐξήρχετο** and plural participles. **James 2:19** has plural **πιστεύουσι** and **φρίσσουσι** with neuter plural subject **τὰ δαιμόνια**. The contrast demonstrates permitted patterns without supplying an automatic semantic decoding rule.
The finite singular pattern does **not** give a general license to make every agreeing adjective or participle singular. Inspect their own controllers and constructions. The traditional historical association with collective formations is background, not a proof of the present meaning of each occurrence; an unsupported detailed Indo-European reconstruction is unnecessary here. **Acts 4:32** is not an example of nominative neuter plural or collective πλῆθος controlling ἦν: **τοῦ πλήθους** is genitive. Its coordinated nominative subject is treated in 52.10.
**HBS BYZ04:** **John 6:9:** **ταῦτα τί ἐστιν εἰς τοσούτους;**, with neuter plural subject and singular verb. **Luke 4:41:** singular **ἐξήρχετο** with plural **δαιμόνια**, accompanied by neuter plural participles **κραυγάζοντα καὶ λέγοντα**. **James 2:19:** **τὰ δαιμόνια πιστεύουσι καὶ φρίσσουσι**, a neuter plural subject with plural verbs. Finite agreement patterns do not impose a singular participle on a plural controller.
Collective Nouns
A **collective noun** denotes a group. A formally singular collective can have singular agreement, while notional plural agreement is also attested, particularly with human groups. This does not make the two forms freely interchangeable in every sentence. Predicate, reference, word order, register, and discourse contribute; a number ending alone cannot establish an intended contrast between unity and individuality.
Use examples in which the collective is actually the **subject**. **John 12:17** has singular **ἐμαρτύρει** with **ὁ ὄχλος**. **Matthew 21:8** has singular **ὁ … ὄχλος** with plural **ἔστρωσαν**, illustrating agreement according to sense. Each later pronoun or participle must still be analyzed rather than assumed to inherit an automatic 'spread' of features.
In **Matthew 14:5**, **τὸν ὄχλον** is the accusative object of Herod's fear; **ἐφοβήθη** agrees with the understood Herod, not with the crowd. In **Acts 4:32**, **τοῦ πλήθους** is a genitive dependent associated with **ἡ καρδία καὶ ἡ ψυχή**. Neither example proves collective-subject number agreement. Corporate unity can be expressed by a passage's wording, but should not be inferred from a falsely identified subject.
**HBS BYZ04:** **John 12:17:** **Ἐμαρτύρει οὖν ὁ ὄχλος**, a singular collective subject with singular imperfect active indicative. **Matthew 21:8:** **ὁ δὲ πλεῖστος ὄχλος ἔστρωσαν**, a singular collective subject with third plural aorist active indicative. **Matthew 14:5's τὸν ὄχλον** is an object; **Acts 4:32's τοῦ πλήθους** is genitive. These latter forms do not control the finite verb as collective subjects.
Gender according to Sense
**Gender according to sense** occurs when a pronoun or modifier reflects the referent rather than simply copying a noun's grammatical gender. It is one kind of notional agreement. Grammatical gender and a person's sex are different properties, and personhood itself is not a masculine/feminine inflectional category. A neuter personal noun does not make its referent impersonal.
In **John 6:9**, **παιδάριον ἓν** is neuter singular, but **ὃς** is masculine singular nominative, referring to the boy. The numeral and the relative do not have the same agreement relation: the numeral matches the neuter noun locally, while the relative reflects the personal referent. Its nominative case follows its subject function with **ἔχει**. Masculine agreement is observed here, not declared obligatory for every pronoun referring to a boy described by a neuter noun.
Do not rank this the 'strongest' NT example without comparative evidence, or derive a frequency contrast with Classical Greek without a study. In the Spirit passages, ordinary anaphora to **παράκλητος** must be considered before positing exceptional agreement with **πνεῦμα** (52.5). A nearby masculine noun is not automatically an attraction trigger; it must be syntactically and referentially relevant.
**John 6:9 — HBS BYZ04:** **παιδάριον ἓν ὧδε, ὃς ἔχει πέντε ἄρτους κριθίνους**. **παιδάριον** and **ἕν** are neuter singular; relative **ὃς** is masculine singular nominative, referring to the boy and serving as subject of **ἔχει**. Natural reference explains the contrast without changing the noun's lexical gender.
Attraction and Assimilation
**Relative case attraction** occurs when a relative takes the case of its antecedent instead of the case ordinarily expected from its function inside the relative clause. Attraction is commonly into a **genitive or dative** antecedent's case; it is not defined as any agreement with a nearby noun. Ordinary relative gender/number concord is not itself case attraction. Other attraction or assimilation patterns need separate evidence and terminology.
**Mark 7:13** has **τῇ παραδόσει ὑμῶν ᾗ παρεδώκατε**. **ᾗ** is feminine singular dative, corresponding to dative **παραδόσει**, although the thing handed on would ordinarily be an accusative object of **παρεδώκατε**. This is a standard case-attraction analysis. The antecedent and relative are related syntactically, not merely adjacent words.
By contrast, in the checked **HBS BYZ04 Ephesians 1:6**, **ἐν ᾗ ἐχαρίτωσεν** has dative **ᾗ** governed by **ἐν**. **First Corinthians 11:23**, **ὃ καὶ παρέδωκα ὑμῖν**, has accusative **ὃ** as direct object; its case follows its object function and does not establish attraction. **John 6:9** has nominative **ὃς** because it is subject of **ἔχει**; its masculine reference is a separate matter.
Agreement with one conjunct, case assimilation in non-finite constructions, and anaphoric reference should not be collapsed into one unexplained process. When several analyses are plausible, identify the actual form, controller, and expected unassimilated relation before selecting a label. See Chapter 61.
**HBS BYZ04:** **Mark 7:13:** **τῇ παραδόσει ὑμῶν ᾗ παρεδώκατε**, a standard relative case-attraction example: dative ᾗ follows the antecedent's dative although the handed-on object would ordinarily be accusative. **Ephesians 1:6:** **ἐν ᾗ ἐχαρίτωσεν**, where ἐν governs the dative. **First Corinthians 11:23:** **ὃ καὶ παρέδωκα ὑμῖν**, ordinary accusative object. Case attraction must be demonstrated, not inferred from matching forms alone.
Irregular Agreement
Coordinated subjects often receive **plural agreement**, but agreement with one conjunct or a unified expression also occurs. Linear position, construction, and discourse matter; a singular verb is not by itself proof that two expressions identify one person or thing. Do not label every such case 'constructio ad finem' or assume nearest-conjunct agreement is necessarily agreement according to sense.
In **Acts 4:32**, **Τοῦ δὲ πλήθους τῶν πιστευσάντων ἦν ἡ καρδία καὶ ἡ ψυχὴ μία**, **τοῦ πλήθους** is **genitive singular**, not the subject of **ἦν**. The nominative expressions **ἡ καρδία καὶ ἡ ψυχή** form the coordinated subject; **ἦν** is singular and **μία** feminine singular. Both coordinated nouns are feminine singular, so μία's form alone cannot select καρδία rather than ψυχή as its sole controller. The expression characterizes the believers' shared heart and soul as one. The interpretation of unity follows the wording and passage, not a claim that πλήθους controls the verb.
**Galatians 3:28**, **πάντες γὰρ ὑμεῖς εἷς ἐστε**, has second plural **ἐστε** agreeing with **ὑμεῖς** and singular masculine **εἷς** in the predicate. **εἷς** is a numeral used predicatively, not an ordinary noun that has failed to agree. Compare **John 10:30**, where **ἕν** is neuter singular and **ἐσμεν** first plural. Predicate classification and verbal agreement are distinct; the full theological implications require contextual interpretation.
Unexpected forms should first be checked for text, parsing, attachment, and construction. Not every mismatch is an intentional signal of emphasis, and no catch-all 'solitary agreement' category replaces that analysis. See Chapter 53 for coordination and Chapter 65 for information structure.
**HBS BYZ04:** **Acts 4:32:** **Τοῦ δὲ πλήθους τῶν πιστευσάντων ἦν ἡ καρδία καὶ ἡ ψυχὴ μία**. The genitive phrase belongs with the coordinated nominative subject 'heart and soul'; it is not itself the subject of ἦν. **Galatians 3:28:** **πάντες γὰρ ὑμεῖς εἷς ἐστε ἐν Χριστῷ Ἰησοῦ**, plural subject/verb with singular predicative numeral. **John 10:30:** **ἐγὼ καὶ ὁ πατὴρ ἕν ἐσμεν**, first plural verb with singular neuter predicative **ἕν**.
Paradigms
Types of Agreement
These are ordinary patterns, not exceptionless attachment rules. Finite verbs lack gender; first/second-person personal pronouns lack gender paradigms; case government, notional agreement, and attraction require separate analysis. Constructed examples are labelled.
| Type | What agrees | Example |
| Subject-verb | Person and number; established constructional qualifications apply | John 12:19: οἱ Φαρισαῖοι … εἶπον (third plural) |
|---|---|---|
| Noun-adjective | Ordinary gender, number, case matching; endings and declension need not match | Matthew 4:5: τὴν ἁγίαν πόλιν (feminine accusative singular) |
| Article-noun | Gender, number, case; individual forms can be syncretic | Matthew 28:19: τοῦ Πατρός (masculine), τοῦ Ἁγίου Πνεύματος (neuter) |
| Pronoun-antecedent | Gender/number for gender-inflected anaphoric forms; case by construction; reference-based patterns also occur | Constructed: ὁ ἄνθρωπος … αὐτός (the man … he) |
| Relative pronoun-antecedent | Ordinary gender/number reference and case by relative-clause function; attraction qualifies this | Constructed: ὁ ἄνθρωπος ὃν εἶδον (the man whom I saw) |
Major Agreement Exceptions
Agreement patterns require identification of the actual controller. Finite number, modifier gender/number, predicative expressions, and relative case attraction are distinct. Morphology alone does not assign theological significance or a compulsory unity/individuality contrast.
| Phenomenon | What happens | NT example |
| Neuter plural subject | Can take singular finite agreement; plural agreement also occurs | John 6:9: ταῦτα … ἐστιν; James 2:19: τὰ δαιμόνια πιστεύουσι καὶ φρίσσουσι |
|---|---|---|
| Collective noun (ὄχλος, λαός) | Formal singular or notional plural agreement is possible; inspect the actual subject | John 12:17: ὁ ὄχλος / ἐμαρτύρει; Matthew 21:8: ὁ … ὄχλος / ἔστρωσαν |
| Constructio ad sensum | Notional gender/number with pronouns/modifiers, or number with finite verbs; finite verbs have no gender | Matthew 21:8: singular collective ὄχλος with plural ἔστρωσαν |
| Natural gender override | A gender-inflected expression can reflect a personal referent despite a neuter noun | John 6:9: παιδάριον … ὃς; the neuter noun remains neuter |
| Coordinated subjects | Resolved plural or agreement with a conjunct/unified expression; no automatic identity inference | Acts 4:32: ἦν with ἡ καρδία καὶ ἡ ψυχή; πλήθους is genitive, not subject |
| Attraction (relative) | Relative case follows antecedent case rather than its ordinary clause-internal case | Mark 7:13: παραδόσει … ᾗ παρεδώκατε; Ephesians 1:6 with ἐν is not this example |
Chapter glossary
- — agreement (concord)
syntax
Relations between grammatical features of syntactically connected expressions: finite person/number and ordinary article/adjectival gender/number/case. Constructional qualifications and reference must also be considered.- — subject-verb agreement
syntax
Finite verb agreement concerns person and number, not gender or case. Ordinary matching is qualified by established neuter-plural, collective, and coordination patterns; identify the actual subject first.- — GNC agreement
syntax
Gender, number, and case matching in ordinary article/adjectival agreement. Matching features need not have identical endings or declension class. Predicate nouns are not required to match subject gender and number like adjectives.- — constructio ad sensum
syntax
Agreement reflecting notional properties of a referent rather than direct formal matching. It can affect pronoun/modifier gender or number and finite-verb number with collectives. Finite verbs have no grammatical gender.- — neuter plural subject
syntax
A neuter plural subject can occur with a singular finite verb; plural agreement is also attested in NT Greek. Do not infer a compulsory collective meaning or automatically extend finite singular agreement to all modifiers.- — collective noun
syntax
A noun denoting a group. Formally singular collective subjects can have formal singular or notional plural agreement, subject to the actual construction and context. First distinguish the subject from a collective used as object or genitive dependent.- — gender according to sense
syntax
Pronoun or modifier gender reflecting the referent rather than the nominal's grammatical gender, as masculine ὃς with neuter παιδάριον in John 6:9. Distinguish this from ordinary reference to an alternative grammatical antecedent.- — attraction
syntax
A construction-specific assimilation of a feature. In relative case attraction, the relative adopts its antecedent's case rather than its ordinary relative-clause case; Mark 7:13 has dative ᾗ with παραδόσει. Ordinary gender/number agreement is not itself case attraction.- — pro-drop (null subject)
syntax
Greek permits unexpressed subjects. Finite endings identify person/number, while context supplies reference; overt subjects have several discourse functions and are not automatically emphatic.- — dual
syntax
An inherited inflectional number for pairs, distinct from singular and plural. It is not a productive inflectional category in NT Greek; words meaning 'two' or 'both' do not by themselves constitute dual inflection.