Determiners and quantifiers

Functional morphemes related to nouns

When we introduced the distinction between grammatical and lexical morphology, we said that determiners and quantifiers were the functional morphemes related to nouns. Considering the kinds of things nouns refer to helps us to see the function of determiners and quantifiers. A prototypical noun like kid refers to an entity in the world, in this case a young person. The problem with a word like kid is that there are so many kids in the world. When a speaker utters something like You should scold the kid who ate the last slice of cake, how is the hearer supposed to know which kid is being referred to if there are multiple kids. In an utterance like this, we are given a description of the culprit – the particular kid to be scolded is the one who ate the last slice of cake. Since only one kid can fit the description (assuming we are referring to a specific slice of cake), we should be able to narrow down who to scold with the right information about the cake-eaters. However, if we change the sentence to You should scold a kid who ate the last slice of cake, the person we are referring to in the world suddenly changes. The most natural interpretation of this sentence is that any kid who eats the last slice of cake should be scolded; there is not one particular kid that the speaker has in mind. In other words, speakers use words like the and a to help the hearer establish who in the world is being referred to. These are determiners – they guide the hearer in determining the real-world object (the referent) of a linguistic description.

Like determiners, quantifiers also help us to establish who is being referred to. When we say something like Every kid is to blame for the cake being gone, we are now referring to each person who fits the description of kid, i.e. we are referring to a group. Of course, the speaker is probably not indicting every kid on the planet with this statement, but only a contextually-salient group (probably ones that are now bouncing off the walls). Again, changing the quantifier changes who we understand the speaker to be referencing – Some kid is to blame for the cake being gone refers to a single individual, presumably unknown to the speaker, who we should be upset with. These are quantifiers – they give information about how many real-world objects the speaker is referring to and who those are.

In the rest of this section, we introduce the basics of determiners and quantifiers, so that you can utilize the categories they encode in exegetical analysis. It should be borne in mind that there is a complex interaction between the meaning of these functional morphemes and the meaning of the lexemes they combine with. A simple illustration is with the English word the and proper names: the normally does not combine with proper names, so while the kid is a perfectly normal thing to say, the Brian is not. A significant aspect of exploring the meaning of determiners and quantifiers is noting how these functional morphemes interact with different categories of lexical items.

Determiners (the article)

The prototypical determiner is the definite article, the. It may seem odd that many, many books have been written on the meaning of this little word, but it has proven to be very difficult to provide a description of the that would account for its many different uses, let alone how near-equivalent morphemes across languages are interpreted. Greek famously exhibits some important differences. Whereas * the Brian is normally ungrammatical, the definite article is readily found with proper names in Koine Greek, e.g. ὁ Ἰησοῦς for ‘Jesus’ is the default. This already suggests that we cannot simply equate ὁ and the – our descriptions of these two functional morphemes must be more nuanced. Two basic questions occupy us in this section: 1) What is it about the and Brian that would make the two incompatible in a language like English but not in Greek; and 2) How does the Greek definite article (and demonstrative ‘this’) differ from a language like English in how it is used?

Morphemes are often incompatible for one of two reasons. Sometimes the meaning of one clashes with the meaning of another so that we cannot interpret the two together. The phrase # one balls falls into this category, since one and the plural morpheme - s have incompatible meanings (semantic incompatibility is often marked with #). At other times, one morpheme expresses a meaning that is already expressed by another, so the two are competing to express the same meaning. This is presumably what is going on with *the Brian (syntactic incompatibility is often marked with *). Lyons (1999:8) suggests that the key feature of the definite article is uniqueness. Proper names are very similar in that they refer to a unique individual; Brian does not have a meaning that would allow us to pick out all the Brians in the world, as we could with a common noun like cat. Even though Brian happens to be reused as a name so that there are multiple Brians, the name itself still serves to refer directly to a unique individual without being a description of the individual. In this way, we might say that both the and Brian have the feature of being [+unique], so adding the to Brian is disallowed by the grammar of English because it is redundant, i.e. it marks the [+unique] feature twice.

However, not all unique nouns are incompatible with the definite article, even in a language like English. A case in point is sun which is a unique noun (referring to the star in our solar system) that requires the definite article, i.e. we always say the sun .2 Comparing this to Greek, we see that ἥλιος is found both with (Acts 13:11) and without (Luke 21:25) the definite article, similar to proper names. Whereas English requires the definite article with some unique nouns and disallows it with others, Greek exhibits varied usage with the article with all kinds of unique nouns. This discussion lends support to the distinction made by Coppock and Beaver (2015) between definiteness and determinacy. According to these authors, it is best to think of definiteness as a morphological category – either the definite article is present or not. Determinacy, on the other hand, is a semantic category whereby a unique individual is being referred to. It is often assumed that definite nouns are determinate, i.e. that nouns with the definite article refer to a unique individual. While Coppock and Beaver (2015) show that this is not the case even in English, other languages such as Greek display an even looser connection between definiteness and determinacy. While the presence of the definite article often signals that the noun being referred to is unique (or determinate), having a unique referent (being determinate) does not require the presence of the definite article in Greek. In other words, a bare singular noun (traditionally called “anarthrous”) like ἄνθρωπος may be equivalent to ‘the person,’ ‘a person,’ or ‘person’ in English.

At this point, we could get into a very lengthy discussion on what precisely the article means and how to define key terms in the debate like “uniqueness” and “referentiality” (for a survey, see Abbott 2010:chs. 1, 2, 6). We must content ourselves, however, with our simple definition of definiteness with a slight modification: “The definite article presupposes that the thing being referred to is unique in the discourse context.” The idea that the presupposed something was introduced by the philosopher Peter Strawson (1952) who pointed out that the use of the word the does not assert the existence of an entity. When the article is used with υἱός in the sentence ὁ υἱὸς μένει εἰς τὸν αἰῶνα ‘the son remains forever’ (John 8:35), Jesus is not making the claim (i.e. asserting) that the son exists and that he remains forever. The only claim being made is that the son remains forever. The existence of the son is taken for granted by Jesus and a characteristic of him is given, namely that he remains forever. The definite article, then, says something like “this noun has a unique reference in this context, and I (as the speaker) am presupposing this identification.”

We have been talking about the definite article as if it had a meaning that was uniform across languages, while at the same time, we have noted the many differences in usage between Greek and English. That the article would have at least a fairly stable meaning across languages is expected if it conveyed a logical meaning that was a foundational part of communication, such as establishing reference (as we have suggested). How do we reconcile these facts? We will assume that the and ὁ essentially mean the same thing, but they are used differently because Greek and English, as language systems, are different.

First, Greek allows reference to individuals with the bare singular, whereas English does not (see 5.2.3; our example with ἥλιος also assumes this). This results in the requirement for the article for unique nouns with content (like sun), but no such requirement holds in Greek not because ὁ is different than the but the bare singular in Greek differs from the bare singular in English. Second, Greek can mark definiteness simultaneously with other functional categories, while this is impossible in English. For example, the article co-occurs with demonstratives, οὗτος ὁ ἄνθρωπος, and possessives, τὸ ποτήριόν σου in Greek, but this is impossible in English where this and your cannot be used with the. Again, this is not necessarily a result of a meaning difference between the and ὁ, but a syntactic difference between the two. In English, the moves to a position in the syntax that this also occupies, blocking the two from co-occurring, but in Greek, we do not see this movement of ὁ. Third, Greek allows for “determiner spreading” whereas English does not. This is where Greek double-marks definiteness in some noun-adjective combinations, such as τὸ βιβλίον τὸ ἅγιον; again, this is impossible in English. At least in Modern Greek, this construction implies contrast with other elements (Alexiadou 2014:20–28), and a case could be made that this same meaning holds in NT Greek as well. English marks a similar kind of contrast with stress (e.g. the holy book) rather than the repetition of the article and the adjective after the noun, but again, this is a difference between Greek and English and not a difference between the articles in the respective languages. Finally, ὁ can be used in some environments without the noun it modifies, such as in the common phrase ὁ δὲ εἶπεν when the subject switches to a person referred to by the masculine article. Notice, however, that ὁ δὲ εἶπεν conveys more information than the (literal) equivalent but the said. Greek has a grammatical gender system that is built into the article (which English lacks), and this allows it to be used to refer to a known entity with a particular gender and without the accompanying noun. Because English lacks gender in its article and nominal system, there would be no way for the hearer to determine who was being referred to with the bare article, so it lacks this use. Even here, however, the same basic meaning of the article we have given applies: ὁ refers to a unique masculine person in the context who is contrasted with someone else.

What remains in our discussion of the definite article is to go through all the various uses and to describe why the article can be used in such and such a way and how it relates back to the meaning we have provided of uniqueness. Space precludes us from doing that (though see Peters 2014 for a recent full-length treatment of the Greek article), but it is worth pointing out that there are many different glosses that can be given to the definite article as we move from Greek to English. Köstenberger, Merkle, and Plummer (2020:157), for example, list the article as sometimes equivalent to the following pronouns: personal, relative, possessive, demonstrative, and “alternate”. These are not necessarily different “meanings” of the Greek article but different English equivalents. The range of glosses, however, should instill caution in us when we move from Greek to another language – equations such as ὁ = the are far too simplistic and quickly lead to errors.

One use of the article that is commonly noted in linguistics but is not normally discussed in NT Greek grammars is to signal argument status. Nouns used as arguments of verbs are often found with the article, and some languages require this (de Swart and Zwarts 2008). Greek grammars do often note that the article is regularly absent in prepositional phrases (Wallace 1996:247), but this is not connected with the noun’s status as an argument. Insofar as nouns in prepositional phrases are often adjuncts (not required by the verb) rather than arguments (required by the verb), this tendency can be explained by ὁ marking argumenthood. Because verbs often refer to specific happenings in the world, languages often require specific (determinate) entities in those specific happenings. An example from the Didache 7:1, 3 is particularly striking in this regard:4

• (184) 1 ταῦτα πάντα προειπόντες βαπτίσατε εἰς τὸ ὄνομα τοῦ πατρὸς καὶ τοῦ υἱοῦ καὶ τοῦ ἁγίου πνεύματος ἐν ὕδατι ζῶντι…3 ἐὰν δὲ ἀμφότερα μὴ ἔχῃς, ἔκχεον εἰς τὴν κεφαλὴν τρὶς ὕδωρ εἰς ὄνομα Πατρὸς καὶ Υἱοῦ καὶ ἁγίου Πνεύματος
• ‘Saying all this first, baptize into the name of the father and the son and the holy spirit in running water … but if you don’t have either, pour water three times on the head for the name of the father and the son and the holy spirit.’

In this example, the phrase εἰς (τὸ) ὄνομα (τοῦ) πατρὸς καὶ (τοῦ) υἱοῦ καὶ (τοῦ) ἁγίου πνεύματος appears twice, the first time with the article and the second time without it. What makes this a particularly good example is the second occurrence of the phrase lacks the article while the first has it. If the second had the article and the first lacked it, the easy explanation would be that the second instance of the phrase was the anaphoric use of the article, referring back to something previously mentioned in the discourse. Since the second use lacks the article, this cannot be the explanation. Careful attention to the semantics of the verbs and the criterion of argumenthood does provide an explanation. The verb βαπτίζω takes three arguments: a baptizer, a baptizee, and something that the baptizee is ‘immersed’ into. The phrase εἰς τὸ ὄνομα… represents the thing baptized into and is, thus, an argument of the verb, even if it is metaphorical. The means of the baptism, given in the phrase ἐν ὕδατι ζῶντι, is an adjunct and is not a necessary part of the verb (more on the special case of ὕδωρ below). Similarly, the verb ἐκχέω in verse 3 takes three arguments: a pourer, something poured, and a location where the thing poured goes (represented by εἰς). The location argument is given first in the phrase εἰς τὴν κεφαλὴν and is found with the article. Next, we have the thing poured, namely ὕδωρ, which is lacking the article. However, as a mass noun rather than a count noun, this is to be expected. There is no specific instance of water that would be poured onto the head of the person; it is any water. The one pouring out is implicit, which means that the final phrase εἰς ὄνομα… cannot be an argument in vs. 3 – it must be an adjunct meaning something like ‘for the name …’ This adjunct lacks the article even though the name of the Father, Son, and Holy Spirit are known entities just mentioned in the discourse and refer to specific individuals. Verse 1, therefore, has the article with εἰς τὸ ὄνομα… because the prepositional phrase is an argument of the verb βαπτίζω, and verse 3 lacks the article with εἰς ὄνομα… because the same prepositional phrase is used as an adjunct of the verb ἐκχέω.

One final difference between the two prepositional phrases is worth pointing out. The definiteness of πατρός is dependent upon the definiteness of ὄνομα, which occurs immediately before it. This is in accordance with Apollonius’s rule, which states that “when genitives occur with the article in possessive constructions, the possessed nouns also take the article” (Householder 1981:72).5 In Rodney Whitacre’s grammar (2021), he provides the following minimal pair to demonstrate the rule:

• (185) ἐν τῷ λόγῳ τῆς ἀληθείας

In the word of truth (Col 1:5)

• (186) ἐν λόγῳ ἀληθείας
• In truthful speech (2 Cor 6:7)

According to the rule, the definiteness of ἀληθείας is dependent upon the definiteness of λόγῳ. In Col 1:5, ἀληθείας is definite, since λόγῳ is also definite, and the opposite holds in 2 Cor 6:7 for the same reason. This kind of minimal pair does not answer the question of why we would have the article in Col 1:5 and not in 2 Cor 6:7 (which we have said often has to do with the function of the entire prepositional phrase in the sentence), but it further demonstrates how the use of the article is affected by the surrounding words and phrases. Depending on what counts as “possession,” this “rule” may also be easily disproven, but there is a general tendency for genitives to be definite when the nouns they modify are definite (and to be anarthrous when the modified noun is anarthrous).

In light of the above discussions, one final example is worth considering, one that has serious implications for exegesis and theology. I belabor this point because the article is, more than any other Greek grammatical feature, routinely abused by theologians. Consider Paula Fredriksen’s discussion of Phil 2:6–11 (2017:138):

The Greek, however, does not quite correspond to the RSV’s English. In Philippians 2.6, Jesus is not “in the form of the [high] God,” but in the form of “[a] god.” Jesus does not demur from equality with God the Father, but from “god-status” or, closer to Paul’s word choice, equality with “[a] god.” The god who exalts Jesus in verse 9, by contrast, is the high god (ho theos, the god), referred to as “God the Father” in verse 11. The conventions of English capitalization – “God” with the upper-case G in all clauses – obscures Paul’s Greek. Paul distinguishes between degrees of divinity here. Jesus is not “God.”

Fredriksen’s claim is fundamentally a grammatical claim about the Greek data. Her equations are as follows: ὁ θεός = the (high) God (= YHWH) and θεός = (a) god = god-status. To falsify her claim, there would have to be unambiguous examples of θεός (without the article) referring to YHWH, the high God. Finding such an example would mean that θεός in Phil 2:6 need not refer to ‘a god’, but might also refer to ‘the God.’ Thankfully, we do not have to look far for an unambiguous reference to “the high God” without the article – Phil 2:11 concludes with εἰς δόξαν θεοῦ πατρός. Fredriksen herself translates

5 Apollonius, a 2nd century grammarian, wrote the first systematic grammar of ancient Greek, or any other language.

this as “God the Father” (italics added)! However, we might want to understand the phrase ἐν μορφῇ θεοῦ ὑπάρχων in Phil 2:6, it is indeed false that the anarthrous θεοῦ cannot mean “the high God.”6

In fact, the preceding discussion gives all the syntactic reasons why we should expect reference to “the high God” to be anarthrous in both ἐν μορφῇ θεοῦ ὑπάρχων and εἰς δόξαν θεοῦ πατρός. We a priori expect there to be a similar definiteness status between the genitive and the noun it modifies. This holds in both phrases. The phrase εἰς δόξαν θεοῦ πατρός is clearly an adjunct, so we expect it to be a bare singular, though it is still determinate (referring to a particular individual, namely “the high God”). The phrase ἐν μορφῇ θεοῦ is serving as a predicate to the verb ὑπάρχω (on which, see 24.3.2). Whenever a noun serves as the predicate of this verb, it is always anarthrous in the NT, including after a prepositional phrase as in Phil 2:6 (ἐν μορφῇ θεοῦ). We should expect, then, both μορφῇ and θεοῦ to be anarthrous, but given that Paul nowhere else refers to a lesser deity as θεός, it is also best to take θεοῦ as determinate, i.e. referring to “the high God.” Fredriksen’s argument is not supported by the data.

Determiners (the demonstrative)

The preceding discussion has focused on the prototypical determiner, the definite article. The other major determiner is the demonstrative, i.e. this, that, these, and those. Like the definite article, demonstratives are used to establish the reference of the noun. Demonstratives differ from definite articles in that they can act like pronouns and replace the entire noun. A speaker can point to a rock and say That is hard where the word that refers directly to something in the world and does not need another noun to go with it. This use of the demonstrative is also found in Greek: when Jesus says τοῦτό ἐστιν τὸ σῶμά μου ‘this is my body’ (Matt 26:26), he is using the demonstrative to point to the thing in his hand, namely the bread. This use makes demonstratives unique: they can refer directly to something in the world (often with an accompanying gesture to indicate what is being pointed at). In contrast, if Jesus were to have used the article and said ὁ ἄρτος ἐστὶν τὸ σῶμά μου ‘the bread is my body,’ he would have been referring to the thing in his hand by describing it as bread. By using τοῦτο, Jesus says that whatever he happens to be pointing to is his body; it so happened to be bread, but he does not describe it as such when he uses the demonstrative. Of course, the demonstrative may be accompanied by a noun (Jesus could have said οὗτος ὁ ἄρτος ‘this bread’), but even in such a case, it is a particular piece of bread in the world of the speaker that is being directly referenced.

As we know, the demonstrative pronoun regularly co-occurs with the definite article in Greek, and it does so in one of two syntactic positions: either come before the article and noun or after both. This same syntactic contrast is also found in Modern Greek and seems to be associated with a difference in meaning: the pre-article demonstrative is typically used to point to things in the world, while the post-nominal demonstrative is more readily used to point to things in the discourse (see Alexiadou, Haegeman, & Stavrou 2007:120–127). These are only tendencies in Modern Greek, however, and even if these tendencies hold in NT Greek, there are clear examples for both world-reference and discourse-reference with the demonstrative before and after the noun phrase:

(187) Post-nom world

• ἔστησάν τε μάρτυρας ψευδεῖς λέγοντας· ὁ ἄνθρωπος οὗτος οὐ παύεται λαλῶν ῥήματα…
• And they presented false witnesses who said, ‘ This man does not stop speaking words …᾽ (Acts 6:13)

(188) Pre-art world

• οἳ καὶ εἶπαν· ἄνδρες Γαλιλαῖοι, τί ἑστήκατε ἐμβλέποντες εἰς τὸν οὐρανόν; οὗτος ὁ Ἰησοῦς ὁ ἀναλημφθεὶς ἀφ’ ὑμῶν εἰς τὸν οὐρανὸν

Who also said, ‘Galilean men, why are you standing staring into heaven? This Jesus, the one taken from you into heaven …’ (Acts 1:11)

(189) Post-nom discourse

τίς ἐστιν ὁ λόγος οὗτος ὃν εἶπεν

What is this word which you spoke? (John 7:36)

(190) Pre-nom discourse

ἐξῆλθεν οὖν οὗτος ὁ λόγος εἰς τοὺς ἀδελφοὺς ὅτι ὁ μαθητὴς ἐκεῖνος οὐκ ἀποθνῄσκει

So this word went out to the brothers that that disciple is not dying (John 21:23)

Syntax often reflects a difference in meaning, and this should largely be the default assumption. It could be that NT Greek exhibits the same contrast as Modern Greek, but there are reasonable explanations for the counterexamples. More research is needed in this area.

Quantifiers

If determiners help to establish reference by picking out individuals, quantifiers do so by specifying how many individuals are being discussed. The prototypical examples are words like all, most, and some. In a sentence like All/Most/Some disciples cast out demons we are dividing those who can be referred to by disciples into two groups, namely a casting-out-demons group and a non-casting-out-demons group. With the quantifier all, we are saying that of the relevant disciples under discussion, every individual is in the casting-out-demons group, and there are no individuals in the non-casting-out-demons group. With most, the assertion would be that there are more in the casting-out-demons group than in the non-casting-out-demons group, and some would specify that there is at least one member in the casting-out-demons group (with an implication that more are in the non-casting-out-demons group). In Greek, these basic ideas are expressed with πᾶς (‘all’), πλείων (‘most’), and τὶς (‘some’), though there are, of course, differences between the Greek and English words that express these ideas.

We will discuss πᾶς in order to illustrate some of the issues involved when analyzing quantifiers. First, and what is true of all quantifiers, πᾶς always makes reference to a particular domain. In other words, πᾶς does not mean every single thing – it means all the individuals that are being discussed, which is the domain of the quantifier. In the sentence All disciples cast out demons, the individuals being referred to must be disciples, and this hints at the most important word for domain restriction: quantifiers are restricted by the noun they immediately combine with, in this case disciples. Non-disciples are not included in the assertion made in this utterance. However, it is not even enough to be a disciple, but the individuals must also be the contextually-relevant disciples. You probably assumed, for example, that the disciples being discussed are Jesus’ disciples and not Rabbi Akiva’s, even though the latter would also count as disciples. Restricting the domain in this way is done implicitly with quantifiers of all kinds – we intuitively use the context to determine which group the quantifier applies to.

While domain restriction is common to all quantifiers, some properties are unique to words like πᾶς, which would be labelled a universal quantifier because it specifies that the description applies universally to every entity in the domain. Despite this, it is not a contradiction to say All disciples cast out demons except for Judas. The fact that exception clauses are not contradictory means that all allows for exceptions, and indeed, we regularly interpret πᾶς to refer to most or a large number of a group and not necessarily every single individual. When Mark 1:5 says καὶ ἐξεπορεύετο πρὸς αὐτὸν πᾶσα ἡ Ἰουδαία χώρα καὶ οἱ Ἱεροσολυμῖται πάντες ‘and all the land of Judah and all the Jerusalemites went out to him,’ we can safely assume that every single individual in Judah and Jerusalem did not go out to see John. The universal quantifiers are probably allowing for many exceptions and is best interpreted as something closer to a lot (we use all similarly in English all the time …). This is not to say that πᾶς should always be interpreted this way, only that this word in natural language is not used equivalently to a universal quantifier in symbolic logic, which would not allow for this kind of interpretation.

On the other end of the spectrum, existential quantifiers say that at least one relevant entity satisfies the description. To say Some disciple(s) cast out demons means that at least one disciple casts out demons. In Greek, this idea is expressed with τὶς. Because the meaning of an existential quantifier specifies that one (or more) entity fits the description, it is often used to introduce entities into the discourse. For example, Jesus begins the parable of the Good Samaritan with ἄνθρωπός τις κατέβαινεν ἀπὸ Ἰερουσαλὴμ εἰς Ἰεριχώ ‘A man was going down from Jerusalem to Jericho.’ This is the start of the parable and the first mention of the man who was traveling. The addition of τις introduces a previously unknown man into the discourse. In a language like English, this is best approximated with the indefinite article a, which also serves a similar function.

Conclusion

Many more distinctions could be made about different kinds of quantifiers and their functions, and doing so would allow us to speak more carefully about how quantifiers interact with the rest of the grammar. To use the categories we developed for determiners and quantifiers in the analysis of a specific text, we would want to look more closely at how they interact with different elements in the grammar. Indeed, that is our task in Exegesis B.