The pistis Christou debate

The meaning of πίστις

When approaching the meaning of a noun in a particular context, it is first helpful to ask ourselves what the noun can mean. In other words, what range of interpretations do we find with πίστις? This gives us at least some options for how to interpret it in a specific passage. In a related vein, the derivational morphology of a word can help to establish a noun’s potential interpretations (I use the words “interpretation,” “reading,” and “sense” equivalently to refer to valid meanings that the word can convey). As we discussed in Morphology B, nominalizing suffixes have their own meaning which they contribute to a word. In the case of the word πίστις, the -ις ending has its own meaning which, along with the stem πιστ to which the suffix attaches, determines the possible senses.

Given the complexity of the problem (and the questions I have received over the years about this issue), it is worth clearly laying out the morphological derivation of πίστις. Although the noun πίστις is not transparently derived from the verb πιστεύω, I argued in my original article that πίστις is a deverbal noun. There were two reasons for this. First, πίστις has the -σι suffix which does regularly convert a verb into a noun (see Morphology B for further discussion). We expect, then, πίστις to be related to an underlying verb. Second, πίστις has E-readings where it shares the argument structure of πιστεύω. Given that nouns do not normally have complex argument structures like verbs do, we expect πίστις to be derived from a verb rather than a verb derived from πίστις:

(468) πιστεύω argument structure

• οὐ μόνον τὸ εἰς αὐτὸν πιστεύειν (Phil 1:29)
• Not only to be faithful to him
• (469) πίστις argument structure (related to πιστεύω)
• τὸ στερέωμα τῆς εἰς Χριστὸν πίστεως ὑμῶν (Col 2:5)
• The steadfastness of your faithfulness to Christ

Although I believe that these two arguments demonstrate the point that there is verbal structure underneath the noun πίστις, the morphological derivation is not so straightforward. Both πίστις and πιστεύω seem to share the stem πιστ, but this stem is itself derived from the root √πειθ (as in πείθω ‘persuade’). The noun πίστις is √πειθ + -σι (n), while the verb πιστεύω is √πειθ + -το (adj) + -ευ (v). This makes the verb πιστεύω derived from the adjective πιστός, which is √πειθ + -το (adj). From a strict morphological perspective, then, it does not appear that πίστις is related to πιστεύω at all except that both share the root √πειθ (which is not itself a verb, but is a root). However, when looking at how πίστις is used, its argument structure more closely resembles the verb πιστεύω than the verb πείθω in the Koine era. Whatever the morphological relationship between the two, πίστις is best analyzed synchronically as inheriting its argument structure from πιστεύω.

Having established the morphological derivation of πίστις, what kinds of interpretations does the noun have? Besides the E-reading given above, the following examples could all arguably be considered distinct senses (more could be added, especially if we look outside of the NT):

• (470) πίστις as quality
• ὁ δὲ καρπὸς τοῦ πνεύματός ἐστιν ἀγάπη χαρὰ… πίστις (Gal 5:22)
• But the fruit of the spirit is love, joy … faithfulness
• (471) πίστις as message
• ἀνάγκην ἔσχον γράψαι ὑμῖν παρακαλῶν ἐπαγωνίζεσθαι τῇ ἅπαξ παραδοθείσῃ τοῖς ἁγίοις πίστει. (Jude 3)
• I had an obligation to write to you encouraging you to contend for the faith which has been finally passed down to the saints
• (472) πίστις as proof
• πίστιν παρασχὼν πᾶσιν ἀναστήσας αὐτὸν ἐκ νεκρῶν. (Acts 17:31)
• Presenting to all the proof of raising him from the dead

The distinctions given above are only useful insofar as we have clear definitions of the terms “quality,” “message,” and “proof.” Part of the problem in this debate has been a lack of clarity and precision in our use of terms. For example, Morgan (2015) calls πίστις an “action-nominal,” and Jensen (2025) says that I myself call πίστις an “action.” Neither of these are true. Actions are things that are done by agents; since πίστις is not a thing that is done, an action-nominal is not a good description. The noun πίστις may refer to a state, more specifically a kind of disposition that one can have towards someone else, but it does not refer to an action. Jensen says that I claim “it would be abnormal for a noun that refers to an action to have an action predicated of it, such as faith coming” (Jensen 2025). He then says that ‘faith’ is “an inner disposition and not an external action” (Jensen 2025). I say something very similar when discussing some readings of πίστις, namely the

4 The morphological mechanism that could account for this would be something like “conversion” where one word is converted into another word in a different class without transparent morphological derivation (see Bauer et al 2013 for a discussion).

readings where the arguments of the underlying verb are present (e.g. (468) above): these readings “refer to an event (or, more accurately, the state)” (Grasso 2020:133).5 Events, actions, and states are not all the same thing, and to confuse the three is to either misunderstand our own descriptions or to misunderstand the words we are trying to describe. These distinctions are not mere pedantic arguing over terminology. In our analysis, we must decide whether a given use of πίστις refers to a ‘quality’ or a ‘message,’ for example, and in order to make a case for one or the other, we need to be clear about what these terms mean. We distinguish some of these senses more precisely below.

As I point out in my article, both the subjective and objective genitive camps assume the same sense for πίστις. Both assume that πίστις refers to a state which is essentially the same as the state referred to by the verb πιστεύω. As shown above, we readily grant that this is a reading of πίστις, but as shown in detail below, it is not the correct reading of πίστις in the pistis Christou phrases. As discussed at length in Morphology B, we expect -σι nouns to have E-readings as well as other R-readings, so the fundamental question is how do we tell which sense is active in a given context. To refer to these senses and their features, I use small caps, a common way of referring to linguistic features (see Semantics A). So what evidence do we have that πίστις should be interpreted as [+state] in this specific context, and how do we determine the proper sense in an ambiguous context in the first place?7

Syntactic environment as the key disambiguating cue

The main two sides have often appealed to “context” to substantiate their claim. The subjective camp, for example, might appeal to God accomplishing his mission through time as a theme of Rom 1–4 and then point to the phrase ‘the faithfulness of Jesus Christ’ as the expression of God accomplishing that mission. On the other side, the objective genitive camp often appeals to the faith of Abraham spoken of in Rom 4 and the many other passages where Paul speaks of people having faith in God, and then this is seen as parallel to the πίστις Χριστοῦ phrase. I have called these kinds of appeals “illegitimate appeals to context” (Grasso 2020), though I myself would appeal to “context” as the primary disambiguating cue. What, then, makes an appeal to context (il)legitimate?

It is true that polysemous words are often used with the same sense in the same context. When a particular sense is used in a discourse, that sense is “activated” in working memory, and it is easier to retrieve in the subsequent discourse. For example, the word camp in the previous two paragraphs consistently means “group of people who ascribe to a view” and not one of its other meanings like “a place where people temporarily lodge, especially in the wilderness” (as in We returned to the camp at midnight). Because camp was used with a particular sense, subsequent uses of camp are assumed to have the same sense unless the context overrides that meaning. This begs the question: what in the context would “override a meaning?” Put differently, what is the fundamental disambiguating cue for polysemous words? It is not as if the use of camp as people prohibits me from using camp in another sense in the same context. Consider the following:

(473) Polysemous word camp disambiguated by syntactic environment

• One camp thought they should return to the camp that night, but the other camp thought it best to wait until the morning

In this simple example, the first and third instance of camp have the people sense, and the second instance has the place sense. What disambiguates these uses is not the larger context in which the sentence is found (we can interpret the sentence in (473) without any larger context), but the verbs with which the nouns combine. The first and third instance of camp combine with the verb thought. Since people can think but locations cannot, we automatically interpret camp thought as referring to a group of people who ascribe to a certain position. In the second example, the verb return forces us to interpret camp as referring to a location. In both cases, the verb with which the noun combines has certain restrictions on the kinds of arguments it can take with each sense, and this allows us to choose the correct sense in context. I call this the syntactic environment of a word. The table below depicts this graphically:

It is not always the case that a word’s syntactic environment provides us with enough information to disambiguate polysemous words. For example, Lauren smiled at the camp could be interpreted in either way. If it is followed with where she had stayed for two weeks, it is the [+location] sense, but if it is followed with that thought she was crazy, it is the [+people] sense. On its own, Lauren smiled at the camp does not give us enough information because Lauren can smile at either a location or a group of people. In such cases, we must resort to other disambiguating cues, such as what senses of camp have been activated in the discourse. Other surrounding words or concepts may also be associated with one of the senses and would favor one interpretation over the other (e.g. the use of S’mores and wilderness might lead to the [+location] sense, while opinion and people might lead to the [+people] sense).

Coming back to the appeals to “context” by the two camps, we can see why something like the narrative sub-structure of Galatians or Abraham’s faith in Rom 4 would only be tangentially relevant to how we interpret πίστις in a given use. These contexts are irrelevant in the sense that they are not going to disambiguate πίστις in a given syntactic environment. Even though the sentence with the three uses of camp might be about camping in general (and not what different groups of people think), this does not determine the meaning of the noun camp in the phrase camp thought. What disambiguates camp is the verb thought, regardless of the larger context. The larger context might affect which verbs Paul chooses to use with the noun πίστις, but it is ultimately the syntactic environment of πίστις that determines its meaning in any given use.

The syntactic environment of πίστις Χριστοῦ in Gal 3:21–26

To clarify where we are in the line of argumentation, we know that πίστις has multiple meanings. We have established that the most reliable way to disambiguate polysemous words is the syntactic environment. Our next step is to look at the syntactic environments of the πίστις Χριστοῦ phrase to see if the verbs/predicates that combine with πίστις Χριστοῦ disambiguate the sense of πίστις. In Gal 3:21–26, several uses of πίστις, including one with Χριστοῦ, are found:

• (474) Πίστις and πιστεύω in Gal 3:21–26
• ὁ οὖν νόμος κατὰ τῶν ἐπαγγελιῶν τοῦ θεοῦ; μὴ γένοιτο. εἰ γὰρ ἐδόθη νόμος ὁ δυνάμενος ζῳοποιῆσαι, ὄντως ἐκ νόμου ἂν ἦν ἡ δικαιοσύνη· ἀλλὰ συνέκλεισεν ἡ γραφὴ τὰ πάντα ὑπὸ ἁμαρτίαν, ἵνα ἡ ἐπαγγελία ἐκ πίστεως Ἰησοῦ Χριστοῦ δοθῇ τοῖς πιστεύουσιν. Πρὸ τοῦ δὲ ἐλθεῖν τὴν πίστιν ὑπὸ νόμον ἐφρουρούμεθα συγκλειόμενοι εἰς τὴν μέλλουσαν πίστιν ἀποκαλυφθῆναι, ὥστε ὁ νόμος παιδαγωγὸς ἡμῶν γέγονεν εἰς Χριστόν, ἵνα ἐκ πίστεως δικαιωθῶμεν· ἐλθούσης δὲ τῆς πίστεως οὐκέτι ὑπὸ παιδαγωγόν ἐσμεν. Πάντες γὰρ υἱοὶ θεοῦ ἐστε διὰ τῆς πίστεως ·
• ἐν Χριστῷ Ἰησοῦ

Is the law then contrary to God’s promises? No way! For if a law had been given which was able to make alive, righteousness would actually be from the law, but the scripture imprisoned everything under sin so that the promise from the Jesus-Christ-faith might be given to those who believe. Before the faith came, we were guarded under the law, being imprisoned so that the future faith would be revealed; therefore, the law became our guide to Christ, in order that we might be made righteous from the faith. Now that the faith has come, we are no longer under the guide. For you are all sons of God through having faith in Christ Jesus

When the meaning of polysemous words like πίστις are debated, it is best to start with the instances with the clearest disambiguating cues. Many scholars on the two dominant sides of the debate recognize that the use of πίστις in vs. 23 and 25 with the predicate ἔρχομαι (ἐλθεῖν and ἐλθούσης) is best interpreted as ‘the faith,’ referring to the message of the faith. The reason for this wide acceptance is, I suggest, that scholars implicitly recognize the principle explicated above and have applied it to the combination of ἔρχομαι and πίστις: the word ‘come’ does not allow for any other interpretation. Once this is granted, it only remains to show that the use of τὴν πίστιν in vs. 23 is a reference back to the πίστις introduced in vs. 22 in the immediately preceding clause, namely ἐκ πίστεως Ἰησοῦ Χριστοῦ. Indeed, the entire paragraph is about how God’s promises relate two different time periods, namely the time period under Law when people were under sin and the time period of the ‘faith’ when the promise from the new Jesus-Christ-faith would be given to those who have faith. Both the subjective and objective camps must posit a different meaning for πίστις when moving from Gal 3:22 to 3:23, despite there being no evidence for a change of thought or a shift in meaning of the word πίστις. The third-view position would straightforwardly regard the sense of πίστις as continuing in Gal 3:23 from 3:22 with the anaphoric use of the article in ἐλθεῖν τὴν πίστιν. Any ambiguity in the first use of πίστις in 3:22 would be immediately resolved by using it with the predicate ‘come’ in 3:23.

Moreover, the same logic that leads scholars to assign the meaning of ‘the faith’ to πίστις in Gal 3:23 should result in the same conclusion in 3:22. Besides the fact that 3:23 is continuing the argument of 3:22 where the promise is given ἐκ πίστεως (the same verb is even repeated in 3:23), the construction ἡ ἐπαγγελία ἐκ πίστεως also suggests an interpretation where the πίστις contains content. The ἐπαγγελία ‘promise’ is best interpreted as coming ‘from’ ἐκ something else that itself contains content, i.e. the ‘promise’ is part of the message of the ‘faith’.11 In other words, the message of the Christ-faith contains a promise, and it is this promise that is given to a particular group of people, namely τοῖς πιστεύουσιν. A very similar construction is found with πίστις in the genitive following the content noun ῥῆμα in Rom 10:8:12

• (475) πίστις qualifying the content of another content noun
• ἀλλὰ τί λέγει; ἐγγύς σου τὸ ῥῆμά ἐστιν ἐν τῷ στόματί σου καὶ ἐν τῇ καρδίᾳ σου, τοῦτ’ ἔστιν τὸ ῥῆμα τῆς πίστεως ὃ κηρύσσομεν (Rom 10:8)
• But what does it say? The word is near to you; it is in your mouth and in your heart, this is the word of the faith, which we preach

9 This includes Jensen (2025:14) and McFadden (2023:261), who have recently critiqued my own position, but who recognize that Gal 3:23 strongly favors the faith-as-message reading.

10 ἔρχομαι is used 74 times in everything labelled “Pauline” in the NT. In the vast majority of cases (62 of the 74), an animate subject is used to refer to physically coming to another location. In the remaining 12 instances, there are examples of events coming (i.e. occurring; Rom 3:8; Eph 5:6; Phil 1:12; Col 3:6; 1 Thess 1:10; 2 Thess 2:3), times coming (i.e. transpiring; 1 Cor 13:10; Gal 4:4; 1 Thess 5:2), and content coming (i.e. being made known; Rom 7:9). When what would otherwise be a state is used (e.g. ἡ ὀργή in Eph 5:6; Col 3:6; 1 Thess 1:10), its use with ἔρχομαι must be interpreted as the state being expressed in an event and not the state itself (the coming of God’s wrath involves him expressing that wrath, not his merely being angry). The uses of πίστις in Gal 3:23, 25 would parallel Rom 7:9 where the content word ἐντολή is said to come in the sense of being made known. Again, it makes no sense for a ‘state’ to come. If the state reading of πίστις were correct, it would have to mean that πίστις was expressed with the verb ἔρχομαι.

11 The alternative is to take ἐκ πίστεως adverbially as that which causes the promise to be given (for the causative use of ἐκ, see 12.4.3.2). This probably is the interpretation of the phrase ἐκ πίστεως δικαιωθῶμεν in Gal 3:24, but it seems less likely here. The verb δικαιόω would need an external causer (it means one entity causes someone to be righteous), which would be supplied by the ἐκ phrase, but δίδωμι would not have this requirement, it only requiring an agentive giver (implicit here), something given (the promise), and recipients (those who believe).

12 Morgan (2024:121), responding directly to my article, states “Paul would only need to specify that the content of his teaching was pistis Christou if pistis were already used to refer to a body of teaching” (italics original). These examples, as well as the example below in Gal 1:23, show that πίστις can refer to a body of teaching, and Rom 10:8 says that Paul preaches this faith. I suggest that in such contexts Ἰησοῦ Χριστοῦ is implied. It is, in fact, given explicitly in Rom 3:22, 26 as content (as well as Gal 3:22–26), but when Paul speaks elsewhere, such as here in Rom 10:8, that he is preaching “the word of the faith,” we are not to understand him as preaching the faith of Osiris or Athena. Paul is, of course, preaching the faith of the Christ.

Whereas Gal 3:22 contains a content noun (ἐπαγγελία) followed by ἐκ plus another content noun (πίστεως), specifying that the ‘promise’ is a part of the larger message of the ‘faith,’ Rom 10:8 contains a content noun followed by a genitive that specifies the nature of the content, namely the ‘faith’ (another reason for the use of ἐκ rather than the genitive in Gal 3:22 is the implicit reference to Paul’s quote of Hab 2:4, on which see below). An example like Rom 10:8 also allows us to clarify what is meant by πίστις as a message, or the [+content] sense. As something that can be preached or believed, it is true or false (for more on truth-conditions, see Semantics D). It is more than a single statement, since it can have something like a promise associated with it, though it is more than the promise itself. I have called it a ‘system of thought’, though something like ‘belief-system’ may also be a close approximation. Given that the syntactic environment of πίστις requires the sense [+content] in Gal 3:23 and that all indications in the context (syntactic or otherwise) point to interpreting πίστις in the same way in 3:22 and 3:23, it is best to take the πίστις Χριστοῦ phrase as referring to the [+content] sense. However, this does not necessarily mean that πίστις Χριστοῦ is interpreted in the same way in every other instance. Other syntactic environments may disambiguate the phrase in a distinct way.

Confirming the [+content] sense and the syntactic environment of Rom 3:26

In Teresa Morgan’s work on πίστις and ‘fides,’ she says that the meaning of πίστις as content is foreign to Paul and only develops in the early church (2015:265):

Hē pistis is sometimes translated, in certain passages of these letters, as ‘the faith’, meaning something close to what modern Christians might mean by ‘the Christian faith’ or ‘the Christian religion’: the complex of doctrines, attitudes, practices, sense of community, and more which makes up a modern worshipper’s understanding of religion. Translations which imply this meaning, however, are to be resisted. There is, notoriously, no word for ‘religion’ in Greek, Latin, or Hebrew, and little sign of the concept in Jewish, Greek, or Roman thought of this period.

She goes on to explain that it would be historically unlikely for the early Christ-followers to invent the concept of ‘faith’ as “religion” because “religion” is a modern concept. However, this unnecessarily equates one particular use of πίστις to the English word religion, and then denies a plausible sense of πίστις on that basis. To say nothing of the data, Morgan’s logic is itself anachronistic. The word πίστις might have a sense that approximates the English word religion at times without being identical to it (and without carrying all the baggage that religion conveys). Indeed, there is a plausible theoretical reason why we would expect πίστις to be able to refer to the content of what one believes. As we explained at length in Morphology B, a derived nominal like ἀποκάλυψις may refer to the event of revealing something or the thing revealed. Nominals formed from the suffix -σι are often related to underlying verbs, and Morgan herself recognizes that πίστις is one such noun (2015:31). We expect, then, that πίστις would have at least two readings: reference to the same eventuality as πιστεύω (with which it shares its E-readings) and reference to the thing ‘believed,’ i.e. the content that would follow ὅτι in the πιστεύω ὅτι ‘believe that’ construction. Given what we know about how these kinds of nouns work, that is our expectation, but is it substantiated by the data?

Besides the examples given above in Gal 3:21–26 and Rom 10:8, Gal 1:23 also requires the [+content] interpretation based on its syntactic environment:

• (476) Πίστις as content
• μόνον δὲ ἀκούοντες ἦσαν ὅτι ὁ διώκων ἡμᾶς ποτε νῦν εὐαγγελίζεται τὴν πίστιν ἥν ποτε ἐπόρθει (Gal 1:23)
• Only they heard that, ‘the one who was formerly persecuting us is now preaching the faith which he was formerly annihilating᾽

The verb εὐαγγελίζεται ‘preach good news’ has two arguments, one referring to the content of what is preached, and the other referring to the people who are preached to. Either may be found in accusative case after the verb:13

13 The hearers of the message are also sometimes found in dative case (e.g. Rom 1:15; and see the alternation in Gal 1:8–9), but the content is always in accusative in the NT. When both are present, the hearers are in dative case (e.g. Acts 13:32; 1 Cor 15:1; 2 Cor 11:7).

• (477) Εὐαγγελίζομαι with content argument
• Οἱ μὲν οὖν διασπαρέντες διῆλθον εὐαγγελιζόμενοι τὸν λόγον. (Acts 8:4)
• Those, then, who had been scattered went around preaching the word
• (478) Εὐαγγελίζομαι with hearer argument
• καὶ διερχόμενος εὐηγγελίζετο τὰς πόλεις πάσας (Acts 8:40)
• And as he was passing through, he preached (the good news) to all the cities

In light of the meaning of εὐαγγελίζομαι, the most straightforward reading of τὴν πίστιν in Gal 1:23, indeed the only real option, is to take it as referring to the content of what is preached. As we have said, this is an expected sense of πίστις, and the syntactic environment of this instance of πίστις demands this interpretation. Morgan’s thesis, then, cannot be sustained.

Galatians 1:23 is not only clear evidence for the [+content] interpretation of πίστις, but it is also evidence that πίστις can refer to a group of people. The relative pronoun ἥν refers back to πίστις, but, I suggest, the meaning is slightly different with this use. It must refer to the people who hold to the ‘faith’ and not to the faith itself. Again, it is the syntactic environment, specifically the verb πορθέω ‘annihilate,’ that demands this. This verb takes a single direct object that must refer to people (or a place where people live).15 As “anachronistic” as it may seem to some today, πίστις in Gal 1:23 must refer to the content of what is preached as well as the group of people that have adopted the truth of that content.

In light of this meaning, one more πίστις Χριστοῦ example is illuminated, namely Rom 3:26:

• (479) πίστις Ἰησοῦ in Rom 3:26
• ἐν τῇ ἀνοχῇ τοῦ θεοῦ, πρὸς τὴν ἔνδειξιν τῆς δικαιοσύνης αὐτοῦ ἐν τῷ νῦν καιρῷ, εἰς τὸ εἶναι αὐτὸν δίκαιον καὶ δικαιοῦντα τὸν ἐκ πίστεως Ἰησοῦ.
• … in the forbearance of God, to display his righteousness in the present time, in order to be righteous and the one who makes righteous the member of the Jesus-faith

The construction article + ἐκ must refer to a member of a group. When ἐκ follows the article, it can only be used in one of two ways: 1) without another noun, it refers to a member of a group (or place), or 2) with another nominal participle or noun, it modifies that participle/noun. This is consistent across all 132 uses of the construction in the LXX/NT:

• (480) Reference to a member of a group as person (a) or thing (b)
• a. ἀσπάζονται ὑμᾶς πάντες οἱ ἅγιοι, μάλιστα δὲ οἱ ἐκ τῆς Καίσαρος οἰκίας. (Phil 4:22)
• ‘All the saints greet you, but especially the members of Caesar’s house .’
• b. ὁ ἐπὶ τοῦ δώματος μὴ καταβάτω ἆραι τὰ ἐκ τῆς οἰκίας αὐτοῦ (Matt 24:17)
• Let the one on the rooftop not come down to take the things from his house

14 This is also not an uncommon use of πίστις in the NT itself. Other especially clear examples (outside of the πίστις Χριστοῦ phrases) include: Acts 13:8; 16:5; Rom 10:8; Gal 6:10; Eph 4:13; Col 2:7; Phil 1:27; 1 Tim 4:6; Jude 3. I would personally argue for others, but these examples are clear. Morgan herself translates this phrase as “the relationship of trust [between God and human beings],” but this is untenable (2015:266). The word πίστις may refer to ‘trust’, but it cannot refer to a ‘relationship of trust’ (i.e. a general category of “relationship” that is characterized by trust—‘trust’ is the relationship/concept itself referred to by πίστις). Morgan reasons that this is the meaning on the basis of the verb πορθέω (the data for which is presented in the following footnote), which she rightly says is “normally used of ravaging or destroying places or communities – towns, areas of country, or groups of people” (2015:266). This fits no better with the abstract meaning of “relationship of trust,” and it still leaves unanswered how τὴν πίστιν could serve as the object of εὐαγγελίζεται.

15 The 4 occurrences outside of Gal 1:23 in the LXX/NT all occur with a group of people: 4 Macc 4:23 (αὐτούς = οἱ Ἱεροσολυμῖται); 4 Macc 11:4 (ἡμᾶς); Acts 9:21 (τοὺς ἐπικαλουμένους); Gal 1:13 (αὐτήν = τὴν ἐκκλησίαν). A meager 4 occurrences is not enough to definitively say that the verb selects for an animate entity and to thereby establish that the 5th occurrence in Gal 1:23 also must refer to an animate entity. However, all 43 occurrences found in Josephus also conform to this pattern (though locations with inhabitants in them are also possible – but this is certainly not the meaning of πίστις): Ant 5.31 (psv = πόλις); 6.106 (χώραν); 7.103 (γῆν = Σαμαρεῖτιν); 7.290 (μητρόπολιν); 9.253 (χώραν); 10.4 (Ἱεροσόλυμα); 10.135 (Ἱεροσόλυμα); 10.237 (Ἱεροσόλυμα); 13.278 (χώραν); 13.334 (psv = χώρας); 13.338 (psv = πόλεως [implied]); 13.348 (Ἰουδαίαν); 13.421 (Ἀρμενίαν); 16.278 (χωρία, κώμας); 16.287 (αὐτήν = βασιλεία); Contra Apion 2.129 (πόλεις); Wars 1.10 (ναόν [implied]); 1.159 (πολλά [χωρία]); 1.315 (χώραν); 2.236 (psv = people); 2.239 (χώραν); 2.322 (πατρίδα); 2.339 (psv = τὰς οἰκίας); 2.351 (ὑμᾶς); 2.458 (πόλεις); 4.134 (psv = people); 4.168 (psv = οἴκων); 4.405 (χώραν); 4.442 (psv = habited places); 4.443 (πόλιν); 4.448 (ὀρεινήν); 4.534 (πόλεις); 4.537 (psv = habited places); 4.552 (Ἰδουμαίαν); 4.590 (psv = πατρίδος); 5.19 (πόλις [implied]); 5.332 (ληφθέν); 5.405 (πόλιν); 5.406 (πόλιν); 5.408 (Γαλιλαίαν); 7.26 (πόλιν); 7.44 (Ἱεροσόλυμα); 7.425 (νεών). Most places are indisputably habitations of people. There are no examples of πορθέω taking an object that refers to a relationship or an abstract concept, as Morgan’s thesis would require.

• (481) Modification of a nominal participle (a) or a noun (b)
• a. καὶ οὐδεὶς ἀναβέβηκεν εἰς τὸν οὐρανὸν εἰ μὴ ὁ ἐκ τοῦ οὐρανοῦ καταβάς, ὁ υἱὸς τοῦ ἀνθρώπου. (John 3:13)
• And no one has gone up to heaven except the one who has descended from heaven, the son of man
• b. Μωϋσῆς γὰρ γράφει τὴν δικαιοσύνην τὴν ἐκ τοῦ νόμου (Rom 10:5)
• For Moses writes of the righteousness from the law

Because the phrase τὸν ἐκ πίστεως Ἰησοῦ in Rom 3:26 refers to an argument of a verb (namely the verbal participle δικαιοῦντα) and is not modifying another noun or nominal participle, it is syntactically parallel to (480) and not (481). People are referred to with a non-neuter form of the article, while things use the neuter, so the syntactic parallel to Rom 3:26 is Phil 4:22. In that example, the genitive τῆς Καίσαρος οἰκίας ‘the household of Caesar’ supplies the group of which the person(s) referred to by the article, in this case οἱ, is/are a member. Notably, every single example where we find this syntactic construction in the LXX/NT has a group of people given in the genitive phrase.

The noun περιτομή is an instructive example. When used in this construction, it does not refer to the act or state of being circumcised or circumcision as a rite, though it may have these meanings. Rather, all examples (listed in fn 147) refer to the group of people who are circumcised. Similarly, πίστις in this construction cannot refer to the state of having faith; it must refer to the group of people who are part of the faith.

To remind the reader why we went on this digression, we said that the syntactic environment is the main disambiguating cue of any polysemous word, including πίστις. The syntactic environment of the pistis Christou phrase in Rom 3:26 is the “article + ἐκ + gen” construction, and this construction, we have just shown, has certain properties. Specifically, the noun in the genitive case must refer to a group of people. Thus, the meaning for πίστις in the phrase τὸν ἐκ πίστεως Ἰησοῦ in Rom 3:26 must be ‘the faith’ in reference to a group of people. It is the member of the Jesus-faith that God makes righteous. Crucially, the syntactic environment of this verse cannot be accounted for by either the subjective or objective camp, since both must posit that πίστις refers to a state (either the believer’s or Jesus’s) and not a group of people. If the objective camp were correct, we would expect the phrase τὸν (τὴν) πίστιν εἰς Ἰησοῦν ἔχοντα ‘the one who has faith in Jesus,’ but this is not what the text says. Despite most major translations adopting this kind of translation (e.g. ESV, NIV, NASB, etc.), the Greek construction found in the NT cannot mean this. Even if the phrase πίστεως Ἰησοῦ could be equivalent to ‘faith in Jesus’ (I have argued that this is not possible in Grasso 2020), a participle at the end of the verse would still be required, i.e. it would be τὸν (τὴν) πίστιν Ἰησοῦν ἔχοντα. Translations such as those cited earlier simply misinterpret the article + ἐκ construction which never has a following genitive that refers to a state; the noun following ἐκ always refers to a group of people.

This phrase is even more problematic for the subjective camp. The referent of τὸν cannot be Jesus, since it is the believer (or member of the Jesus-faith) that is made righteous. If the phrase ἐκ πίστεως Ἰησοῦ refers to Jesus’ faithfulness, then we are left without any description of who the person is who is ‘made righteous,’ since all we have is τὸν to refer to that person. Translations that adopt the subjective position must then supply material that is simply not there

16 The following nouns (with their references) are found with ἐκ after the article in the LXX/NT, all of which refer to a group of people (or things, which are marked with *): αἰχμαλωσία (EsdA 2:11; 5:64; 6:27; 7:6, 10, 13; 9:3, 15; Zech 6:10); ἄκρα (1 Macc 6:18; 10:7, 9; 11:41; 2 Macc 15:31); γένος (Tob 1:10); δήμος (Num 36:8); ἔθνος (EsdA 1:32; 1 Macc 11:25; 2 Macc 1:11); Ἰσραήλ (EsdA 9:37); κατοικία (EsdA 9:12); κόσμος (τοῦ οὐρανοῦ) (Deut 17:3); λαός (EsdA 5:45; Jdt 11:13; EsdA 1:13); νόμος (*EsdA 8:7); οἰκία (Ep Jer 19; Matt 24:17; Phil 4:22); περιτομή (Acts 11:2; Gal 2:12; Titus 1:10); πόλις (1 Macc 5:47; 6:3; 10:75, 76, 86; 11:46, 60); συναγωγή (Acts 6:9); υἱός (Judg 18:16); φυλή (EsdA 9:5). Proper names may also be used to refer to the members of one’s family, e.g. τῶν Ἀριστοβούλου (Rom 16:10); τῶν Ναρκίσσου (Rom 16:11).

In addition to these unambiguous references to groups of people, it is also possible to have places. Since places are often used metonymically to refer to the people from that place (and that is the sense in these contexts), these are not counterexamples, but I list the verses separately here for ease of use (without the name of each individual place): Josh 6:26; 4 Kgdms 17:24; 1 Chr 27:10, 13, 14, 27, 30; EsdA 5:18 (2x), 19 (2x), 20, 21 (2x); 1 Macc 5:23; 6:49; 8:9; 9:36. The only “exceptions” to this rule are Lev 24:10 where the noun after ἐκ refers to an individual’s mother rather than a group of people as well as adverbial phrases with ἐκ, such as ἐκ δεξιῶν (e.g. Matt 25:34) and ἐκ μέρους (1 Cor 13:10). Again, states are never found in this construction, which would be the required sense of πίστις for both the subjective and objective camps.

17 It is crucial to note the subtle difference between the hypothetical construction with the participle ἔχοντα and the construction as it stands in Rom 3:26. The fundamental difference between the two readings is whether πίστις is referring to people which have a certain property or to a property which people may have. The former understanding is the view advocated for here, and the latter is the traditional objective view, which I have said requires the construction πίστιν…ἔχοντα. It is certainly the case that the people who belong to the Jesus-faith have faith in Jesus, but even if the two groups refer to the same people, the descriptions do not mean the same thing. In Rom 3:26, the construction demands the following interpretation: ‘God makes righteous the (member of the) Jesus-faith’ and not ‘those who have faith in Jesus.’

in Greek (italics are mine and reflect what has no Greek counterpart): ‘the justifier of the one who lives because of Jesus’ faithfulness’ (NET) or ‘declares to be in the right everyone who trusts in the faithfulness of Jesus’ (Wright’s Kingdom New Testament). The reason why these translations are so different is because they are both attempts to supply something needed in the “article + ἐκ + gen” construction, namely a description of the person as the object of the verbal participle δικαιοῦντα. To put it more succinctly, both translations have entirely made up the italicized portions. Like the problems with the objective camp, this is grammatically impossible. The NET translation would require δικαιοῦντα τὸν ζῶντα ἐκ πίστεως Ἰησοῦ, and Wright’s translation would require δικαιοῦντα τὸν πιστεύοντα εἰς τὴν πίστιν τοῦ Ἰησοῦ. The bare article cannot be used to refer to an individual without a description of that individual in the article’s immediate syntactic environment (see 36.4.4); the fact that what is supplied is very different in these two translations shows that the syntactic environment does not make clear what the description of that individual is. Both translations must make something up.

At this point, scholars from both camps are ready to hit me with Hab 2:4, which is indeed a crucial text for all three camps. Space does not permit a detailed analysis, but I present here the well-known differences between Paul’s quotation of the text and the text as Paul might have seen it in the Greek or Hebrew versions of his day (chart adapted from Harris 2024; I provide a word-for-word translation to show the differences between the versions, italicizing the differences in the pronouns):

What is striking about Paul’s quotation of Hab 2:4 is that he seems to (intentionally?) omit the pronoun, either ‘my’ (in the Greek) or ‘his’ (in the Hebrew). This alone suggests an interpretation of πίστεως as something not possessed by someone or a state that holds of someone, but as something external to them – such as a message. If Paul intended the meaning of either the subjective or objective genitive in Hab 2:4, he could have easily left the pronoun in and made it clear that it referred either to the believer’s faith or the Messiah’s faithfulness. He did neither.

Since ἐκ πίστεως is probably referring back to Paul’s use of Hab 2:4 in both Gal 3:22 and Rom 3:26, it has been argued that ἐκ πίστεως is being used as shorthand to refer to whatever it refers to. I agree with this entirely, but it does not license Paul to use nonsensical Greek. The τὸν ἐκ πίστεως Ἰησοῦ construction in Rom 3:26 can only have one interpretation, namely as πίστις referring to a group of people. If he wanted to allude to Hab 2:4 while also intending the meaning posited by either the subjective or objective camps, he would have had plenty of grammatical options at his disposal. It seems, rather, that Paul changed the text of Hab 2:4 in order to communicate a message that differs from both the subjective and objective positions.

Given that we have already established that πίστις must refer to a group of people in Gal 1:23, the construction τὸν ἐκ πίστεως Ἰησοῦ may refer to a particular member of the group referred to by πίστεως Ἰησοῦ, i.e. the Jesus-faith. This is consistent with the other 58 times this construction is found in the LXX/NT, all of which conform to the pattern described above. We are then left with two meanings for the πίστις Ἰησοῦ (Χριστοῦ) phrase: in Gal 3:22 it refers to the content of the Jesus-Christ faith that has promises associated with it, and in Rom 3:26 it refers to the people who adopt that faith. Given the ease with which Paul moves between the [+content] and [+person] senses in Gal 1:23, it is no surprise that he does the same in Rom 3:22–26 where he moves from talking about the content of the Jesus-Christ faith that reveals God’s righteousness (Rom 3:22) to the person who belongs to the Jesus-faith who is made righteous by God (Rom 3:26). These readings necessarily follow from the syntactic environments in which the phrase is found, and they are consistent with how the noun πίστις is used in other places in Paul and in Koine Literature in general.

Suggested exegetical exercise

Look for disambiguating cues for one of the following polysemous words in the listed (debated) passages. Argue for a particular meaning in that passage based on the grammar:

1. τέλειος in Matt 5:48

4. ἁμαρτία in 2 Cor 5:21